(j3.2006) Question about IEEE assignment and derived types
Fri Dec 14 12:31:24 EST 2007
Dick Hendrickson wrote:
> This is a multi-part question and everything depends on the
> answer to the first question. If that is NO, then everything
> that follows isn't interesting.
> 1) It's my understanding that IEEE conformance requires raising
> an exception whenever a NaN is used (assuming that the correct
> IEEE modules are used and that enough IEEE_SET routines have
> been called). If so, does this mean that
> A = B
> must raise an exception if B is a NaN?
No. Assignment is not an "operation" and the IEEE standard explicitly
says that raising an exception in this case in not required.
> If NO, hit the delete key now!
> 2) Then obviously
> C%E = D%E
> must raise an exception if C and D are of a type that has a real
> component and D%E is a NaN.
> 3) Again, then obviously
> C = D
> must raise an exception if intrinsic assignment is used and
> D%E is still a NaN
> 4) How about if defined assignment is used? The defined
> assignment section (18.104.22.168.3) says C=D is equivalent to
> CALL USER_SUB (C, (D))
> 12.5.4 says that the argument expression, (D), is evaluated.
> Does that mean that an exception must be raised on the calling
> side? Can't be raised, or is it processor dependent?
> Can the expression be evaluated without using the value of
> its elements?
> 5) How about allocation with a source = term.
> ALLOCATE (C, SOURCE = D)
> 22.214.171.124 says the value of C "becomes" that of D. Again,
> should an exception be raised if D%E is a NaN?
> Dick Hendrickson
> J3 mailing list
> J3 at j3-fortran.org
Bill Long longb at cray.com
Fortran Technical Support & voice: 651-605-9024
Bioinformatics Software Development fax: 651-605-9142
Cray Inc., 1340 Mendota Heights Rd., Mendota Heights, MN, 55120
More information about the J3