forums.silverfrost.com

[shahid]

hi,

i am using personal version of salford FTN95.i am compiling one program but it is not compiling and giving the following error message.

error 137 - Operand incompatible with opcode

and it is pointing at one of the expression in the following function:


REAL FUNCTION FA(X,Y)
! .. Scalar Arguments ..
REAL X, Y, R
complex VALUE
REAL pi,umu,ep0,c,Frequency,Alam,kay0,er,a,m_l,m_u,m_m,n_l,n_u,n_m,zt,zs
COMPLEX jay
INTEGER ReIm
COMMON /C1/ pi,umu,ep0,c,Frequency,Alam,kay0,er,a,jay,ReIm,m_l,m_u,m_m,n_l,n_u,n_m,zt,zs
INTRINSIC SQRT,CEXP
R = SQRT(a*a+(X-Y)*(X-Y))
VALUE = 1.0/(4.*pi*R**5)*(cos(kay0*R)-jay*sin(kay0*R))
VALUE = VALUE * ((1+jay*kay0*R)*(2.*R**2-3*a**2)+(kay0*a*R)**2)
VALUE = VALUE*sin(kay0(X-n_l))*sin(kay0(y-m_l))

IF(ReIm.EQ.0)THEN
FA=REAL(VALUE)
ELSE
FA=AIMAG(VALUE)
ENDIF

RETURN
END


the error is pointed at this expression : VALUE = 1.0/(4.*pi*R**5)*(cos(kay0*R)-jay*sin(kay0*R))

i dont know whats the cause of this error mesage.
any help would be highly appreciated.

thanks

[JohnCampbell]

Try

REAL FUNCTION FA (X,Y)
! .. Scalar Arguments ..
REAL X, Y
!
COMMON /C1/ pi,umu,ep0,c,Frequency,Alam,kay0,er,a,jay,ReIm,m_l,m_u,m_m,n_l,n_u,n_m,zt,zs
REAL pi,umu,ep0,c,Frequency,Alam,kay0,er,a,m_l,m_u,m_m,n_l,n_u,n_m,zt,zs
COMPLEX jay
INTEGER ReIm
!
INTRINSIC SQRT,COS,SIN
real r, co, si, t1, t2
complex VALUE
!
R = SQRT(a*a+(X-Y)*(X-Y))

! VALUE = 1.0/(4.*pi*R**5)*(cos(kay0*R)-jay*sin(kay0*R))
t1 = 1.0/(4.*pi*R**5)
co = cos(kay0*R)
si = sin(kay0*R)
value = t1 * co - jay*t1*si
!
! VALUE = VALUE * ((1+jay*kay0*R)*(2.*R**2-3*a**2)+(kay0*a*R)**2)
t1 = 2.*R**2-3*a**2
t2 = (kay0*a*R)**2
VALUE = VALUE*(t1+t2) + jay * (value*t1*kay0*R)
!
! VALUE = VALUE*sin(kay0*(X-n_l))*sin(kay0*(y-m_l))
t1 = sin(kay0*(X-n_l))
t2 = sin(kay0*(y-m_l))
value = value * t1 * t2
!
IF (ReIm == 0)THEN
FA=REAL(VALUE)
ELSE
FA=AIMAG(VALUE)
ENDIF

RETURN
END

[PaulLaidler]

Shahid

The error is on the line

VALUE = VALUE*sin(kay0(X-n_l))*sin(kay0(y-m_l))

which should be

VALUE = VALUE*sin(kay0*(X-n_l))*sin(kay0*(y-m_l))

We have now changed the compiler so that it gives a sensible diagnostic in this situation.

[shahid]

thank you very much Paul.
you are absolutely right.
after i make this change my program is successfully compiled.

regards,
shahid