[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gcc-bugs
Subject:    [Bug fortran/91716] [9/10 Regression] ICE in output_constant, at varasm.c:5026
From:       "bernd.edlinger at hotmail dot de" <gcc-bugzilla () gcc ! gnu ! org>
Date:       2019-09-30 17:44:48
Message-ID: bug-91716-4-lR2fww9zVK () http ! gcc ! gnu ! org/bugzilla/
[Download RAW message or body]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91716

--- Comment #7 from Bernd Edlinger <bernd.edlinger at hotmail dot de> ---
Hmm, I tried it out and built gcc-9.2.0 out of the release tar ball,
with no checking flag....
and, actually the test case still ICEs, just in a different
place:

$ gfortran -c z1.f90
z1.f90:5:0:

    5 | end
      |
internal compiler error: in output_constructor_regular_field, at varasm.c:5207
0x5e6e68 output_constructor_regular_field
        ../../gcc-9.2.0/gcc/varasm.c:5207
0x5e6e68 output_constructor
        ../../gcc-9.2.0/gcc/varasm.c:5513
0xdac6aa output_constant
        ../../gcc-9.2.0/gcc/varasm.c:4885
0xdac6aa output_constructor_regular_field
        ../../gcc-9.2.0/gcc/varasm.c:5246
0xdac6aa output_constructor
        ../../gcc-9.2.0/gcc/varasm.c:5513
0xdad03c output_constant
        ../../gcc-9.2.0/gcc/varasm.c:4885
0xdad03c assemble_variable_contents
        ../../gcc-9.2.0/gcc/varasm.c:2144
0xdb3b97 assemble_variable(tree_node*, int, int, int)
        ../../gcc-9.2.0/gcc/varasm.c:2323
0xdb59e9 varpool_node::assemble_decl()
        ../../gcc-9.2.0/gcc/varpool.c:584
0xdb59e9 varpool_node::assemble_decl()
        ../../gcc-9.2.0/gcc/varpool.c:552
0x7cc17c output_in_order
        ../../gcc-9.2.0/gcc/cgraphunit.c:2446
0x7cc17c symbol_table::compile()
        ../../gcc-9.2.0/gcc/cgraphunit.c:2686
0x7ce14c symbol_table::compile()
        ../../gcc-9.2.0/gcc/cgraphunit.c:2599
0x7ce14c symbol_table::finalize_compilation_unit()
        ../../gcc-9.2.0/gcc/cgraphunit.c:2865
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.


And the patch applies cleanly and fixes the issue.

So I guess I will look at running a reg-test cycle
and send the patch to the fortran mailing list when that is
done.=
[prev in list] [next in list] [prev in thread] [next in thread]