This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: PR fortran/19269 (partial fix): Lengths of reshaped arrays


Tobias Schlüter <tobias.schlueter@physik.uni-muenchen.de> writes:
> (I assume you meant resolve_character_array_constructor, I can't seem
> to find the function you're quoting, and the name looks fairly
> nonsensical: resolve -> frontend, descriptor -> code generation.)

OK, OK, give me a break ;)  It's nearly 11 here.

> For PARAMETERs the characters may have different length, and this
> function only determines the maximum length in the constant case, at
> least if this comment is correct:
>
>      /* Find the maximum length of the elements. Do nothing for
>         variable array constructor.  */

Right.  I assumed it was just trying to save people from adding
explicit blanks.

> I don't think it is an extension that we want.

Fair enough.  Like I say, I really hadn't expected the testcase
to be so controversial, and the variable length thing has nothing
to do with the bug that I'm trying to fix.

So would the patch be OK with a testcase in which the constants
were all of equal length?  I can remove the maximum code from
resolve_character_array_constructor too if that's thought to be
the best thing.

Richard


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]