DLAZQ4(3) an approximation TAU to the smallest eigenvalue using values of d from the previous transform

SYNOPSIS

SUBROUTINE DLAZQ4(
I0, N0, Z, PP, N0IN, DMIN, DMIN1, DMIN2, DN, DN1, DN2, TAU, TTYPE, G )

    
INTEGER I0, N0, N0IN, PP, TTYPE

    
DOUBLE PRECISION DMIN, DMIN1, DMIN2, DN, DN1, DN2, G, TAU

    
DOUBLE PRECISION Z( * )

PURPOSE

DLAZQ4 computes an approximation TAU to the smallest eigenvalue using values of d from the previous transform.

I0 (input) INTEGER

      First index.

N0 (input) INTEGER

      Last index.

Z (input) DOUBLE PRECISION array, dimension ( 4*N )

      Z holds the qd array.

PP (input) INTEGER

      PP=0 for ping, PP=1 for pong.

N0IN (input) INTEGER

      The value of N0 at start of EIGTEST.

DMIN (input) DOUBLE PRECISION

      Minimum value of d.

DMIN1 (input) DOUBLE PRECISION

      Minimum value of d, excluding D( N0 ).

DMIN2 (input) DOUBLE PRECISION

      Minimum value of d, excluding D( N0 ) and D( N0-1 ).

DN (input) DOUBLE PRECISION

      d(N)

DN1 (input) DOUBLE PRECISION

      d(N-1)

DN2 (input) DOUBLE PRECISION

      d(N-2)

TAU (output) DOUBLE PRECISION

      This is the shift.

TTYPE (output) INTEGER

      Shift type.

G (input/output) DOUBLE PRECISION

      G is passed as an argument in order to save its value between
      calls to DLAZQ4

FURTHER DETAILS

This is a thread safe version of DLASQ4, which passes G through the argument list in place of declaring G in a SAVE statment.