From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756626AbaEIJSc (ORCPT ); Fri, 9 May 2014 05:18:32 -0400 Received: from mail-lb0-f177.google.com ([209.85.217.177]:63542 "EHLO mail-lb0-f177.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754042AbaEIJS2 (ORCPT ); Fri, 9 May 2014 05:18:28 -0400 From: =?UTF-8?q?Matias=20Bj=C3=B8rling?= To: axboe@kernel.dk, sbradshaw@micron.com Cc: linux-kernel@vger.kernel.org, =?UTF-8?q?Matias=20Bj=C3=B8rling?= Subject: [PATCH] block: small performance optimization Date: Fri, 9 May 2014 11:17:40 +0200 Message-Id: <1399627061-5960-1-git-send-email-m@bjorling.me> X-Mailer: git-send-email 1.9.1 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Jens, The IO accounting stats contribute significant overhead on high IO/multi-node workloads. By moving the in_flight IO tracking to per-cpu counters, we can elevate some of the cache coherence overhead, at the cost of setup/teardown and percpu counters structure within the hd_struct structure. Sam Bradshaw reported an IOPS increase of 20% on his fio workloads. Thanks, Matias Matias Bjørling (1): blk: per-cpu counters for in-flight IO accounting block/partition-generic.c | 4 ++-- include/linux/genhd.h | 25 ++++++++++++++++++------- 2 files changed, 20 insertions(+), 9 deletions(-) -- 1.9.1