internal compiler error

Tom Hull thull@kscable.com
Tue Nov 30 23:39:00 GMT 1999


I encountered the following error whilst trying to built ftwalk-1.5.2
(see http://www.ocston.org/~thull/ftwalk/download.html ).

Error message:

  ft_ios.C: In function `int ostream_special(class ostream &, class ft_objt &)':
  ft_ios.C:371: Internal compiler error.
  ft_ios.C:371: Please submit a full bug report to 'egcs-bugs@egcs.cygnus.com'.
  ft_ios.C:371: See <URL: http://egcs.cygnus.com/faq.html#bugreport > for details.
  make[1]: *** [ft_ios.o] Error 1
  make[1]: Leaving directory `/home/tom/ftwork/ftwalk-1.5.2/src'
  make: *** [ftwalk] Error 2

GCC Version:

  $ g++ --version
  egcs-2.91.66

System Type:

  $ uname -a
  Linux tux.kinich.com 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999 i686 unknown

I.e., Intel Pentium II 333MHz, 128MB RAM, Red Hat Linux 6.1

All options you passed to the compiler:

  c++ -c -g -O2 -I. -D_GNU_SOURCE=1 ft_ios.C

Preprocessed output of the source file that caused the compiler error:

See attachment.

Note: I reworked the code a bit, which provides a workaround to this
problem:

  class ft_enum : public ft_objt {
    public:
      ...
      int operator () () const { return int(*this); }
+     int val() const          { return (int) *this; }
      ...
  }

  static int ostream_special(ostream &s, ft_objt &o)
  {
      if (o.type() == t_enum) {
-         int n = int(o);
+         int n = (int) o;
-         if (n == sym_endl()) {
+         if (n == sym_endl.val()) {
              s << endl;
              return 1;
          }
-         else if (n == sym_flush()) {
+         else if (n == sym_flush.val()) {
              s << flush;
              return 1;
          }
      }
      return 0;
  }

-- Tom Hull - thull@kscable.com
ft_ios.ii.bz2

-------------- next part --------------
A non-text attachment was scrubbed...
Name: ft_ios.ii.bz2
Type: application/x-bzip2
Size: 23244 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-bugs/attachments/19991130/e0684b78/attachment-0001.bz2>


More information about the Gcc-bugs mailing list