[Patch, Fortran] PR fortran/35837: Fix type-resolving and gfc_current_ns

[Daniel Kraft]

Tobias Burnus wrote:
> Hi Daniel,
> 
> Daniel Kraft wrote:
>> Currently regression testing on GNU/Linux-x86-32, and
>> I hope there won't be any despite this might have some
>> consequences; but I hope all are positive :)
>> Ok to commit if no regressions?
> 
> OK. (Let's see whether a user finds out that it causs a regression ;-)
> 
>> PS: The gfc_save_all change remained from some experiments there;
>> I can take it out if you prefer. Should I add a ChangeLog entry
>> like "removed blank line"?
> 
> Leave it in, but I think the rules say that one should add a
> ChangeLog item.

Committed as revision 140046, added that ChangeLog entry of course. 
Thanks for all your reviews the last time :)

> Thanks for this patch and thanks for looking into the F95 bugs!
> (Currently, we have 23 "Fortran95 bugs" and 4 regressions w/o patch)

Well, the last ones were rather "easy" and went more or less smooth, 
luckily.  I hope we can get them down a bit during this stage3.

> PS: I wonder why the number of new bug reports is quite low. Is there
> no intensive testing anymore? Or is it indeed that difficult to hit
> bugs in gfortran? (Well, every day programs clearly work.) I think
> in a few weeks we should pre-announce GCC 4.4 in c.l.f and call for
> testing it (in general plus new features). By then GCC should have
> stabilized and we have still enough time to address newly found bugs.
> -- Hopefully, Stage3 will be shorter than last time (5 months).

Sounds like a good idea to me!  And yes, 5 months sound really a bit 
long...  For the final 4.4 release, there'll be another delay after 
branching off the 4.4 release, won't there?

Daniel

-- 
Done:     Arc-Bar-Cav-Sam-Val-Wiz, Dwa-Elf-Gno-Hum-Orc, Law-Neu-Cha, Fem-Mal
To go:    Hea-Kni-Mon-Pri-Ran-Rog-Tou