SUBROUTINE PCHEEVX( JOBZ, RANGE, UPLO, N, A, IA, JA, DESCA, VL,
$ VU, IL, IU, ABSTOL, M, NZ, W, ORFAC, Z, IZ,
$ JZ, DESCZ, WORK, LWORK, RWORK, LRWORK, IWORK,
$ LIWORK, IFAIL, ICLUSTR, GAP, INFO )
*
* -- ScaLAPACK routine (version 1.7) --
* University of Tennessee, Knoxville, Oak Ridge National Laboratory,
* and University of California, Berkeley.
* May 25, 2001
*
* .. Scalar Arguments ..
CHARACTER JOBZ, RANGE, UPLO
INTEGER IA, IL, INFO, IU, IZ, JA, JZ, LIWORK, LRWORK,
$ LWORK, M, N, NZ
REAL ABSTOL, ORFAC, VL, VU
* ..
* .. Array Arguments ..
INTEGER DESCA( * ), DESCZ( * ), ICLUSTR( * ),
$ IFAIL( * ), IWORK( * )
REAL GAP( * ), RWORK( * ), W( * )
COMPLEX A( * ), WORK( * ), Z( * )
* ..
*
* Purpose
* =======
*
* PCHEEVX computes selected eigenvalues and, optionally, eigenvectors
* of a complex hermitian matrix A by calling the recommended sequence
* of ScaLAPACK routines. Eigenvalues/vectors can be selected by
* specifying a range of values or a range of indices for the desired
* eigenvalues.
*
* 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
*
* PCHEEVX assumes IEEE 754 standard compliant arithmetic. To port
* to a system which does not have IEEE 754 arithmetic, modify
* the appropriate SLmake.inc file to include the compiler switch
* -DNO_IEEE. This switch only affects the compilation of pslaiect.c.
*
* Arguments
* =========
*
* NP = the number of rows local to a given process.
* NQ = the number of columns local to a given process.
*
* JOBZ (global input) CHARACTER*1
* Specifies whether or not to compute the eigenvectors:
* = 'N': Compute eigenvalues only.
* = 'V': Compute eigenvalues and eigenvectors.
*
* RANGE (global input) CHARACTER*1
* = 'A': all eigenvalues will be found.
* = 'V': all eigenvalues in the interval [VL,VU] will be found.
* = 'I': the IL-th through IU-th eigenvalues will be found.
*
* UPLO (global input) CHARACTER*1
* Specifies whether the upper or lower triangular part of the
* Hermitian matrix A is stored:
* = 'U': Upper triangular
* = 'L': Lower triangular
*
* N (global input) INTEGER
* The number of rows and columns of the matrix A. N >= 0.
*
* A (local input/workspace) block cyclic COMPLEX array,
* global dimension (N, N),
* local dimension ( LLD_A, LOCc(JA+N-1) )
*
* On entry, the Hermitian matrix A. If UPLO = 'U', only the
* upper triangular part of A is used to define the elements of
* the Hermitian matrix. If UPLO = 'L', only the lower
* triangular part of A is used to define the elements of the
* Hermitian matrix.
*
* On exit, the lower triangle (if UPLO='L') or the upper
* triangle (if UPLO='U') of A, including the diagonal, is
* destroyed.
*
* IA (global input) INTEGER
* A's global row index, which points to the beginning of the
* submatrix which is to be operated on.
*
* JA (global input) INTEGER
* A's global column index, which points to the beginning of
* the submatrix which is to be operated on.
*
* DESCA (global and local input) INTEGER array of dimension DLEN_.
* The array descriptor for the distributed matrix A.
* If DESCA( CTXT_ ) is incorrect, PCHEEVX cannot guarantee
* correct error reporting.
*
* VL (global input) REAL
* If RANGE='V', the lower bound of the interval to be searched
* for eigenvalues. Not referenced if RANGE = 'A' or 'I'.
*
* VU (global input) REAL
* If RANGE='V', the upper bound of the interval to be searched
* for eigenvalues. Not referenced if RANGE = 'A' or 'I'.
*
* IL (global input) INTEGER
* If RANGE='I', the index (from smallest to largest) of the
* smallest eigenvalue to be returned. IL >= 1.
* Not referenced if RANGE = 'A' or 'V'.
*
* IU (global input) INTEGER
* If RANGE='I', the index (from smallest to largest) of the
* largest eigenvalue to be returned. min(IL,N) <= IU <= N.
* Not referenced if RANGE = 'A' or 'V'.
*
* ABSTOL (global input) REAL
* If JOBZ='V', setting ABSTOL to PSLAMCH( CONTEXT, 'U') yields
* the most orthogonal eigenvectors.
*
* The absolute error tolerance for the eigenvalues.
* An approximate eigenvalue is accepted as converged
* when it is determined to lie in an interval [a,b]
* of width less than or equal to
*
* ABSTOL + EPS * max( |a|,|b| ) ,
*
* where EPS is the machine precision. If ABSTOL is less than
* or equal to zero, then EPS*norm(T) will be used in its place,
* where norm(T) is the 1-norm of the tridiagonal matrix
* obtained by reducing A to tridiagonal form.
*
* Eigenvalues will be computed most accurately when ABSTOL is
* set to twice the underflow threshold 2*PSLAMCH('S') not zero.
* If this routine returns with ((MOD(INFO,2).NE.0) .OR.
* (MOD(INFO/8,2).NE.0)), indicating that some eigenvalues or
* eigenvectors did not converge, try setting ABSTOL to
* 2*PSLAMCH('S').
*
* See "Computing Small Singular Values of Bidiagonal Matrices
* with Guaranteed High Relative Accuracy," by Demmel and
* Kahan, LAPACK Working Note #3.
*
* See "On the correctness of Parallel Bisection in Floating
* Point" by Demmel, Dhillon and Ren, LAPACK Working Note #70
*
* M (global output) INTEGER
* Total number of eigenvalues found. 0 <= M <= N.
*
* NZ (global output) INTEGER
* Total number of eigenvectors computed. 0 <= NZ <= M.
* The number of columns of Z that are filled.
* If JOBZ .NE. 'V', NZ is not referenced.
* If JOBZ .EQ. 'V', NZ = M unless the user supplies
* insufficient space and PCHEEVX is not able to detect this
* before beginning computation. To get all the eigenvectors
* requested, the user must supply both sufficient
* space to hold the eigenvectors in Z (M .LE. DESCZ(N_))
* and sufficient workspace to compute them. (See LWORK below.)
* PCHEEVX is always able to detect insufficient space without
* computation unless RANGE .EQ. 'V'.
*
* W (global output) REAL array, dimension (N)
* On normal exit, the first M entries contain the selected
* eigenvalues in ascending order.
*
* ORFAC (global input) REAL
* Specifies which eigenvectors should be reorthogonalized.
* Eigenvectors that correspond to eigenvalues which are within
* tol=ORFAC*norm(A) of each other are to be reorthogonalized.
* However, if the workspace is insufficient (see LWORK),
* tol may be decreased until all eigenvectors to be
* reorthogonalized can be stored in one process.
* No reorthogonalization will be done if ORFAC equals zero.
* A default value of 10^-3 is used if ORFAC is negative.
* ORFAC should be identical on all processes.
*
* Z (local output) COMPLEX array,
* global dimension (N, N),
* local dimension ( LLD_Z, LOCc(JZ+N-1) )
* If JOBZ = 'V', then on normal exit the first M columns of Z
* contain the orthonormal eigenvectors of the matrix
* corresponding to the selected eigenvalues. If an eigenvector
* fails to converge, then that column of Z contains the latest
* approximation to the eigenvector, and the index of the
* eigenvector is returned in IFAIL.
* If JOBZ = 'N', then Z is not referenced.
*
* IZ (global input) INTEGER
* Z's global row index, which points to the beginning of the
* submatrix which is to be operated on.
*
* JZ (global input) INTEGER
* Z's global column index, which points to the beginning of
* the submatrix which is to be operated on.
*
* DESCZ (global and local input) INTEGER array of dimension DLEN_.
* The array descriptor for the distributed matrix Z.
* DESCZ( CTXT_ ) must equal DESCA( CTXT_ )
*
* WORK (local workspace/output) COMPLEX array,
* dimension (LWORK)
* WORK(1) returns workspace adequate workspace to allow
* optimal performance.
*
* LWORK (local input) INTEGER
* Size of WORK array. If only eigenvalues are requested:
* LWORK >= N + MAX( NB * ( NP0 + 1 ), 3 )
* If eigenvectors are requested:
* LWORK >= N + ( NP0 + MQ0 + NB ) * NB
* with NQ0 = NUMROC( NN, NB, 0, 0, NPCOL ).
*
* For optimal performance, greater workspace is needed, i.e.
* LWORK >= MAX( LWORK, NHETRD_LWORK )
* Where LWORK is as defined above, and
* NHETRD_LWORK = N + 2*( ANB+1 )*( 4*NPS+2 ) +
* ( NPS + 1 ) * NPS
*
* ICTXT = DESCA( CTXT_ )
* ANB = PJLAENV( ICTXT, 3, 'PCHETTRD', 'L', 0, 0, 0, 0 )
* SQNPC = SQRT( DBLE( NPROW * NPCOL ) )
* NPS = MAX( NUMROC( N, 1, 0, 0, SQNPC ), 2*ANB )
*
* NUMROC is a ScaLAPACK tool functions;
* PJLAENV is a ScaLAPACK envionmental inquiry function
* MYROW, MYCOL, NPROW and NPCOL can be determined by calling
* the subroutine BLACS_GRIDINFO.
*
* If LWORK = -1, then LWORK is global input and a workspace
* query is assumed; the routine only calculates the
* optimal size for all work arrays. Each of these
* values is returned in the first entry of the corresponding
* work array, and no error message is issued by PXERBLA.
*
* RWORK (local workspace/output) REAL array,
* dimension max(3,LRWORK)
* On return, WORK(1) contains the optimal amount of
* workspace required for efficient execution.
* if JOBZ='N' RWORK(1) = optimal amount of workspace
* required to compute eigenvalues efficiently
* if JOBZ='V' RWORK(1) = optimal amount of workspace
* required to compute eigenvalues and eigenvectors
* efficiently with no guarantee on orthogonality.
* If RANGE='V', it is assumed that all eigenvectors
* may be required.
*
* LRWORK (local input) INTEGER
* Size of RWORK
* See below for definitions of variables used to define LRWORK.
* If no eigenvectors are requested (JOBZ = 'N') then
* LRWORK >= 5 * NN + 4 * N
* If eigenvectors are requested (JOBZ = 'V' ) then
* the amount of workspace required to guarantee that all
* eigenvectors are computed is:
* LRWORK >= 4*N + MAX( 5*NN, NP0 * MQ0 ) +
* ICEIL( NEIG, NPROW*NPCOL)*NN
*
* The computed eigenvectors may not be orthogonal if the
* minimal workspace is supplied and ORFAC is too small.
* If you want to guarantee orthogonality (at the cost
* of potentially poor performance) you should add
* the following to LRWORK:
* (CLUSTERSIZE-1)*N
* where CLUSTERSIZE is the number of eigenvalues in the
* largest cluster, where a cluster is defined as a set of
* close eigenvalues: { W(K),...,W(K+CLUSTERSIZE-1) |
* W(J+1) <= W(J) + ORFAC*2*norm(A) }
* Variable definitions:
* NEIG = number of eigenvectors requested
* NB = DESCA( MB_ ) = DESCA( NB_ ) =
* DESCZ( MB_ ) = DESCZ( NB_ )
* NN = MAX( N, NB, 2 )
* DESCA( RSRC_ ) = DESCA( NB_ ) = DESCZ( RSRC_ ) =
* DESCZ( CSRC_ ) = 0
* NP0 = NUMROC( NN, NB, 0, 0, NPROW )
* MQ0 = NUMROC( MAX( NEIG, NB, 2 ), NB, 0, 0, NPCOL )
* ICEIL( X, Y ) is a ScaLAPACK function returning
* ceiling(X/Y)
*
* When LRWORK is too small:
* If LRWORK is too small to guarantee orthogonality,
* PCHEEVX attempts to maintain orthogonality in
* the clusters with the smallest
* spacing between the eigenvalues.
* If LRWORK is too small to compute all the eigenvectors
* requested, no computation is performed and INFO=-25
* is returned. Note that when RANGE='V', PCHEEVX does
* not know how many eigenvectors are requested until
* the eigenvalues are computed. Therefore, when RANGE='V'
* and as long as LRWORK is large enough to allow PCHEEVX to
* compute the eigenvalues, PCHEEVX will compute the
* eigenvalues and as many eigenvectors as it can.
*
* Relationship between workspace, orthogonality & performance:
* If CLUSTERSIZE >= N/SQRT(NPROW*NPCOL), then providing
* enough space to compute all the eigenvectors
* orthogonally will cause serious degradation in
* performance. In the limit (i.e. CLUSTERSIZE = N-1)
* PCSTEIN will perform no better than CSTEIN on 1
* processor.
* For CLUSTERSIZE = N/SQRT(NPROW*NPCOL) reorthogonalizing
* all eigenvectors will increase the total execution time
* by a factor of 2 or more.
* For CLUSTERSIZE > N/SQRT(NPROW*NPCOL) execution time will
* grow as the square of the cluster size, all other factors
* remaining equal and assuming enough workspace. Less
* workspace means less reorthogonalization but faster
* execution.
*
* If LRWORK = -1, then LRWORK is global input and a workspace
* query is assumed; the routine only calculates the size
* required for optimal performance for all work arrays. Each of
* these values is returned in the first entry of the
* corresponding work arrays, and no error message is issued by
* PXERBLA.
*
* IWORK (local workspace) INTEGER array
* On return, IWORK(1) contains the amount of integer workspace
* required.
*
* LIWORK (local input) INTEGER
* size of IWORK
* LIWORK >= 6 * NNP
* Where:
* NNP = MAX( N, NPROW*NPCOL + 1, 4 )
* If LIWORK = -1, then LIWORK is global input and a workspace
* query is assumed; the routine only calculates the minimum
* and optimal size for all work arrays. Each of these
* values is returned in the first entry of the corresponding
* work array, and no error message is issued by PXERBLA.
*
* IFAIL (global output) INTEGER array, dimension (N)
* If JOBZ = 'V', then on normal exit, the first M elements of
* IFAIL are zero. If (MOD(INFO,2).NE.0) on exit, then
* IFAIL contains the
* indices of the eigenvectors that failed to converge.
* If JOBZ = 'N', then IFAIL is not referenced.
*
* ICLUSTR (global output) integer array, dimension (2*NPROW*NPCOL)
* This array contains indices of eigenvectors corresponding to
* a cluster of eigenvalues that could not be reorthogonalized
* due to insufficient workspace (see LWORK, ORFAC and INFO).
* Eigenvectors corresponding to clusters of eigenvalues indexed
* ICLUSTR(2*I-1) to ICLUSTR(2*I), could not be
* reorthogonalized due to lack of workspace. Hence the
* eigenvectors corresponding to these clusters may not be
* orthogonal. ICLUSTR() is a zero terminated array.
* (ICLUSTR(2*K).NE.0 .AND. ICLUSTR(2*K+1).EQ.0) if and only if
* K is the number of clusters
* ICLUSTR is not referenced if JOBZ = 'N'
*
* GAP (global output) REAL array,
* dimension (NPROW*NPCOL)
* This array contains the gap between eigenvalues whose
* eigenvectors could not be reorthogonalized. The output
* values in this array correspond to the clusters indicated
* by the array ICLUSTR. As a result, the dot product between
* eigenvectors correspoding to the I^th cluster may be as high
* as ( C * n ) / GAP(I) where C is a small constant.
*
* INFO (global output) INTEGER
* = 0: successful exit
* < 0: If the i-th argument is an array and the j-entry had
* an illegal value, then INFO = -(i*100+j), if the i-th
* argument is a scalar and had an illegal value, then
* INFO = -i.
* > 0: if (MOD(INFO,2).NE.0), then one or more eigenvectors
* failed to converge. Their indices are stored
* in IFAIL. Ensure ABSTOL=2.0*PSLAMCH( 'U' )
* Send e-mail to scalapack@cs.utk.edu
* if (MOD(INFO/2,2).NE.0),then eigenvectors corresponding
* to one or more clusters of eigenvalues could not be
* reorthogonalized because of insufficient workspace.
* The indices of the clusters are stored in the array
* ICLUSTR.
* if (MOD(INFO/4,2).NE.0), then space limit prevented
* PCHEEVX from computing all of the eigenvectors
* between VL and VU. The number of eigenvectors
* computed is returned in NZ.
* if (MOD(INFO/8,2).NE.0), then PCSTEBZ failed to compute
* eigenvalues. Ensure ABSTOL=2.0*PSLAMCH( 'U' )
* Send e-mail to scalapack@cs.utk.edu
*
* Alignment requirements
* ======================
*
* The distributed submatrices A(IA:*, JA:*) and C(IC:IC+M-1,JC:JC+N-1)
* must verify some alignment properties, namely the following
* expressions should be true:
*
* ( MB_A.EQ.NB_A.EQ.MB_Z .AND. IROFFA.EQ.IROFFZ .AND. IROFFA.EQ.0 .AND.
* IAROW.EQ.IZROW )
* where
* IROFFA = MOD( IA-1, MB_A ) and ICOFFA = MOD( JA-1, NB_A ).
*
* =====================================================================
*
* Differences between PCHEEVX and CHEEVX
* ======================================
*
* A, LDA -> A, IA, JA, DESCA
* Z, LDZ -> Z, IZ, JZ, DESCZ
* WORKSPACE needs are larger for PCHEEVX.
* LIWORK parameter added
*
* ORFAC, ICLUSTER() and GAP() parameters added
* meaning of INFO is changed
*
* Functional differences:
* PCHEEVX does not promise orthogonality for eigenvectors associated
* with tighly clustered eigenvalues.
* PCHEEVX does not reorthogonalize eigenvectors
* that are on different processes. The extent of reorthogonalization
* is controlled by the input parameter LWORK.
*
* Version 1.4 limitations:
* DESCA(MB_) = DESCA(NB_)
* DESCA(M_) = DESCZ(M_)
* DESCA(N_) = DESCZ(N_)
* DESCA(MB_) = DESCZ(MB_)
* DESCA(NB_) = DESCZ(NB_)
* DESCA(RSRC_) = DESCZ(RSRC_)
*
* .. Parameters ..
INTEGER BLOCK_CYCLIC_2D, DLEN_, DTYPE_, CTXT_, M_, N_,
$ MB_, NB_, RSRC_, CSRC_, LLD_
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 ZERO, ONE, TEN, FIVE
PARAMETER ( ZERO = 0.0E+0, ONE = 1.0E+0, TEN = 10.0E+0,
$ FIVE = 5.0E+0 )
INTEGER IERREIN, IERRCLS, IERRSPC, IERREBZ
PARAMETER ( IERREIN = 1, IERRCLS = 2, IERRSPC = 4,
$ IERREBZ = 8 )
* ..
* .. Local Scalars ..
LOGICAL ALLEIG, INDEIG, LOWER, LQUERY, QUICKRETURN,
$ VALEIG, WANTZ
CHARACTER ORDER
INTEGER ANB, CSRC_A, I, IAROW, ICOFFA, ICTXT, IINFO,
$ INDD, INDD2, INDE, INDE2, INDIBL, INDISP,
$ INDRWORK, INDTAU, INDWORK, IROFFA, IROFFZ,
$ ISCALE, ISIZESTEBZ, ISIZESTEIN, IZROW,
$ LALLWORK, LIWMIN, LLRWORK, LLWORK, LRWMIN,
$ LRWOPT, LWMIN, LWOPT, MAXEIGS, MB_A, MQ0,
$ MYCOL, MYROW, NB, NB_A, NEIG, NHETRD_LWOPT, NN,
$ NNP, NP0, NPCOL, NPROCS, NPROW, NPS, NQ0,
$ NSPLIT, NZZ, OFFSET, RSRC_A, RSRC_Z, SIZEHEEVX,
$ SIZESTEIN, SQNPC
REAL ABSTLL, ANRM, BIGNUM, EPS, RMAX, RMIN, SAFMIN,
$ SIGMA, SMLNUM, VLL, VUU
* ..
* .. Local Arrays ..
INTEGER IDUM1( 4 ), IDUM2( 4 )
* ..
* .. External Functions ..
LOGICAL LSAME
INTEGER ICEIL, INDXG2P, NUMROC, PJLAENV
REAL PCLANHE, PSLAMCH
EXTERNAL LSAME, ICEIL, INDXG2P, NUMROC, PJLAENV,
$ PCLANHE, PSLAMCH
* ..
* .. External Subroutines ..
EXTERNAL BLACS_GRIDINFO, CHK1MAT, IGAMN2D, PCELGET,
$ PCHENTRD, PCHK1MAT, PCHK2MAT, PCLASCL, PCSTEIN,
$ PCUNMTR, PSLARED1D, PSSTEBZ, PXERBLA, SGEBR2D,
$ SGEBS2D, SLASRT, SSCAL
* ..
* .. Intrinsic Functions ..
INTRINSIC ABS, CMPLX, DBLE, ICHAR, INT, MAX, MIN, MOD,
$ REAL, SQRT
* ..
* .. Executable Statements ..
* This is just to keep ftnchek and toolpack/1 happy
IF( BLOCK_CYCLIC_2D*CSRC_*CTXT_*DLEN_*DTYPE_*LLD_*MB_*M_*NB_*N_*
$ RSRC_.LT.0 )RETURN
*
QUICKRETURN = ( N.EQ.0 )
*
* Test the input arguments.
*
ICTXT = DESCA( CTXT_ )
CALL BLACS_GRIDINFO( ICTXT, NPROW, NPCOL, MYROW, MYCOL )
INFO = 0
*
WANTZ = LSAME( JOBZ, 'V' )
IF( NPROW.EQ.-1 ) THEN
INFO = -( 800+CTXT_ )
ELSE IF( WANTZ ) THEN
IF( ICTXT.NE.DESCZ( CTXT_ ) ) THEN
INFO = -( 2100+CTXT_ )
END IF
END IF
IF( INFO.EQ.0 ) THEN
CALL CHK1MAT( N, 4, N, 4, IA, JA, DESCA, 8, INFO )
IF( WANTZ )
$ CALL CHK1MAT( N, 4, N, 4, IZ, JZ, DESCZ, 21, INFO )
*
IF( INFO.EQ.0 ) THEN
*
* Get machine constants.
*
SAFMIN = PSLAMCH( ICTXT, 'Safe minimum' )
EPS = PSLAMCH( ICTXT, 'Precision' )
SMLNUM = SAFMIN / EPS
BIGNUM = ONE / SMLNUM
RMIN = SQRT( SMLNUM )
RMAX = MIN( SQRT( BIGNUM ), ONE / SQRT( SQRT( SAFMIN ) ) )
*
NPROCS = NPROW*NPCOL
LOWER = LSAME( UPLO, 'L' )
ALLEIG = LSAME( RANGE, 'A' )
VALEIG = LSAME( RANGE, 'V' )
INDEIG = LSAME( RANGE, 'I' )
*
* Set up pointers into the WORK array
*
INDTAU = 1
INDWORK = INDTAU + N
LLWORK = LWORK - INDWORK + 1
*
* Set up pointers into the RWORK array
*
INDE = 1
INDD = INDE + N
INDD2 = INDD + N
INDE2 = INDD2 + N
INDRWORK = INDE2 + N
LLRWORK = LRWORK - INDRWORK + 1
*
* Set up pointers into the IWORK array
*
ISIZESTEIN = 3*N + NPROCS + 1
ISIZESTEBZ = MAX( 4*N, 14, NPROCS )
INDIBL = ( MAX( ISIZESTEIN, ISIZESTEBZ ) ) + 1
INDISP = INDIBL + N
*
* Compute the total amount of space needed
*
LQUERY = .FALSE.
IF( LWORK.EQ.-1 .OR. LIWORK.EQ.-1 .OR. LRWORK.EQ.-1 )
$ LQUERY = .TRUE.
*
NNP = MAX( N, NPROCS+1, 4 )
LIWMIN = 6*NNP
*
NPROCS = NPROW*NPCOL
NB_A = DESCA( NB_ )
MB_A = DESCA( MB_ )
NB = NB_A
NN = MAX( N, NB, 2 )
*
RSRC_A = DESCA( RSRC_ )
CSRC_A = DESCA( CSRC_ )
IROFFA = MOD( IA-1, MB_A )
ICOFFA = MOD( JA-1, NB_A )
IAROW = INDXG2P( 1, NB_A, MYROW, RSRC_A, NPROW )
NP0 = NUMROC( N+IROFFA, NB, 0, 0, NPROW )
MQ0 = NUMROC( N+ICOFFA, NB, 0, 0, NPCOL )
IF( WANTZ ) THEN
RSRC_Z = DESCZ( RSRC_ )
IROFFZ = MOD( IZ-1, MB_A )
IZROW = INDXG2P( 1, NB_A, MYROW, RSRC_Z, NPROW )
ELSE
IROFFZ = 0
IZROW = 0
END IF
*
IF( ( .NOT.WANTZ ) .OR. ( VALEIG .AND. ( .NOT.LQUERY ) ) )
$ THEN
LWMIN = N + MAX( NB*( NP0+1 ), 3 )
LWOPT = LWMIN
LRWMIN = 5*NN + 4*N
IF( WANTZ ) THEN
MQ0 = NUMROC( MAX( N, NB, 2 ), NB, 0, 0, NPCOL )
LRWOPT = 4*N + MAX( 5*NN, NP0*MQ0 ) +
$ ICEIL( N, NPROW*NPCOL )*NN
ELSE
LRWOPT = LRWMIN
END IF
NEIG = 0
ELSE
IF( ALLEIG .OR. VALEIG ) THEN
NEIG = N
ELSE IF( INDEIG ) THEN
NEIG = IU - IL + 1
END IF
MQ0 = NUMROC( MAX( NEIG, NB, 2 ), NB, 0, 0, NPCOL )
NQ0 = NUMROC( NN, NB, 0, 0, NPCOL )
LWMIN = N + ( NP0+NQ0+NB )*NB
LRWMIN = 4*N + MAX( 5*NN, NP0*MQ0 ) +
$ ICEIL( NEIG, NPROW*NPCOL )*NN
LRWOPT = LRWMIN
LWOPT = LWMIN
*
END IF
*
* Compute how much workspace is needed to use the
* new TRD code
*
ANB = PJLAENV( ICTXT, 3, 'PCHETTRD', 'L', 0, 0, 0, 0 )
SQNPC = INT( SQRT( DBLE( NPROW*NPCOL ) ) )
NPS = MAX( NUMROC( N, 1, 0, 0, SQNPC ), 2*ANB )
NHETRD_LWOPT = 2*( ANB+1 )*( 4*NPS+2 ) + ( NPS+2 )*NPS
LWOPT = MAX( LWOPT, N+NHETRD_LWOPT )
*
END IF
IF( INFO.EQ.0 ) THEN
IF( MYROW.EQ.0 .AND. MYCOL.EQ.0 ) THEN
RWORK( 1 ) = ABSTOL
IF( VALEIG ) THEN
RWORK( 2 ) = VL
RWORK( 3 ) = VU
ELSE
RWORK( 2 ) = ZERO
RWORK( 3 ) = ZERO
END IF
CALL SGEBS2D( ICTXT, 'ALL', ' ', 3, 1, RWORK, 3 )
ELSE
CALL SGEBR2D( ICTXT, 'ALL', ' ', 3, 1, RWORK, 3, 0, 0 )
END IF
IF( .NOT.( WANTZ .OR. LSAME( JOBZ, 'N' ) ) ) THEN
INFO = -1
ELSE IF( .NOT.( ALLEIG .OR. VALEIG .OR. INDEIG ) ) THEN
INFO = -2
ELSE IF( .NOT.( LOWER .OR. LSAME( UPLO, 'U' ) ) ) THEN
INFO = -3
ELSE IF( VALEIG .AND. N.GT.0 .AND. VU.LE.VL ) THEN
INFO = -10
ELSE IF( INDEIG .AND. ( IL.LT.1 .OR. IL.GT.MAX( 1, N ) ) )
$ THEN
INFO = -11
ELSE IF( INDEIG .AND. ( IU.LT.MIN( N, IL ) .OR. IU.GT.N ) )
$ THEN
INFO = -12
ELSE IF( LWORK.LT.LWMIN .AND. LWORK.NE.-1 ) THEN
INFO = -23
ELSE IF( LRWORK.LT.LRWMIN .AND. LRWORK.NE.-1 ) THEN
INFO = -25
ELSE IF( LIWORK.LT.LIWMIN .AND. LIWORK.NE.-1 ) THEN
INFO = -27
ELSE IF( VALEIG .AND. ( ABS( RWORK( 2 )-VL ).GT.FIVE*EPS*
$ ABS( VL ) ) ) THEN
INFO = -9
ELSE IF( VALEIG .AND. ( ABS( RWORK( 3 )-VU ).GT.FIVE*EPS*
$ ABS( VU ) ) ) THEN
INFO = -10
ELSE IF( ABS( RWORK( 1 )-ABSTOL ).GT.FIVE*EPS*
$ ABS( ABSTOL ) ) THEN
INFO = -13
ELSE IF( IROFFA.NE.0 ) THEN
INFO = -6
ELSE IF( DESCA( MB_ ).NE.DESCA( NB_ ) ) THEN
INFO = -( 800+NB_ )
END IF
IF( WANTZ ) THEN
IF( IROFFA.NE.IROFFZ ) THEN
INFO = -19
ELSE IF( IAROW.NE.IZROW ) THEN
INFO = -19
ELSE IF( DESCA( M_ ).NE.DESCZ( M_ ) ) THEN
INFO = -( 2100+M_ )
ELSE IF( DESCA( N_ ).NE.DESCZ( N_ ) ) THEN
INFO = -( 2100+N_ )
ELSE IF( DESCA( MB_ ).NE.DESCZ( MB_ ) ) THEN
INFO = -( 2100+MB_ )
ELSE IF( DESCA( NB_ ).NE.DESCZ( NB_ ) ) THEN
INFO = -( 2100+NB_ )
ELSE IF( DESCA( RSRC_ ).NE.DESCZ( RSRC_ ) ) THEN
INFO = -( 2100+RSRC_ )
ELSE IF( DESCA( CSRC_ ).NE.DESCZ( CSRC_ ) ) THEN
INFO = -( 2100+CSRC_ )
ELSE IF( ICTXT.NE.DESCZ( CTXT_ ) ) THEN
INFO = -( 2100+CTXT_ )
END IF
END IF
END IF
IF( WANTZ ) THEN
IDUM1( 1 ) = ICHAR( 'V' )
ELSE
IDUM1( 1 ) = ICHAR( 'N' )
END IF
IDUM2( 1 ) = 1
IF( LOWER ) THEN
IDUM1( 2 ) = ICHAR( 'L' )
ELSE
IDUM1( 2 ) = ICHAR( 'U' )
END IF
IDUM2( 2 ) = 2
IF( ALLEIG ) THEN
IDUM1( 3 ) = ICHAR( 'A' )
ELSE IF( INDEIG ) THEN
IDUM1( 3 ) = ICHAR( 'I' )
ELSE
IDUM1( 3 ) = ICHAR( 'V' )
END IF
IDUM2( 3 ) = 3
IF( LQUERY ) THEN
IDUM1( 4 ) = -1
ELSE
IDUM1( 4 ) = 1
END IF
IDUM2( 4 ) = 4
IF( WANTZ ) THEN
CALL PCHK2MAT( N, 4, N, 4, IA, JA, DESCA, 8, N, 4, N, 4, IZ,
$ JZ, DESCZ, 21, 4, IDUM1, IDUM2, INFO )
ELSE
CALL PCHK1MAT( N, 4, N, 4, IA, JA, DESCA, 8, 4, IDUM1,
$ IDUM2, INFO )
END IF
WORK( 1 ) = CMPLX( LWOPT )
RWORK( 1 ) = REAL( LRWOPT )
IWORK( 1 ) = LIWMIN
END IF
*
IF( INFO.NE.0 ) THEN
CALL PXERBLA( ICTXT, 'PCHEEVX', -INFO )
RETURN
ELSE IF( LQUERY ) THEN
RETURN
END IF
*
* Quick return if possible
*
IF( QUICKRETURN ) THEN
IF( WANTZ ) THEN
NZ = 0
ICLUSTR( 1 ) = 0
END IF
M = 0
WORK( 1 ) = CMPLX( LWOPT )
RWORK( 1 ) = REAL( LRWMIN )
IWORK( 1 ) = LIWMIN
RETURN
END IF
*
* Scale matrix to allowable range, if necessary.
*
ABSTLL = ABSTOL
ISCALE = 0
IF( VALEIG ) THEN
VLL = VL
VUU = VU
ELSE
VLL = ZERO
VUU = ZERO
END IF
*
ANRM = PCLANHE( 'M', UPLO, N, A, IA, JA, DESCA,
$ RWORK( INDRWORK ) )
*
IF( ANRM.GT.ZERO .AND. ANRM.LT.RMIN ) THEN
ISCALE = 1
SIGMA = RMIN / ANRM
ANRM = ANRM*SIGMA
ELSE IF( ANRM.GT.RMAX ) THEN
ISCALE = 1
SIGMA = RMAX / ANRM
ANRM = ANRM*SIGMA
END IF
*
IF( ISCALE.EQ.1 ) THEN
CALL PCLASCL( UPLO, ONE, SIGMA, N, N, A, IA, JA, DESCA, IINFO )
IF( ABSTOL.GT.0 )
$ ABSTLL = ABSTOL*SIGMA
IF( VALEIG ) THEN
VLL = VL*SIGMA
VUU = VU*SIGMA
IF( VUU.EQ.VLL ) THEN
VUU = VUU + 2*MAX( ABS( VUU )*EPS, SAFMIN )
END IF
END IF
END IF
*
* Call PCHENTRD to reduce Hermitian matrix to tridiagonal form.
*
LALLWORK = LLRWORK
*
CALL PCHENTRD( UPLO, N, A, IA, JA, DESCA, RWORK( INDD ),
$ RWORK( INDE ), WORK( INDTAU ), WORK( INDWORK ),
$ LLWORK, RWORK( INDRWORK ), LLRWORK, IINFO )
*
*
* Copy the values of D, E to all processes
*
* Here PxLARED1D is used to redistribute the tridiagonal matrix.
* PxLARED1D, however, doesn't yet work with arbritary matrix
* distributions so we have PxELGET as a backup.
*
OFFSET = 0
IF( IA.EQ.1 .AND. JA.EQ.1 .AND. RSRC_A.EQ.0 .AND. CSRC_A.EQ.0 )
$ THEN
CALL PSLARED1D( N, IA, JA, DESCA, RWORK( INDD ),
$ RWORK( INDD2 ), RWORK( INDRWORK ), LLRWORK )
*
CALL PSLARED1D( N, IA, JA, DESCA, RWORK( INDE ),
$ RWORK( INDE2 ), RWORK( INDRWORK ), LLRWORK )
IF( .NOT.LOWER )
$ OFFSET = 1
ELSE
DO 10 I = 1, N
CALL PCELGET( 'A', ' ', WORK( INDD2+I-1 ), A, I+IA-1,
$ I+JA-1, DESCA )
RWORK( INDD2+I-1 ) = REAL( WORK( INDD2+I-1 ) )
10 CONTINUE
IF( LSAME( UPLO, 'U' ) ) THEN
DO 20 I = 1, N - 1
CALL PCELGET( 'A', ' ', WORK( INDE2+I-1 ), A, I+IA-1,
$ I+JA, DESCA )
RWORK( INDE2+I-1 ) = REAL( WORK( INDE2+I-1 ) )
20 CONTINUE
ELSE
DO 30 I = 1, N - 1
CALL PCELGET( 'A', ' ', WORK( INDE2+I-1 ), A, I+IA,
$ I+JA-1, DESCA )
RWORK( INDE2+I-1 ) = REAL( WORK( INDE2+I-1 ) )
30 CONTINUE
END IF
END IF
*
* Call PSSTEBZ and, if eigenvectors are desired, PCSTEIN.
*
IF( WANTZ ) THEN
ORDER = 'B'
ELSE
ORDER = 'E'
END IF
*
CALL PSSTEBZ( ICTXT, RANGE, ORDER, N, VLL, VUU, IL, IU, ABSTLL,
$ RWORK( INDD2 ), RWORK( INDE2+OFFSET ), M, NSPLIT, W,
$ IWORK( INDIBL ), IWORK( INDISP ), RWORK( INDRWORK ),
$ LLRWORK, IWORK( 1 ), ISIZESTEBZ, IINFO )
*
*
* IF PSSTEBZ fails, the error propogates to INFO, but
* we do not propogate the eigenvalue(s) which failed because:
* 1) This should never happen if the user specifies
* ABSTOL = 2 * PSLAMCH( 'U' )
* 2) PSSTEIN will confirm/deny whether the eigenvalues are
* close enough.
*
IF( IINFO.NE.0 ) THEN
INFO = INFO + IERREBZ
DO 40 I = 1, M
IWORK( INDIBL+I-1 ) = ABS( IWORK( INDIBL+I-1 ) )
40 CONTINUE
END IF
IF( WANTZ ) THEN
*
IF( VALEIG ) THEN
*
* Compute the maximum number of eigenvalues that we can
* compute in the
* workspace that we have, and that we can store in Z.
*
* Loop through the possibilities looking for the largest
* NZ that we can feed to PCSTEIN and PCUNMTR
*
* Since all processes must end up with the same value
* of NZ, we first compute the minimum of LALLWORK
*
CALL IGAMN2D( ICTXT, 'A', ' ', 1, 1, LALLWORK, 1, 1, 1, -1,
$ -1, -1 )
*
MAXEIGS = DESCZ( N_ )
*
DO 50 NZ = MIN( MAXEIGS, M ), 0, -1
MQ0 = NUMROC( NZ, NB, 0, 0, NPCOL )
SIZESTEIN = ICEIL( NZ, NPROCS )*N + MAX( 5*N, NP0*MQ0 )
SIZEHEEVX = SIZESTEIN
IF( SIZEHEEVX.LE.LALLWORK )
$ GO TO 60
50 CONTINUE
60 CONTINUE
ELSE
NZ = M
END IF
NZ = MAX( NZ, 0 )
IF( NZ.NE.M ) THEN
INFO = INFO + IERRSPC
*
DO 70 I = 1, M
IFAIL( I ) = 0
70 CONTINUE
*
* The following code handles a rare special case
* - NZ .NE. M means that we don't have enough room to store
* all the vectors.
* - NSPLIT .GT. 1 means that the matrix split
* In this case, we cannot simply take the first NZ eigenvalues
* because PSSTEBZ sorts the eigenvalues by block when
* a split occurs. So, we have to make another call to
* PSSTEBZ with a new upper limit - VUU.
*
IF( NSPLIT.GT.1 ) THEN
CALL SLASRT( 'I', M, W, IINFO )
NZZ = 0
IF( NZ.GT.0 ) THEN
*
VUU = W( NZ ) - TEN*( EPS*ANRM+SAFMIN )
IF( VLL.GE.VUU ) THEN
NZZ = 0
ELSE
CALL PSSTEBZ( ICTXT, RANGE, ORDER, N, VLL, VUU, IL,
$ IU, ABSTLL, RWORK( INDD2 ),
$ RWORK( INDE2+OFFSET ), NZZ, NSPLIT,
$ W, IWORK( INDIBL ), IWORK( INDISP ),
$ RWORK( INDRWORK ), LLRWORK,
$ IWORK( 1 ), ISIZESTEBZ, IINFO )
END IF
*
IF( MOD( INFO / IERREBZ, 1 ).EQ.0 ) THEN
IF( NZZ.GT.NZ .OR. IINFO.NE.0 ) THEN
INFO = INFO + IERREBZ
END IF
END IF
END IF
NZ = MIN( NZ, NZZ )
*
END IF
END IF
CALL PCSTEIN( N, RWORK( INDD2 ), RWORK( INDE2+OFFSET ), NZ, W,
$ IWORK( INDIBL ), IWORK( INDISP ), ORFAC, Z, IZ,
$ JZ, DESCZ, RWORK( INDRWORK ), LALLWORK,
$ IWORK( 1 ), ISIZESTEIN, IFAIL, ICLUSTR, GAP,
$ IINFO )
*
IF( IINFO.GE.NZ+1 )
$ INFO = INFO + IERRCLS
IF( MOD( IINFO, NZ+1 ).NE.0 )
$ INFO = INFO + IERREIN
*
* Z = Q * Z
*
*
IF( NZ.GT.0 ) THEN
CALL PCUNMTR( 'L', UPLO, 'N', N, NZ, A, IA, JA, DESCA,
$ WORK( INDTAU ), Z, IZ, JZ, DESCZ,
$ WORK( INDWORK ), LLWORK, IINFO )
END IF
*
END IF
*
* If matrix was scaled, then rescale eigenvalues appropriately.
*
IF( ISCALE.EQ.1 ) THEN
CALL SSCAL( M, ONE / SIGMA, W, 1 )
END IF
*
WORK( 1 ) = CMPLX( LWOPT )
RWORK( 1 ) = REAL( LRWOPT )
IWORK( 1 ) = LIWMIN
*
RETURN
*
* End of PCHEEVX
*
END