From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-116986-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 11482 invoked by alias); 3 Nov 2004 22:24:14 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Received: (qmail 11466 invoked by uid 48); 3 Nov 2004 22:24:13 -0000
Date: Wed, 03 Nov 2004 22:24:00 -0000
From: "steven at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Message-ID: <20041103222411.18293.steven@gcc.gnu.org>
Reply-To: gcc-bugzilla@gcc.gnu.org
Subject: [Bug middle-end/18293] New: Redundant copy operation introduced by expand
X-Bugzilla-Reason: CC
X-SW-Source: 2004-11/txt/msg00395.txt.bz2
List-Id: <gcc-bugs.sourceware.org>

Given this piece of code, 
 
struct deferred_access; 
extern struct deferred_access *D25481; 
struct deferred_access * 
foo (void) 
{ 
  return (struct deferred_access *) ((long unsigned int) D25481 * 16); 
} 
 
 
which generates on i686 in .00.expand (-fdump-rtl-expand-details): 
 
 
;; return (struct deferred_access *) ((long unsigned int) (int) D25481 * 16) 
(insn 10 9 11 (set (reg:SI 60) 
        (mem/f/i:SI (symbol_ref:SI ("D25481") [flags 0x40] <var_decl 
0x401d489c D25481>) [2 D25481+0 S4 A32])) -1 (nil) 
    (nil)) 
 
(insn 11 10 12 (set (reg:SI 61) 
        (reg:SI 60)) -1 (nil) 
    (nil)) 
 
(insn 12 11 13 (parallel [ 
            (set (reg:SI 62) 
                (ashift:SI (reg:SI 61) 
                    (const_int 4 [0x4]))) 
            (clobber (reg:CC 17 flags)) 
        ]) -1 (nil) 
    (expr_list:REG_EQUAL (mult:SI (reg:SI 60) 
            (const_int 16 [0x10])) 
        (nil))) 
 
(insn 13 12 14 (set (reg:SI 58 [ <result> ]) 
        (reg:SI 62)) -1 (nil) 
    (nil)) 
 
 
 
Notice the unnecessary copy insn 11: 
(insn 11 10 12 (set (reg:SI 61) (reg:SI 60)) -1 (nil) (nil)) 
 
Is there some way, oh RTL gurus tell me!, to avoid introducing this 
garbage RTL that only introduces work for the RTL optimizers?

-- 
           Summary: Redundant copy operation introduced by expand
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: middle-end
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: steven at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org,sayle at gcc dot gnu dot
                    org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18293