Re: optimization/7145: [PATCH] g++ -O with structure initializer & return value optimization generates bad code

Re: optimization/7145: [PATCH] g++ -O with structure initializer & return value optimization generates bad code

[Jason Merrill]

The following reply was made to PR optimization/7145; it has been noted by GNATS.

From: Jason Merrill <jason@redhat.com>
To: "H. J. Lu" <hjl@lucon.org>
Cc: tlb@tlb.org, gcc-gnats@gcc.gnu.org, gcc@gcc.gnu.org
Subject: Re: optimization/7145: [PATCH] g++ -O with structure initializer &
 return value optimization generates bad code
Date: Fri, 05 Jul 2002 15:59:53 +0100

 --=-=-=
 
 Here's the patch I'm applying.
 
 2002-07-05  Jason Merrill  <jason@redhat.com>
 
 	* tree.c (cp_copy_res_decl_for_inlining): Also copy DECL_INITIAL.
 
 
 --=-=-=
 Content-Type: text/x-patch
 Content-Disposition: inline
 
 *** tree.c.~1~	Tue Jul  2 14:56:10 2002
 --- tree.c	Thu Jul  4 16:08:15 2002
 *************** cp_copy_res_decl_for_inlining (result, f
 *** 2292,2297 ****
 --- 2292,2298 ----
   	  DECL_SOURCE_FILE (var) = DECL_SOURCE_FILE (nrv);
   	  DECL_SOURCE_LINE (var) = DECL_SOURCE_LINE (nrv);
   	  DECL_ABSTRACT_ORIGIN (var) = DECL_ORIGIN (nrv);
 + 	  DECL_INITIAL (var) = DECL_INITIAL (nrv);
   	  splay_tree_insert (decl_map,
   			     (splay_tree_key) nrv,
   			     (splay_tree_value) var);
 
 --=-=-=--

Re: optimization/7145: [PATCH] g++ -O with structure initializer & return value optimization generates bad code

[H. J. Lu]

The following reply was made to PR optimization/7145; it has been noted by GNATS.

From: "H. J. Lu" <hjl@lucon.org>
To: Jason Merrill <jason@redhat.com>
Cc: tlb@tlb.org, gcc-gnats@gcc.gnu.org, gcc@gcc.gnu.org
Subject: Re: optimization/7145: [PATCH] g++ -O with structure initializer & return value optimization generates bad code
Date: Fri, 5 Jul 2002 08:01:37 -0700

 On Fri, Jul 05, 2002 at 03:59:53PM +0100, Jason Merrill wrote:
 > Here's the patch I'm applying.
 > 
 > 2002-07-05  Jason Merrill  <jason@redhat.com>
 > 
 > 	* tree.c (cp_copy_res_decl_for_inlining): Also copy DECL_INITIAL.
 > 
 
 Thanks. I assume you will also fix the same regression in gcc 3.1.
 
 
 H.J.