|
|
| |
| # lines: |
309 | | # code: |
309 | | # comment: | 0 | |
# blank: | 0 |
| # Variables: | 42 |
| # Callers: | 1 |
| # Callings: | 2 |
| # Words: | 153 |
| # Keywords: | 87 |
|
|
|
|
|
..
.. Array Arguments ..
..
Purpose
=======
PSGEQL2 computes a QL factorization of a real distributed M-by-N
matrix sub( A ) = A(IA:IA+M-1,JA:JA+N-1) = Q * L.
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
=========
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.
A (local input/local output) REAL pointer into the
local memory to an array of dimension (LLD_A, LOCc(JA+N-1)).
On entry, the local pieces of the M-by-N distributed matrix
sub( A ) which is to be factored. On exit, if M >= N, the
lower triangle of the distributed submatrix
A( IA+M-N:IA+M-1, JA:JA+N-1 ) contains the N-by-N lower
triangular matrix L; if M <= N, the elements on and below
the (N-M)-th superdiagonal contain the M by N lower
trapezoidal matrix L; the remaining elements, with the
array TAU, represent the orthogonal matrix Q as a product of
elementary reflectors (see Further Details).
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.
TAU (local output) REAL, array, dimension LOCc(JA+N-1)
This array contains the scalar factors of the elementary
reflectors. TAU is tied to the distributed matrix A.
WORK (local workspace/local output) REAL array,
dimension (LWORK)
On exit, WORK(1) returns the minimal and optimal LWORK.
LWORK (local or global input) INTEGER
The dimension of the array WORK.
LWORK is local input and must be at least
LWORK >= Mp0 + MAX( 1, Nq0 ), where
IROFF = MOD( IA-1, MB_A ), ICOFF = 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+IROFF, MB_A, MYROW, IAROW, NPROW ),
Nq0 = NUMROC( N+ICOFF, NB_A, MYCOL, IACOL, NPCOL ),
and NUMROC, INDXG2P are ScaLAPACK tool functions;
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 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.
INFO (local 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.
Further Details
===============
The matrix Q is represented as a product of elementary reflectors
Q = H(ja+k-1) . . . H(ja+1) H(ja), where k = min(m,n).
Each H(i) has the form
H(i) = I - tau * v * v'
where tau is a real scalar, and v is a real vector with
v(m-k+i+1:m) = 0 and v(m-k+i) = 1; v(1:m-k+i-1) is stored on exit in
A(ia:ia+m-k+i-2,ja+n-k+i-1), and tau in TAU(ja+n-k+i-1).
=====================================================================
.. Parameters ..
|
|
|
|
001 SUBROUTINE PSGEQL2( M , N , A , IA , JA , DESCA , TAU , WORK , LWORK ,
002 $INFO )
003
004 * -- ScaLAPACK routine(version 1.7) --
005 * University of Tennessee , Knoxville , Oak Ridge National Laboratory ,
006 * and University of California , Berkeley.
007 * May 25 , 2001
008
009 * .. Scalar Arguments ..
010 INTEGER IA , INFO , JA , LWORK , M , N
011 INTEGER BLOCK_CYCLIC_2D , CSRC_ , CTXT_ , DLEN_ , DTYPE_ ,
012 $LLD_ , MB_ , M_ , NB_ , N_ , RSRC_
013 PARAMETER( BLOCK_CYCLIC_2D = 1 , DLEN_ = 9 , DTYPE_ = 1 ,
014 $CTXT_ = 2 , M_ = 3 , N_ = 4 , MB_ = 5 , NB_ = 6 ,
015 $RSRC_ = 7 , CSRC_ = 8 , LLD_ = 9 )
016 REAL ONE
017 PARAMETER( ONE = 1.0E + 0 )
018 * ..
019 * .. Local Scalars ..
020 LOGICAL LQUERY
021 CHARACTER COLBTOP , ROWBTOP
022 INTEGER I , IACOL , IAROW , ICTXT , II , J , JJ , K , LWMIN ,
023 $MP , MYCOL , MYROW , NPCOL , NPROW , NQ
024 REAL AJJ , ALPHA
025 * ..
026 * .. External Subroutines ..
027 EXTERNAL BLACS_ABORT , BLACS_GRIDINFO , CHK1MAT , INFOG2L ,
028 $PSELSET , PSLARF , PSLARFG , PB_TOPGET ,
029 $PB_TOPSET , PXERBLA , SGEBR2D , SGEBS2D ,
030 $SLARFG , SSCAL
031 * ..
032 * .. External Functions ..
033 INTEGER INDXG2P , NUMROC
034 EXTERNAL INDXG2P , NUMROC
035 * ..
036 * .. Intrinsic Functions ..
037 INTRINSIC MAX , MIN , MOD , REAL
038 * ..
039 * .. Executable Statements ..
040
041 * Get grid parameters
042
043 ICTXT = DESCA( CTXT_ )
044 CALL BLACS_GRIDINFO( ICTXT , NPROW , NPCOL , MYROW , MYCOL )
045
046 * Test the input parameters
047
048 INFO = 0
049 IF( NPROW.EQ. - 1 ) THEN
049
050 INFO = - (600 + CTXT_)
051 ELSE
051
052 CALL CHK1MAT( M , 1 , N , 2 , IA , JA , DESCA , 6 , INFO )
053 IF( INFO.EQ.0 ) THEN
053
054 IAROW = INDXG2P( IA , DESCA( MB_ ) , MYROW , DESCA( RSRC_ ) ,
055 $ NPROW )
056 IACOL = INDXG2P( JA , DESCA( NB_ ) , MYCOL , DESCA( CSRC_ ) ,
057 $ NPCOL )
058 MP = NUMROC( M + MOD( IA - 1 , DESCA( MB_ ) ) , DESCA( MB_ ) ,
059 $ MYROW , IAROW , NPROW )
060 NQ = NUMROC( N + MOD( JA - 1 , DESCA( NB_ ) ) , DESCA( NB_ ) ,
061 $ MYCOL , IACOL , NPCOL )
062 LWMIN = MP + MAX( 1 , NQ )
063
064 WORK( 1 ) = REAL( LWMIN )
065 LQUERY =( LWORK.EQ. - 1 )
066 IF( LWORK.LT.LWMIN .AND. .NOT.LQUERY )
066
067 $ INFO = - 9
068 END IF
069 END IF
070
071 IF( INFO.NE.0 ) THEN
071
072 CALL PXERBLA( ICTXT , 'PSGEQL2' , - INFO )
073 CALL BLACS_ABORT( ICTXT , 1 )
074 RETURN
075 ELSE IF( LQUERY ) THEN
075
076 RETURN
077 END IF
078
079 * Quick return if possible
080
081 IF( M.EQ.0 .OR. N.EQ.0 )
081
082 $ RETURN
083
084 CALL PB_TOPGET( ICTXT , 'Broadcast' , 'Rowwise' , ROWBTOP )
085 CALL PB_TOPGET( ICTXT , 'Broadcast' , 'Columnwise' , COLBTOP )
086 CALL PB_TOPSET( ICTXT , 'Broadcast' , 'Rowwise' , 'D - ring' )
087 CALL PB_TOPSET( ICTXT , 'Broadcast' , 'Columnwise' , ' ' )
088
089 IF( DESCA( M_ ).EQ.1 ) THEN
089
090 IF( MYCOL.EQ.IACOL )
090
091 $ NQ = NQ - MOD( JA - 1 , DESCA( NB_ ) )
092 CALL INFOG2L( IA , JA , DESCA , NPROW , NPCOL , MYROW , MYCOL , II ,
093 $ JJ , IAROW , IACOL )
094 IACOL = INDXG2P( JA + N - 1 , DESCA( NB_ ) , MYCOL , DESCA( CSRC_ ) ,
095 $ NPCOL )
096 IF( MYROW.EQ.IAROW ) THEN
096
097 IF( MYCOL.EQ.IACOL ) THEN
097
098 I = II + (JJ + NQ - 2)*DESCA( LLD_ )
099 AJJ = A( I )
100 CALL SLARFG( 1 , AJJ , A( I ) , 1 , TAU( JJ + NQ - 1 ) )
101 IF( N.GT.1 ) THEN
101
102 ALPHA = ONE - TAU( JJ + NQ - 1 )
103 CALL SGEBS2D( ICTXT , 'Rowwise' , ' ' , 1 , 1 , ALPHA , 1 )
104 CALL SSCAL( NQ - 1 , ALPHA , A( II + (JJ - 1)*DESCA( LLD_ ) ) ,
105 $ DESCA( LLD_ ) )
106 END IF
107 CALL SGEBS2D( ICTXT , 'Columnwise' , ' ' , 1 , 1 ,
108 $ TAU( JJ + NQ - 1 ) , 1 )
109 A( I ) = AJJ
110 ELSE
110
111 IF( N.GT.1 ) THEN
111
112 CALL SGEBR2D( ICTXT , 'Rowwise' , ' ' , 1 , 1 , ALPHA ,
113 $ 1 , IAROW , IACOL )
114 CALL SSCAL( NQ , ALPHA , A( II + (JJ - 1)*DESCA( LLD_ ) ) ,
115 $ DESCA( LLD_ ) )
116 END IF
117 END IF
118 ELSE IF( MYCOL.EQ.IACOL ) THEN
118
119 CALL SGEBR2D( ICTXT , 'Columnwise' , ' ' , 1 , 1 ,
120 $ TAU( JJ + NQ - 1 ) , 1 , IAROW , IACOL )
121 END IF
122
123 ELSE
124
124
125 K = MIN( M , N )
126 DO 10 J = JA + K - 1 , JA , - 1
126
127 I = IA + J - JA
128
129 * Generate elementary reflector H(j) to annihilate
130 * A(ia : i + m - k - 1 , j + n - k)
131
132 CALL PSLARFG ( M - K + I - IA + 1 , AJJ , M - K + I , N - K + J , A , IA ,
133 $ N - K + J , DESCA , 1 , TAU )
134
135 * Apply H(j) to A(ia : i + m - k , ja : j + n - k - 1) from the left
136
137 CALL PSELSET( A , I + M - K , J + N - K , DESCA , ONE )
138 CALL PSLARF ( 'Left' , M - K + I - IA + 1 , N - K + J - JA , A , IA , N - K + J ,
139 $ DESCA , 1 , TAU , A , IA , JA , DESCA , WORK )
140 CALL PSELSET( A , I + M - K , J + N - K , DESCA , AJJ )
141
142 10 CONTINUE
143
143
144 END IF
145
146 CALL PB_TOPSET( ICTXT , 'Broadcast' , 'Rowwise' , ROWBTOP )
147 CALL PB_TOPSET( ICTXT , 'Broadcast' , 'Columnwise' , COLBTOP )
148
149 WORK( 1 ) = REAL( LWMIN )
150
151 RETURN
152
153 * End of PSGEQL2
154
155 END25
18
|
|
Variables in Routine PSGEQL2()
| Summary Report |
| Data Type | Quantity | Size(byte) |
| CHARACTER | 2 | 2 |
| INTEGER | 34 | 136 |
| LOGICAL | 1 | 1 |
| REAL | 5 | 20 |
| TOTAL | 42 | 159 |
List of Variables
CHARACTER
INTEGER
| BLOCK_CYCLIC_2D | CSRC_ | CTXT_ | DLEN_ | DTYPE_ |
| I | IA | IACOL | IAROW | ICTXT |
| II | INDXG2P | INFO | J | JA |
| JJ | K | LLD_ | LWMIN | LWORK |
| M | M_ | MB_ | MP | MYCOL |
| MYROW | N | N_ | NB_ | NPCOL |
| NPROW | NQ | NUMROC | RSRC_ | |
LOGICAL
REAL
Variables Dependence Graph Put the mouse over a right hand side variable to display the corresponding line of the dependence | | - | | - | - | | A | <--- | AJJA( I ) = AJJ |
| AJJ | <--- | AAJJ = A( I ), IAJJ = A( I ) |
| ALPHA | <--- | JJALPHA = ONE - TAU( JJ+NQ-1 ), NQALPHA = ONE - TAU( JJ+NQ-1 ), ONEALPHA = ONE - TAU( JJ+NQ-1 ) |
| I | <--- | IAI = IA + J - JA, III = II+(JJ+NQ-2)*DESCA( LLD_ ), JI = IA + J - JA, JAI = IA + J - JA, JJI = II+(JJ+NQ-2)*DESCA( LLD_ ), LLD_I = II+(JJ+NQ-2)*DESCA( LLD_ ), NQI = II+(JJ+NQ-2)*DESCA( LLD_ ) |
| IACOL | <--- | INDXG2PIACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),}, JAIACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),}, MYCOLIACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),}, NIACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),, NB_IACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),}, NPCOLIACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),}, CSRC_IACOL = INDXG2P( JA, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),{2IACOL = INDXG2P( JA+N-1, DESCA( NB_ ), MYCOL, DESCA( CSRC_ ),} |
| IAROW | <--- | IAIAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ),, INDXG2PIAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ),, MB_IAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ),, MYROWIAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ),, NPROWIAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ),, RSRC_IAROW = INDXG2P( IA, DESCA( MB_ ), MYROW, DESCA( RSRC_ ), |
| ICTXT | <--- | CTXT_ICTXT = DESCA( CTXT_ ) |
| INFO | <--- | CTXT_INFO = -(600+CTXT_) |
| J | <--- | JADO 10 J = JA+K-1, JA, -1, KDO 10 J = JA+K-1, JA, -1 |
| K | <--- | MK = MIN( M, N ), NK = MIN( M, N ) |
| LWMIN | <--- | MPLWMIN = MP + MAX( 1, NQ ), NQLWMIN = MP + MAX( 1, NQ ) |
| MP | <--- | IAMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, IAROWMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, MMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, MB_MP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, MYROWMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, NPROWMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ),, NUMROCMP = NUMROC( M+MOD( IA-1, DESCA( MB_ ) ), DESCA( MB_ ), |
| NQ | <--- | IACOLNQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, JANQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, MYCOLNQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, NNQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, NB_NQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, NPCOLNQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ),, NUMROCNQ = NUMROC( N+MOD( JA-1, DESCA( NB_ ) ), DESCA( NB_ ), |
| WORK | <--- | LWMINWORK( 1 ) = REAL( LWMIN ){2WORK( 1 ) = REAL( LWMIN )} |
|
|
Analysis elements of the routine PSGEQL2() Put the mouse over each element to display detailed matching information
Assigned variables |
| | | AJJ , ALPHA , BLOCK_CYCLIC_2D , CSRC_ , CTXT_ , DLEN_ , DTYPE_ , I , IACOL , IAROW , ICTXT , INFO , J , K , LLD_ , LQUERY , LWMIN , M_ , MB_ , MP , N_ , NB_ , NQ , ONE , RSRC_ , WORK |
|
Active variables |
| | | A , AJJ , ALPHA , BLOCK_CYCLIC_2D , COLBTOP , CSRC_ , CTXT_ , DESCA , DLEN_ , DTYPE_ , I , IA , IACOL , IAROW , ICTXT , II , INDXG2P , INFO , J , JA , JJ , K , LLD_ , LQUERY , LWMIN , LWORK , M , M_ , MB_ , MP , MYCOL , MYROW , N , N_ , NB_ , NPCOL , NPROW , NQ , NUMROC , ONE , ROWBTOP , RSRC_ , TAU , WORK |
|
Accessed arrays [ array name : associated index ] |
| | A | : I , I , I , ia:i+m-k,ja:j+n-k-1 , ia:i+m-k-1,j+n-k , II+(JJ-1)*DESCA( LLD_ ) , II+(JJ-1)*DESCA( LLD_ ) |
| | DESCA | : CSRC_ , CSRC_ , CTXT_ , LLD_ , LLD_ , LLD_ , LLD_ , LLD_ , M_ , MB_ , MB_ , NB_ , NB_ , NB_ , NB_ , RSRC_ |
| | TAU | : JJ+NQ-1 , JJ+NQ-1 , JJ+NQ-1 , JJ+NQ-1 |
| | WORK | : 1 , 1 |
|
Conditional statements [ statement : associated predicate ] |
| | do | : ( 10 J = JA + K - 1 , JA , - 1 ) |
| | if | : ( NPROW.EQ. - 1 ) , ( INFO.EQ.0 ) , ( LWORK.LT.LWMIN .AND. .NOT.LQUERY ) , ( INFO.NE.0 ) , ( LQUERY ) , ( possible ) , ( M.EQ.0 .OR. N.EQ.0 ) , ( (DESCA( M_ ).EQ.1 ) ) , ( MYCOL.EQ.IACOL ) , ( MYROW.EQ.IAROW ) , ( MYCOL.EQ.IACOL ) , ( N.GT.1 ) , ( N.GT.1 ) , ( MYCOL.EQ.IACOL ) |
|
| List of variables | A AJJ ALPHA BLOCK_CYCLIC_2D COLBTOP CSRC_ CTXT_
| DLEN_ DTYPE_ I IA IACOL IAROW ICTXT II
| INDXG2P INFO J JA JJ K LLD_ LQUERY
| LWMIN LWORK M M_ MB_ MP MYCOL MYROW
| N N_ NB_ NPCOL NPROW NQ NUMROC ONE
| ROWBTOP RSRC_ WORK | | close
| |
A
AJJ
ALPHA
BLOCK_CYCLIC_2D
COLBTOP
CSRC_
CTXT_
DLEN_
DTYPE_
I
IA
IACOL
IAROW
ICTXT
II
INDXG2P
INFO
J
JA
JJ
K
LLD_
LQUERY
LWMIN
LWORK
M
M_
MB_
MP
MYCOL
MYROW
N
N_
NB_
NPCOL
NPROW
NQ
NUMROC
ONE
ROWBTOP
RSRC_
WORK
384#382
| |