Dear All,
The fix for the failing testcases in comments #6 and #7 in this PR
have been fixed as 'obvious' in r220140.
Thanks to Mikael for reopening the PR as a placeholder.
This is such an embarrassing bug that I intend to commit to 4.9 as
well even if it is not a regression, unless there are objections.
Notice that in 4.9, the testcase will change to that of comment #7.
Cheers
Paul
2015-01-26 Paul Thomas <pa...@gcc.gnu.org>
PR fortran/62044
* resolve.c (resolve_allocate_expr): If the default initializer
is NULL, keep the original MOLD expression so that the correct
typespec is available.
2015-01-26 Paul Thomas <pa...@gcc.gnu.org>
PR fortran/62044
* gfortran.dg/allocate_with_mold_1.f90: New test
On 25 January 2015 at 14:49, Mikael Morin <mikael.mo...@sfr.fr> wrote:
> Hello,
>
> I have had a look at PR62044 where the compiler ICEs when loading the
> extensions of a derived type from a module file, because one use-renamed
> extended type is not found under its original name.
>
> After scratching my head on the tree of extended types (rooted at
> derived->f2k_derived->sym_root), wondering whether the renamed or
> original name should be used, I decided to remove all that nonsense.
>
> So here is the patch, regression tested on x86_64-unknown-linux-gnu.
> There is a failure, graphite/pr42393.f90, also present on mainline.
> OK for trunk/4.9/4.8 ?
>
> Mikael
>
>
--
Outside of a dog, a book is a man's best friend. Inside of a dog it's
too dark to read.
Groucho Marx
