From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752103AbbIQPOE (ORCPT ); Thu, 17 Sep 2015 11:14:04 -0400 Received: from mail-pa0-f52.google.com ([209.85.220.52]:33444 "EHLO mail-pa0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751471AbbIQPOB (ORCPT ); Thu, 17 Sep 2015 11:14:01 -0400 From: Ming Lei To: Jens Axboe , linux-kernel@vger.kernel.org Cc: Ming Lei , Christoph Hellwig , Kent Overstreet , Ming Lin , Dongsu Park Subject: [PATCH] block: blk-merge: fast-clone bio when splitting rw bios Date: Thu, 17 Sep 2015 23:13:27 +0800 Message-Id: <1442502807-24377-1-git-send-email-ming.lei@canonical.com> X-Mailer: git-send-email 1.9.1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org biovecs has become immutable since v3.13, so it isn't necessary to allocate biovecs for the new cloned bios, then we can save one extra biovecs allocation/copy, and the allocation is often not fixed-length and a bit more expensive. For example, if the 'max_sectors_kb' of null blk's queue is set as 16(32 sectors) via sysfs just for making more splits, this patch can increase throught about ~70% in the sequential read test over null_blk(direct io, bs: 1M). Cc: Christoph Hellwig Cc: Kent Overstreet Cc: Ming Lin Cc: Dongsu Park Signed-off-by: Ming Lei --- block/blk-merge.c | 19 ++++--------------- 1 file changed, 4 insertions(+), 15 deletions(-) diff --git a/block/blk-merge.c b/block/blk-merge.c index d088cff..be2493f 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -66,16 +66,13 @@ static struct bio *blk_bio_segment_split(struct request_queue *q, struct bio *bio, struct bio_set *bs) { - struct bio *split; struct bio_vec bv, bvprv; struct bvec_iter iter; unsigned seg_size = 0, nsegs = 0, sectors = 0; int prev = 0; bio_for_each_segment(bv, bio, iter) { - sectors += bv.bv_len >> 9; - - if (sectors > queue_max_sectors(q)) + if (sectors + (bv.bv_len >> 9) > queue_max_sectors(q)) goto split; /* @@ -96,6 +93,7 @@ static struct bio *blk_bio_segment_split(struct request_queue *q, seg_size += bv.bv_len; bvprv = bv; prev = 1; + sectors += bv.bv_len >> 9; continue; } new_segment: @@ -106,21 +104,12 @@ new_segment: bvprv = bv; prev = 1; seg_size = bv.bv_len; + sectors += bv.bv_len >> 9; } return NULL; split: - split = bio_clone_bioset(bio, GFP_NOIO, bs); - - split->bi_iter.bi_size -= iter.bi_size; - bio->bi_iter = iter; - - if (bio_integrity(bio)) { - bio_integrity_advance(bio, split->bi_iter.bi_size); - bio_integrity_trim(split, 0, bio_sectors(split)); - } - - return split; + return bio_split(bio, sectors, GFP_NOIO, bs); } void blk_queue_split(struct request_queue *q, struct bio **bio, -- 1.9.1