🚀 go-pugleaf

The problem of getting "failed assertion" in core.c is almost
certainly a GCC bug.  But I cannot debug it because I do not
understand how that code works.

I could study it and fix the problem, but it would take time.  Many of
you who are reading this could do that with no more time than I would
need.

To get the maximum progress on GNU with the resources available, I
think it is best if I work on other things and let this wait until one
of you decides to debug it.  Our rate of progress is measured by how
many people contribute.

Here is a patch I was sent to fix a bug in core.c.  I believe it does
not fix this problem, but you might as well have the latest version
to work from.

===================================================================
RCS file: /home/fsf/rms/cc/objc/RCS/core.c,v
retrieving revision 1.14
diff -a -c1 -r1.14 core.c
*** core.c	1993/01/11 05:10:46	1.14
--- core.c	1993/01/14 20:46:05
***************
*** 1172,1174 ****
  	 node = hash_next (module_hash_table, node))
!       record_store (NULL, record);
  			
--- 1172,1180 ----
  	 node = hash_next (module_hash_table, node))
!       {
! 	int nclasses = ((Module_t) node->value)->symtab->cls_def_cnt;
! 	int j;
! 
! 	for (j = 0; j < nclasses; j++)
! 	  record_store (NULL, record);
!       }
  			
***************
*** 1177,1179 ****
  	 node = hash_next (module_hash_table, node))
!       record_store (NULL, record);
    }	
--- 1183,1191 ----
  	 node = hash_next (module_hash_table, node))
!       {
! 	int nclasses = ((Module_t) node->value)->symtab->cls_def_cnt;
! 	int j;
! 
! 	for (j = 0; j < nclasses; j++)
! 	  record_store (NULL, record);
!       }
    }