This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, fortran] PR20441 -finit-local-zero
- From: "Asher Langton" <langton2 at llnl dot gov>
- To: "Tobias Burnus" <burnus at net-b dot de>
- Cc: fortran at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Date: Fri, 14 Sep 2007 07:50:33 -0500
- Subject: Re: [PATCH, fortran] PR20441 -finit-local-zero
- References: <1953a4560708171312u705f5bf4ie9d8ce645d2f02b6@mail.gmail.com> <1953a4560708311456g550c60a8xd4a0ffa9335c4325@mail.gmail.com> <46EA4E98.2050809@net-b.de>
On 9/14/07, Tobias Burnus <burnus@net-b.de> wrote:
> + else if (!strcmp (arg, "nan"))
>
> I am wondering whether one should use strcasecmp instead (as people like
> to write INF, NaN,...), but I do not have a real opinion on this.
That's a good idea, and it certainly won't harm anything.
>
> +@option{-finit-real=@var{<zero|inf|-inf|nan>}} (which also initializes
> +the real and imaginary parts of local @code{COMPLEX} variables),
>
> I wonder whether one should add something like
> "Note: @var{nan} initializes REAL and COMPLEX variables with a quiet NaN."
> as some might expect that the NaN are signalling.*
I'll add something like this. Eventually, I'd like to add
finit-real=<snan,qnan> options, but so far I haven't come up with a
nice kludge to deal with sNaNs in mpfr.
-Asher