From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 8384 invoked by alias); 13 May 2011 08:29:39 -0000 Received: (qmail 8375 invoked by uid 22791); 13 May 2011 08:29:38 -0000 X-SWARE-Spam-Status: No, hits=-5.8 required=5.0 tests=AWL,BAYES_00,RCVD_IN_DNSWL_HI,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from cantor.suse.de (HELO mx1.suse.de) (195.135.220.2) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Fri, 13 May 2011 08:29:23 +0000 Received: from relay1.suse.de (charybdis-ext.suse.de [195.135.221.2]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.suse.de (Postfix) with ESMTP id C950393987 for ; Fri, 13 May 2011 10:29:21 +0200 (CEST) Date: Fri, 13 May 2011 09:51:00 -0000 From: Richard Guenther To: gcc-patches@gcc.gnu.org Subject: [PATCH] Optimize fix for PR48172 Message-ID: User-Agent: Alpine 2.00 (LNX 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org X-SW-Source: 2011-05/txt/msg00946.txt.bz2 The following avoids the multiplication by niters when the step is equal. Boostrapped and tested on x86_64-unknown-linux-gnu, applied to trunk. Richard. 2011-05-13 Richard Guenther PR tree-optimization/48172 * tree-vect-loop-manip.c (vect_vfa_segment_size): Avoid multiplying by number of iterations for equal step. (vect_create_cond_for_alias_checks): Likewise. Index: gcc/tree-vect-loop-manip.c =================================================================== *** gcc/tree-vect-loop-manip.c (revision 173704) --- gcc/tree-vect-loop-manip.c (working copy) *************** vect_create_cond_for_align_checks (loop_ *** 2353,2375 **** Input: DR: The data reference. ! VECT_FACTOR: vectorization factor. ! SCALAR_LOOP_NITERS: number of iterations. Return an expression whose value is the size of segment which will be accessed by DR. */ static tree ! vect_vfa_segment_size (struct data_reference *dr, int vect_factor, ! tree scalar_loop_niters) { tree segment_length; segment_length = size_binop (MULT_EXPR, fold_convert (sizetype, DR_STEP (dr)), ! size_int (vect_factor)); ! segment_length = size_binop (MULT_EXPR, ! segment_length, ! fold_convert (sizetype, scalar_loop_niters)); if (vect_supportable_dr_alignment (dr, false) == dr_explicit_realign_optimized) { --- 2353,2370 ---- Input: DR: The data reference. ! LENGTH_FACTOR: segment length to consider. Return an expression whose value is the size of segment which will be accessed by DR. */ static tree ! vect_vfa_segment_size (struct data_reference *dr, tree length_factor) { tree segment_length; segment_length = size_binop (MULT_EXPR, fold_convert (sizetype, DR_STEP (dr)), ! fold_convert (sizetype, length_factor)); if (vect_supportable_dr_alignment (dr, false) == dr_explicit_realign_optimized) { *************** vect_create_cond_for_alias_checks (loop_ *** 2417,2423 **** ddr_p ddr; unsigned int i; ! tree part_cond_expr; /* Create expression ((store_ptr_0 + store_segment_length_0) < load_ptr_0) --- 2412,2418 ---- ddr_p ddr; unsigned int i; ! tree part_cond_expr, length_factor; /* Create expression ((store_ptr_0 + store_segment_length_0) < load_ptr_0) *************** vect_create_cond_for_alias_checks (loop_ *** 2465,2474 **** vect_create_addr_base_for_vector_ref (stmt_b, cond_expr_stmt_list, NULL_TREE, loop); ! segment_length_a = vect_vfa_segment_size (dr_a, vect_factor, ! scalar_loop_iters); ! segment_length_b = vect_vfa_segment_size (dr_b, vect_factor, ! scalar_loop_iters); if (vect_print_dump_info (REPORT_DR_DETAILS)) { --- 2460,2471 ---- vect_create_addr_base_for_vector_ref (stmt_b, cond_expr_stmt_list, NULL_TREE, loop); ! if (!operand_equal_p (DR_STEP (dr_a), DR_STEP (dr_b), 0)) ! length_factor = scalar_loop_iters; ! else ! length_factor = size_int (vect_factor); ! segment_length_a = vect_vfa_segment_size (dr_a, length_factor); ! segment_length_b = vect_vfa_segment_size (dr_b, length_factor); if (vect_print_dump_info (REPORT_DR_DETAILS)) {