This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran 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]

Fwd: [EXTERNAL] Re: [Patch, F03, RFC] FINAL support in 4.8


Hi all,

updated FINAL patch in attachment.

I did not make as much progress as I had hoped (due to rather tight
time constraints). Relative to the last version I sent, I mostly did
some cleanup.


> Note that the patch also does end-of-program finalization as a GNU
> extension (i.e. it is done with -std=gnu, but not with -std=f2003),


I have removed these parts now, in order to be fully compatible with
the standard.

In related news, I have opened PR 55207 in order to track the issues
with auto-deallocation and implicit SAVE (which also affect
finalization).


>
> At this point, I would say, the major to-do item is to implement the
> missing six items of F08:4.5.6.3, namely:
>  * finalization of local vars at the end of a BLOCK construct
>  * finalization of function results (in executable statements &
> specifications expr's)
>  * finalization of structure/array constructors (in executable
> statements & specification expr's)
>  * finalization in intrinsic assignments


These to-do items are still open.


Unfortunately I will not have any time in the coming weeks to take
this patch any further. So, if this is still supposed to make it in
the 4.8 release, I hope that someone else can pick it up from here.

If this is not the case and/or we decide that finalization is not
ready for prime-time in 4.8 after all, I would at least try to single
out the parts that are necessary for polymorphic deallocation (which
is definitely the least we should do for 4.8).

Feedback welcome ...

Cheers,
Janus

Attachment: final-2012-11-04.diff
Description: Binary data


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