------- Comment #4 from janus at gcc dot gnu dot org 2010-06-08 01:19 ------- The backtrace for comment #3 is:
#0 gfc_add_component_ref (e=0x1806e80, name=0x7ffff7e7ceb8 "doit") at /home/jweil/gcc46/trunk/gcc/fortran/class.c:77 #1 0x000000000052b323 in resolve_typebound_subroutine (code=0x1806bf0) at /home/jweil/gcc46/trunk/gcc/fortran/resolve.c:5530 #2 0x0000000000530ed4 in resolve_code (code=0x1806bf0, ns=0x17fb030) at /home/jweil/gcc46/trunk/gcc/fortran/resolve.c:8454 #3 0x000000000053acfa in resolve_codes (ns=0x17fb030) at /home/jweil/gcc46/trunk/gcc/fortran/resolve.c:12837 #4 0x000000000053abeb in resolve_codes (ns=0x17febe0) at /home/jweil/gcc46/trunk/gcc/fortran/resolve.c:12823 #5 0x000000000053adf5 in gfc_resolve (ns=0x17febe0) at /home/jweil/gcc46/trunk/gcc/fortran/resolve.c:12864 #6 0x0000000000519bbf in gfc_parse_file () at /home/jweil/gcc46/trunk/gcc/fortran/parse.c:4275 #7 0x000000000055a7bb in gfc_be_parse_file (set_yydebug=0) at /home/jweil/gcc46/trunk/gcc/fortran/f95-lang.c:234 #8 0x0000000000a19514 in compile_file () at /home/jweil/gcc46/trunk/gcc/toplev.c:1042 #9 0x0000000000a1b80e in do_compile () at /home/jweil/gcc46/trunk/gcc/toplev.c:2407 #10 0x0000000000a1b8e4 in toplev_main (argc=2, argv=0x7fffffffe2b8) at /home/jweil/gcc46/trunk/gcc/toplev.c:2449 #11 0x00000000005e28b0 in main (argc=2, argv=0x7fffffffe2b8) at /home/jweil/gcc46/trunk/gcc/main.c:35 Apparently the problem is that the generic's vtype still is '$empty', while it should be 'vtype$foo$do'. Paul, can you remind me why we even need this $empty thing? -- janus at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |pault at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44434