[Bug fortran/50410] [8/9/10/11 Regression] ICE in record_reference, pointer variable in data statement

kargl at gcc dot gnu.org gcc-bugzilla@gcc.gnu.org
Mon May 11 06:09:56 GMT 2020


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=50410

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
                 CC|                            |kargl at gcc dot gnu.org
         Resolution|FIXED                       |---
   Target Milestone|8.5                         |11.0

--- Comment #39 from kargl at gcc dot gnu.org ---
The bug is not fixed.  The code from comment #3

% cat a.f90
     type t
       integer g
      end type
      type(t) :: u=t(1)
      data u%g /2/
      end
troutmask:sgk[205] gfcx -c a.f90
a.f90:1:0:

    1 |      type t
      | 
internal compiler error: Segmentation fault
0xdf6e3a crash_signal
        ../../gccx/gcc/toplev.c:328
0x8acbae gfc_conv_structure(gfc_se*, gfc_expr*, int)
        ../../gccx/gcc/fortran/trans-expr.c:8528
0x8c1f37 gfc_conv_initializer(gfc_expr*, gfc_typespec*, tree_node*, bool, bool,
bool)
        ../../gccx/gcc/fortran/trans-expr.c:7679
0x8a0020 gfc_get_symbol_decl(gfc_symbol*)
        ../../gccx/gcc/fortran/trans-decl.c:1892
0x8a33e5 generate_local_decl
        ../../gccx/gcc/fortran/trans-decl.c:5894
0x85a0e2 do_traverse_symtree
        ../../gccx/gcc/fortran/symbol.c:4165
0x8a454c generate_local_vars
        ../../gccx/gcc/fortran/trans-decl.c:6100
0x8a454c gfc_generate_function_code(gfc_namespace*)
        ../../gccx/gcc/fortran/trans-decl.c:6758
0x81e516 translate_all_program_units
        ../../gccx/gcc/fortran/parse.c:6302
0x81e516 gfc_parse_file()
        ../../gccx/gcc/fortran/parse.c:6541
0x871858 gfc_be_parse_file
        ../../gccx/gcc/fortran/f95-lang.c:210

The code in the first example in comment #9 also causes an ICE.

Codes z1.f90, z2.f90, and z3.f90 in comment #23 cause ICEs.


More information about the Gcc-bugs mailing list