public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
* pre-release-g++: cold ICE
@ 1998-03-08 13:06 Max Lawson
  1998-03-10 18:13 ` Mark Mitchell
  0 siblings, 1 reply; 3+ messages in thread
From: Max Lawson @ 1998-03-08 13:06 UTC (permalink / raw)
  To: egcs; +Cc: egcs-bugs

typedef unsigned int size_t;


struct dummy { };

struct arrrrrgh { };


template<class Par,class Rand = arrrrrgh>
struct whyyyyyyy { };	 




template<class T, class S =dummy> 
struct grrrrrrrr { };         


template<class Par, class Par2 =Par, class Rand =arrrrrgh>
class no_future
{
public:

   
  template<class S>
  no_future(const grrrrrrrr<whyyyyyyy<Par,Rand>*,S>& man )  { }

  ~no_future( ) { }

private:

   
  no_future(const no_future&);
  no_future& operator=(const no_future&);
};	 
 

int main( )
{
  grrrrrrrr<whyyyyyyy<double>*> man;

  no_future<double> here(man);

  return 0;
}


/*

  Here I am again :-) 

the above program give an ICE output when I try to compile 
with the -g flag on (and u guess it, I need this flag to debug 
a program which is segfaulting):

$ g++ -g bug.cc
bug.cc:42: Internal compiler error.
bug.cc:42: Please submit a full bug report to `egcs-bugs@cygnus.com'.

the implied compiler is 
$ g++ -v          
Reading specs from /usr/local/egcs-release-cvs-980303/lib/gcc-lib/i586-pc-linux-gnulibc1/egcs-2.90.25/specs
gcc version egcs-2.90.25 980302 (egcs-1.0.2 prerelease)


Max

*/

^ permalink raw reply	[flat|nested] 3+ messages in thread

* pre-release-g++: cold ICE
  1998-03-08 13:06 pre-release-g++: cold ICE Max Lawson
@ 1998-03-10 18:13 ` Mark Mitchell
  1998-03-11 16:23   ` Jim Wilson
  0 siblings, 1 reply; 3+ messages in thread
From: Mark Mitchell @ 1998-03-10 18:13 UTC (permalink / raw)
  To: Max Lawson; +Cc: egcs, egcs-bugs

>>>>> "Max" == Max Lawson <mlawson@drfmc.ceng.cea.fr> writes:

    Max> the above program give an ICE output when I try to compile
    Max> with the -g flag on (and u guess it, I need this flag to
    Max> debug a program which is segfaulting):

The problem was that dbxout_type_methods did not deal gracefully with
a TEMPLATE_DECL.  I've attached a patch.  Jeff, may I check this in?

-- 
Mark Mitchell		mmitchell@usa.net
Stanford University	http://www.stanford.edu

1998-03-10  Mark Mitchell  <mmitchell@usa.net>

	* dbxout.c (dbxout_type_methods): Only treat TYPE_METHODS as a
	TREE_VEC if that's what it really is.

Index: dbxout.c
===================================================================
RCS file: /egcs/carton/cvsfiles/egcs/gcc/dbxout.c,v
retrieving revision 1.8
diff -c -p -r1.8 dbxout.c
*** dbxout.c	1998/02/07 22:54:16	1.8
--- dbxout.c	1998/03/11 00:33:43
*************** dbxout_type_methods (type)
*** 824,830 ****
  
    sprintf(formatted_type_identifier_length, "%d", type_identifier_length);
  
!   if (TREE_CODE (methods) == FUNCTION_DECL)
      fndecl = methods;
    else if (TREE_VEC_ELT (methods, 0) != NULL_TREE)
      fndecl = TREE_VEC_ELT (methods, 0);
--- 824,830 ----
  
    sprintf(formatted_type_identifier_length, "%d", type_identifier_length);
  
!   if (TREE_CODE (methods) != TREE_VEC)
      fndecl = methods;
    else if (TREE_VEC_ELT (methods, 0) != NULL_TREE)
      fndecl = TREE_VEC_ELT (methods, 0);

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: pre-release-g++: cold ICE
  1998-03-10 18:13 ` Mark Mitchell
@ 1998-03-11 16:23   ` Jim Wilson
  0 siblings, 0 replies; 3+ messages in thread
From: Jim Wilson @ 1998-03-11 16:23 UTC (permalink / raw)
  To: mmitchell; +Cc: Max Lawson, egcs, egcs-bugs

This is OK to check in.

Jim

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~1998-03-11 16:23 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-03-08 13:06 pre-release-g++: cold ICE Max Lawson
1998-03-10 18:13 ` Mark Mitchell
1998-03-11 16:23   ` Jim Wilson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).