| Re: Q Re using gfortran instead of g77 |
|
 |
|
 |
|
 |
|
 |
Group: comp.lang.fortran · Group Profile
Author: Tobias BurnusTobias Burnus Date: Sep 21, 2007 03:42
On Sep 19, 9:32 pm, Craig Powers hal-pc.org> wrote:>>> gfortran will have such a switch for version 4.3. I've seen the
>>> patch proposed, but I don't recall offhand if it's included in
>>> the source tree yet.
It is but only since a few hours. (Thus the nightly binaries might not
have it, yet.)
>> Does it allow one to initialize to other than zero?
> As I recall, yes. There's also a plan to eventually allow initializing
> real variables to snan or qnan, but I don't think that's going to make
> it in for 4.3.
It allows other values including qNaN, but unfortunately not yet sNaN.
Available options:
-finit-local-zero
-finit-integer=n
-finit-real=
-finit-logical=
-finit-character=n
The -finit-local-zero option instructs the compiler to
initialize local "INTEGER", "REAL", and "COMPLEX" vari-
ables to zero, "LOGICAL" variables to false, and "CHARAC-
TER" variables to a string of null bytes. Finer-grained
initialization options are provided by the -finit-inte-
ger=n, -finit-real= (which also ini-
tializes the real and imaginary parts of local "COMPLEX"
variables), -finit-logical=, and -finit-char-
acter=n (where n is an ASCII character value) options.
These options do not initialize components of derived
type variables, nor do they initialize variables that
appear in an "EQUIVALENCE" statement. (This limitation
may be removed in future releases).
Note that the -finit-real=nan option initializes "REAL"
and "COMPLEX" variables with a quiet NaN.
Tobias
| 
