[Bug fortran/86220] [9 Regression] ICE in gfc_conv_structure, at fortran/trans-expr.c:7789

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86220

Dominique d'Humieres  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |DUPLICATE

--- Comment #3 from Dominique d'Humieres  ---
I think this is a duplicate of pr49278: mixing initialization with data is
invalid.
Replacing

  real :: a = 1

with

  real :: a

allows the test to compile and give the right result.

*** This bug has been marked as a duplicate of bug 49278 ***

[Bug fortran/86220] [9 Regression] ICE in gfc_conv_structure, at fortran/trans-expr.c:7789

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86220

Dominique d'Humieres  changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEW
   Last reconfirmed||2018-07-08
 Ever confirmed|0   |1

--- Comment #2 from Dominique d'Humieres  ---
> > Introduced between 20180520 and 20180527.
>
> It wasn't introduced.  It was uncovered. :-)

May be r260808.

Note that the test compiles, but gives a wrong result with the patch in pr49278
comment 2.

[Bug fortran/86220] [9 Regression] ICE in gfc_conv_structure, at fortran/trans-expr.c:7789

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86220

Richard Biener  changed:

   What|Removed |Added

   Priority|P3  |P4
   Target Milestone|--- |9.0

[Bug fortran/86220] [9 Regression] ICE in gfc_conv_structure, at fortran/trans-expr.c:7789

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86220

--- Comment #1 from kargl at gcc dot gnu.org ---
(In reply to G. Steinmetz from comment #0)
> Introduced between 20180520 and 20180527.

It wasn't introduced.  It wa uncovered. :-)

> 
> $ cat z1.f90
> program p
>type t
>   real :: a = 1
>   integer, pointer :: b
>end type
>type(t) :: z
>integer, save, target :: x = 789
>data z%b /x/
> end
> 

The code is invalid Fortran, so technically gfortran
can do anything including ICE.