65 GCC regressions, 46 new, with your patch on 2003-07-02T02:10:25Z.

Jan Hubicka jh@suse.cz
Wed Jul 2 12:14:00 GMT 2003


> 
> 
> All of these are have the following ICE (all at -O3):
> 
> /Users/regress/tbox/cvs-gcc/gcc/gcc/testsuite/gcc.c-torture/execute/ 
> 920501-7.c: In function `main':^M
> /Users/regress/tbox/cvs-gcc/gcc/gcc/testsuite/gcc.c-torture/execute/ 
> 920501-7.c:17: internal compiler error: in lookup_static_chain, at  
> function.c:5647^M
> Please submit a full bug report,^M
> with preprocessed source if appropriate.^M
> See <URL:http://gcc.gnu.org/bugs.html> for instructions.^M

I am looking into it, however I have no idea how to reproduce it as it
works well for me on both i386 and PPC boxes.  I am building the darwin
crosscompiler too.  Does these still reproduce with the current tree?
> 
> Or this error (all at -O3):
> 
> /Users/regress/tbox/cvs-gcc/gcc/gcc/testsuite/gcc.c-torture/compile/ 
> inline-1.c: In function `file_name_as_prefix':^M
> /Users/regress/tbox/cvs-gcc/gcc/gcc/testsuite/gcc.c-torture/compile/ 
> inline-1.c:10: error: `va_start' used in function with fixed args^M

I've fixed this one already.

Honza
> 
> Or is a failure on execution (all at -O3).
> 
> These look like they are caused by one of these two patches:
> +Wed Jul  2 02:16:48 CEST 2003  Jan Hubicka  <jh@suse.cz>
> +
> +	* cgraph.c (cgraph_mark_needed_node, cgraph_varpool_mark_needed_node,
> +	cgraph_varpool_finalize_decl, cgraph_varpool_assemble_pending_decls):
> +	Use next_needed field instead of aux to maintain the queue.
> +	* cgraph.h (cgraph_node): Add next_needed.
> +	(cgraph_varpool_node): Add next_needed; remove aux.
> +	* cgraphunit.c (cgraph_finalize_compilation_unit): Use next_needed.
> +
> +Wed Jul  2 02:12:51 CEST 2003  Jan Hubicka  <jh@suse.cz>
> +
> +	* cgraphunit.c (cgraph_finalize_function): Set finalized.
> +	(cgraph_finalize_function): Do not examine inlinablility.
> +	(cgraph_finalize_compilation_unit): Do it here.
> +	* cgraph.h (cgraph_local_info): Add finalized field.
> 



More information about the Gcc-regression mailing list