psqrt14()

real function psqrt14	(	character	trans,
		integer	m,
		integer	n,
		integer	nrhs,
		real, dimension( * )	a,
		integer	ia,
		integer	ja,
		integer, dimension( * )	desca,
		real, dimension( * )	x,
		integer	ix,
		integer	jx,
		integer, dimension( * )	descx,
		real, dimension( * )	work
	)

Definition at line 1 of file psqrt14.f.

*
*  -- ScaLAPACK routine (version 1.7) --
*     University of Tennessee, Knoxville, Oak Ridge National Laboratory,
*     and University of California, Berkeley.
*     May 1, 1997
*
*     .. Scalar Arguments ..
      CHARACTER          TRANS
      INTEGER            IA, IX, JA, JX, M, N, NRHS
*     ..
*     .. Array Arguments ..
      INTEGER            DESCA( * ), DESCX( * )
      REAL               A( * ), WORK( * ), X( * )
*     ..
*
*  Purpose
*  =======
*
*  PSQRT14 checks whether sub( X ) is in the row space of sub( A ) or
*  sub( A )', where sub( A ) denotes A( IA:IA+M-1, JA:JA+N-1 ) and
*  sub( X ) denotes X( IX:IX+N-1, JX:JX+NRHS-1 ) if TRANS = 'N', and
*  X( IX:IX+N-1, JX:JX+NRHS-1 ) otherwise.  It does so by scaling both
*  sub( X ) and sub( A ) such that their norms are in the range
*  [sqrt(eps), 1/sqrt(eps)], then computing an LQ factorization of
*  [sub( A )',sub( X )]' (if TRANS = 'N') or a QR factorization of
*  [sub( A ),sub( X )] otherwise, and returning the norm of the trailing
*  triangle, scaled by MAX(M,N,NRHS)*eps.
*
*  Notes
*  =====
*
*  Each global data object is described by an associated description
*  vector.  This vector stores the information required to establish
*  the mapping between an object element and its corresponding process
*  and memory location.
*
*  Let A be a generic term for any 2D block cyclicly distributed array.
*  Such a global array has an associated description vector DESCA.
*  In the following comments, the character _ should be read as
*  "of the global array".
*
*  NOTATION        STORED IN      EXPLANATION
*  --------------- -------------- --------------------------------------
*  DTYPE_A(global) DESCA( DTYPE_ )The descriptor type.  In this case,
*                                 DTYPE_A = 1.
*  CTXT_A (global) DESCA( CTXT_ ) The BLACS context handle, indicating
*                                 the BLACS process grid A is distribu-
*                                 ted over. The context itself is glo-
*                                 bal, but the handle (the integer
*                                 value) may vary.
*  M_A    (global) DESCA( M_ )    The number of rows in the global
*                                 array A.
*  N_A    (global) DESCA( N_ )    The number of columns in the global
*                                 array A.
*  MB_A   (global) DESCA( MB_ )   The blocking factor used to distribute
*                                 the rows of the array.
*  NB_A   (global) DESCA( NB_ )   The blocking factor used to distribute
*                                 the columns of the array.
*  RSRC_A (global) DESCA( RSRC_ ) The process row over which the first
*                                 row of the array A is distributed.
*  CSRC_A (global) DESCA( CSRC_ ) The process column over which the
*                                 first column of the array A is
*                                 distributed.
*  LLD_A  (local)  DESCA( LLD_ )  The leading dimension of the local
*                                 array.  LLD_A >= MAX(1,LOCr(M_A)).
*
*  Let K be the number of rows or columns of a distributed matrix,
*  and assume that its process grid has dimension p x q.
*  LOCr( K ) denotes the number of elements of K that a process
*  would receive if K were distributed over the p processes of its
*  process column.
*  Similarly, LOCc( K ) denotes the number of elements of K that a
*  process would receive if K were distributed over the q processes of
*  its process row.
*  The values of LOCr() and LOCc() may be determined via a call to the
*  ScaLAPACK tool function, NUMROC:
*          LOCr( M ) = NUMROC( M, MB_A, MYROW, RSRC_A, NPROW ),
*          LOCc( N ) = NUMROC( N, NB_A, MYCOL, CSRC_A, NPCOL ).
*  An upper bound for these quantities may be computed by:
*          LOCr( M ) <= ceil( ceil(M/MB_A)/NPROW )*MB_A
*          LOCc( N ) <= ceil( ceil(N/NB_A)/NPCOL )*NB_A
*
*  Arguments
*  =========
*
*  TRANS   (global input) CHARACTER*1
*          = 'N':  No transpose, check for sub( X ) in the row space of
*                  sub( A ),
*          = 'T':  Transpose, check for sub( X ) in row space of
*                  sub( A )'.
*
*  M       (global input) INTEGER
*          The number of rows to be operated on, i.e. the number of rows
*          of the distributed submatrix sub( A ). M >= 0.
*
*  N       (global input) INTEGER
*          The number of columns to be operated on, i.e. the number of
*          columns of the distributed submatrix sub( A ). N >= 0.
*
*  NRHS    (global input) INTEGER
*          The number of right hand sides, i.e., the number of columns
*          of the distributed submatrix sub( X ). NRHS >= 0.
*
*  A       (local input) REAL pointer into the local memory
*          to an array of dimension (LLD_A, LOCc(JA+N-1)). This array
*          contains the local pieces of the M-by-N distributed matrix
*          sub( A ).
*
*  IA      (global input) INTEGER
*          The row index in the global array A indicating the first
*          row of sub( A ).
*
*  JA      (global input) INTEGER
*          The column index in the global array A indicating the
*          first column of sub( A ).
*
*  DESCA   (global and local input) INTEGER array of dimension DLEN_.
*          The array descriptor for the distributed matrix A.
*
*  X       (local input) REAL pointer into the local
*          memory to an array of dimension (LLD_X,LOCc(JX+NRHS-1)).
*          On entry, this array contains the local pieces of the
*          N-by-NRHS distributed submatrix sub( X ) if TRANS = 'N',
*          and the M-by-NRHS distributed submatrix sub( X ) otherwise.
*
*  IX      (global input) INTEGER
*          The row index in the global array X indicating the first
*          row of sub( X ).
*
*  JX      (global input) INTEGER
*          The column index in the global array X indicating the
*          first column of sub( X ).
*
*  DESCX   (global and local input) INTEGER array of dimension DLEN_.
*          The array descriptor for the distributed matrix X.
*
*  WORK    (local workspace) REAL array dimension (LWORK)
*          If TRANS='N', LWORK >= MNRHSP * NQ + LTAU + LWF and
*          LWORK >= MP * NNRHSQ + LTAU + LWF otherwise, where
*
*          IF TRANS='N', (LQ fact)
*            MNRHSP = NUMROC( M+NRHS+IROFFA, MB_A, MYROW, IAROW,
*                             NPROW )
*            LTAU   = NUMROC( IA+MIN( M+NRHS, N )-1, MB_A, MYROW,
*                             RSRC_A, NPROW )
*            LWF    = MB_A * ( MB_A + MNRHSP + NQ0 )
*          ELSE         (QR fact)
*            NNRHSQ = NUMROC( N+NRHS+ICOFFA, NB_A, MYCOL, IACOL,
*                             NPCOL )
*            LTAU   = NUMROC( JA+MIN( M, N+NRHS )-1, NB_A, MYCOL,
*                             CSRC_A, NPCOL )
*            LWF    = NB_A * ( NB_A + MP0 + NNRHSQ )
*          END IF
*
*          and,
*
*          IROFFA = MOD( IA-1, MB_A ), ICOFFA = MOD( JA-1, NB_A ),
*          IAROW = INDXG2P( IA, MB_A, MYROW, RSRC_A, NPROW ),
*          IACOL = INDXG2P( JA, NB_A, MYCOL, CSRC_A, NPCOL ),
*          MP0 = NUMROC( M+IROFFA, MB_A, MYROW, IAROW, NPROW ),
*          NQ0 = NUMROC( N+ICOFFA, NB_A, MYCOL, IACOL, NPCOL ).
*
*          INDXG2P and NUMROC are ScaLAPACK tool functions;
*          MYROW, MYCOL, NPROW and NPCOL can be determined by calling
*          the subroutine BLACS_GRIDINFO.
*
*
*  =====================================================================
*
*     .. Parameters ..
      INTEGER            BLOCK_CYCLIC_2D, CSRC_, CTXT_, DLEN_, DTYPE_,
     $                   LLD_, MB_, M_, NB_, N_, RSRC_
      parameter( block_cyclic_2d = 1, dlen_ = 9, dtype_ = 1,
     $                     ctxt_ = 2, m_ = 3, n_ = 4, mb_ = 5, nb_ = 6,
     $                     rsrc_ = 7, csrc_ = 8, lld_ = 9 )
      REAL               ONE, ZERO
      parameter( zero = 0.0e+0, one = 1.0e+0 )
*     ..
*     .. Local Scalars ..
      LOGICAL            TPSD
      INTEGER            IACOL, IAROW, ICOFFA, ICTXT, IDUM, IIA, INFO,
     $                   IPTAU, IPW, IPWA, IROFFA, IWA, IWX, J, JJA,
     $                   JWA, JWX, LDW, LWORK, MPWA, MPW, MQW, MYCOL,
     $                   MYROW, NPCOL, NPROW, NPW, NQWA, NQW
      REAL               AMAX, ANRM, ERR, XNRM
*     ..
*     .. Local Arrays ..
      INTEGER            DESCW( DLEN_ ), IDUM1( 1 ), IDUM2( 1 )
      REAL               RWORK( 1 )
*     ..
*     .. External Functions ..
      LOGICAL            LSAME
      INTEGER            NUMROC
      REAL               PSLANGE, PSLAMCH
      EXTERNAL           lsame, numroc, pslange, pslamch
*     ..
*     .. External Subroutines ..
      EXTERNAL           blacs_gridinfo, descset, infog2l, psamax,
     $                   pscopy, psgelqf, psgeqrf, pslacpy,
     $                   pslascl, pxerbla, sgamx2d
*     ..
*     .. Intrinsic Functions ..
      INTRINSIC          abs, max, min, mod, real
*     ..
*     .. Executable Statements ..
*
*     Get grid parameters
*
      ictxt = desca( ctxt_ )
      CALL blacs_gridinfo( ictxt, nprow, npcol, myrow, mycol )
*
      psqrt14 = zero
*
      ipwa = 1
      iroffa = mod( ia-1, desca( mb_ ) )
      icoffa = mod( ja-1, desca( nb_ ) )
      iwa = iroffa + 1
      jwa = icoffa + 1
      CALL infog2l( ia, ja, desca, nprow, npcol, myrow, mycol, iia,
     $              jja, iarow, iacol )
      mpwa = numroc( m+iroffa, desca( mb_ ), myrow, iarow, nprow )
      nqwa = numroc( n+icoffa, desca( nb_ ), mycol, iacol, npcol )
*
      info = 0
      IF( lsame( trans, 'N' ) ) THEN
         IF( n.LE.0 .OR. nrhs.LE.0 )
     $      RETURN
         tpsd = .false.
         mpw = numroc( m+nrhs+iroffa, desca( mb_ ), myrow, iarow,
     $                 nprow )
         nqw = nqwa
*
*        Assign descriptor DESCW for workspace WORK and pointers to
*        matrices sub( A ) and sub( X ) in workspace
*
         iwx = iwa + m
         jwx = jwa
         ldw = max( 1, mpw )
         CALL descset( descw, m+nrhs+iroffa, n+icoffa, desca( mb_ ),
     $                 desca( nb_ ), iarow, iacol, ictxt, ldw )
*
      ELSE IF( lsame( trans, 'T' ) ) THEN
         IF( m.LE.0 .OR. nrhs.LE.0 )
     $      RETURN
         tpsd = .true.
         mpw = mpwa
         nqw = numroc( n+nrhs+icoffa, desca( nb_ ), mycol, iacol,
     $                 npcol )
*
*        Assign descriptor DESCW for workspace WORK and pointers to
*        matrices sub( A ) and sub( X ) in workspace
*
         iwx = iwa
         jwx = jwa + n
         ldw = max( 1, mpw )
         CALL descset( descw, m+iroffa, n+nrhs+icoffa, desca( mb_ ),
     $                 desca( nb_ ), iarow, iacol, ictxt, ldw )
      ELSE
         CALL pxerbla( ictxt, 'PSQRT14', -1 )
         RETURN
      END IF
*
*     Copy and scale sub( A )
*
      iptau = ipwa + mpw*nqw
      CALL pslacpy( 'All', m, n, a, ia, ja, desca, work( ipwa ), iwa,
     $              jwa, descw )
      rwork( 1 ) = zero
      anrm = pslange( 'M', m, n, work( ipwa ), iwa, jwa, descw, rwork )
      IF( anrm.NE.zero )
     $   CALL pslascl( 'G', anrm, one, m, n, work( ipwa ), iwa,
     $                 jwa, descw, info )
*
*     Copy sub( X ) or sub( X )' into the right place and scale it
*
      IF( tpsd ) THEN
*
*        Copy sub( X ) into columns jwa+n:jwa+n+nrhs-1 of work
*
         DO 10 j = 1, nrhs
            CALL pscopy( m, x, ix, jx+j-1, descx, 1, work( ipwa ), iwx,
     $                   jwx+j-1, descw, 1 )
   10    CONTINUE
         xnrm = pslange( 'M', m, nrhs, work( ipwa ), iwx, jwx, descw,
     $                   rwork )
         IF( xnrm.NE.zero )
     $      CALL pslascl( 'G', xnrm, one, m, nrhs, work( ipwa ), iwx,
     $                    jwx, descw, info )
*
*        Compute QR factorization of work(iwa:iwa+m-1,jwa:jwa+n+nrhs-1)
*
         mqw = numroc( m+icoffa, desca( nb_ ), mycol, iacol, npcol )
         ipw = iptau + min( mqw, nqw )
         lwork = descw( nb_ ) * ( mpw + nqw + descw( nb_ ) )
         CALL psgeqrf( m, n+nrhs, work( ipwa ), iwa, jwa, descw,
     $                work( iptau ), work( ipw ), lwork, info )
*
*        Compute largest entry in upper triangle of
*        work(iwa+n:iwa+m-1,jwa+n:jwa+n+nrhs-1)
*
         err = zero
         IF( n.LT.m ) THEN
            DO 20 j = jwx, jwa+n+nrhs-1
               CALL psamax( min(m-n,j-jwx+1), amax, idum, work( ipwa ),
     $                      iwa+n, j, descw, 1 )
               err = max( err, abs( amax ) )
   20       CONTINUE
         END IF
         CALL sgamx2d( ictxt, 'All', ' ', 1, 1, err, 1, idum1, idum2,
     $                 -1, -1, 0 )
*
      ELSE
*
*        Copy sub( X )' into rows iwa+m:iwa+m+nrhs-1 of work
*
         DO 30 j = 1, nrhs
            CALL pscopy( n, x, ix, jx+j-1, descx, 1, work( ipwa ),
     $                   iwx+j-1, jwx, descw, descw( m_ ) )
   30    CONTINUE
*
         xnrm = pslange( 'M', nrhs, n, work( ipwa ), iwx, jwx, descw,
     $                   rwork )
         IF( xnrm.NE.zero )
     $      CALL pslascl( 'G', xnrm, one, nrhs, n, work( ipwa ), iwx,
     $                    jwx, descw, info )
*
*        Compute LQ factorization of work(iwa:iwa+m+nrhs-1,jwa:jwa+n-1)
*
         npw = numroc( n+iroffa, desca( mb_ ), myrow, iarow, nprow )
         ipw = iptau + min( mpw, npw )
         lwork = descw( mb_ ) * ( mpw + nqw + descw( mb_ ) )
         CALL psgelqf( m+nrhs, n, work( ipwa ), iwa, jwa, descw,
     $                 work( iptau ), work( ipw ), lwork, info )
*
*        Compute largest entry in lower triangle in
*        work(iwa+m:iwa+m+nrhs-1,jwa+m:jwa+n-1)
*
         err = zero
         DO 40 j = jwa+m, min( jwa+n-1, jwa+m+nrhs-1 )
            CALL psamax( jwa+m+nrhs-j, amax, idum, work( ipwa ),
     $                   iwx+j-jwa-m, j, descw, 1 )
            err = max( err, abs( amax ) )
   40    CONTINUE
         CALL sgamx2d( ictxt, 'All', ' ', 1, 1, err, 1, idum1, idum2,
     $                 -1, -1, 0 )
*
      END IF
*
      psqrt14 = err / ( real( max( m, n, nrhs ) ) *
     $          pslamch( ictxt, 'Epsilon' ) )
*
      RETURN
*
*     End of PSQRT14
*

Here is the call graph for this function:

Here is the caller graph for this function: