[BACK]Return to dlag2.f CVS log [TXT] [DIR] Up to [local] / rpl / lapack / lapack

Diff for /rpl/lapack/lapack/dlag2.f between versions 1.13 and 1.14

version 1.13, 2014/01/27 09:28:19 version 1.14, 2016/08/27 15:27:09
Line 99 Line 99
 *>          will always be positive.  If the eigenvalues are real, then  *>          will always be positive.  If the eigenvalues are real, then
 *>          the first (real) eigenvalue is  WR1 / SCALE1 , but this may  *>          the first (real) eigenvalue is  WR1 / SCALE1 , but this may
 *>          overflow or underflow, and in fact, SCALE1 may be zero or  *>          overflow or underflow, and in fact, SCALE1 may be zero or
 *>          less than the underflow threshhold if the exact eigenvalue  *>          less than the underflow threshold if the exact eigenvalue
 *>          is sufficiently large.  *>          is sufficiently large.
 *> \endverbatim  *> \endverbatim
 *>  *>
Line 112 Line 112
 *>          eigenvalues are real, then the second (real) eigenvalue is  *>          eigenvalues are real, then the second (real) eigenvalue is
 *>          WR2 / SCALE2 , but this may overflow or underflow, and in  *>          WR2 / SCALE2 , but this may overflow or underflow, and in
 *>          fact, SCALE2 may be zero or less than the underflow  *>          fact, SCALE2 may be zero or less than the underflow
 *>          threshhold if the exact eigenvalue is sufficiently large.  *>          threshold if the exact eigenvalue is sufficiently large.
 *> \endverbatim  *> \endverbatim
 *>  *>
 *> \param[out] WR1  *> \param[out] WR1
Line 148 Line 148
 *> \author Univ. of Colorado Denver   *> \author Univ. of Colorado Denver 
 *> \author NAG Ltd.   *> \author NAG Ltd. 
 *  *
 *> \date September 2012  *> \date June 2016
 *  *
 *> \ingroup doubleOTHERauxiliary  *> \ingroup doubleOTHERauxiliary
 *  *
Line 156 Line 156
       SUBROUTINE DLAG2( A, LDA, B, LDB, SAFMIN, SCALE1, SCALE2, WR1,        SUBROUTINE DLAG2( A, LDA, B, LDB, SAFMIN, SCALE1, SCALE2, WR1,
      $                  WR2, WI )       $                  WR2, WI )
 *  *
 *  -- LAPACK auxiliary routine (version 3.4.2) --  *  -- LAPACK auxiliary routine (version 3.6.1) --
 *  -- LAPACK is a software package provided by Univ. of Tennessee,    --  *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
 *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--  *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
 *     September 2012  *     June 2016
 *  *
 *     .. Scalar Arguments ..  *     .. Scalar Arguments ..
       INTEGER            LDA, LDB        INTEGER            LDA, LDB
Line 266 Line 266
 *     Note: the test of R in the following IF is to cover the case when  *     Note: the test of R in the following IF is to cover the case when
 *           DISCR is small and negative and is flushed to zero during  *           DISCR is small and negative and is flushed to zero during
 *           the calculation of R.  On machines which have a consistent  *           the calculation of R.  On machines which have a consistent
 *           flush-to-zero threshhold and handle numbers above that  *           flush-to-zero threshold and handle numbers above that
 *           threshhold correctly, it would not be necessary.  *           threshold correctly, it would not be necessary.
 *  *
       IF( DISCR.GE.ZERO .OR. R.EQ.ZERO ) THEN        IF( DISCR.GE.ZERO .OR. R.EQ.ZERO ) THEN
          SUM = PP + SIGN( R, PP )           SUM = PP + SIGN( R, PP )
Removed from v.1.13  
changed lines
  Added in v.1.14

CVSweb interface <joel.bertrand@systella.fr>