From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753916Ab2AaKof (ORCPT ); Tue, 31 Jan 2012 05:44:35 -0500 Received: from mga14.intel.com ([143.182.124.37]:40596 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753793Ab2AaKoY (ORCPT ); Tue, 31 Jan 2012 05:44:24 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.71,315,1320652800"; d="scan'208";a="101476656" Date: Tue, 31 Jan 2012 18:34:16 +0800 From: Wu Fengguang To: Shaohua Li Cc: lkml , linux-mm , Andrew Morton , Jens Axboe , Herbert Poetzl , Eric Dumazet , Vivek Goyal Subject: Re: [PATCH] fix readahead pipeline break caused by block plug Message-ID: <20120131103416.GA1661@localhost> References: <1327996780.21268.42.camel@sli10-conroe> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1327996780.21268.42.camel@sli10-conroe> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I'd like to propose a sister patch on the write part. It may not be as easy to measure any performance impacts of it, but I'll try. --- Subject: remove plugging at buffered write time Date: Tue Jan 31 18:25:48 CST 2012 Buffered write(2) is not directly tied to IO, so no need to handle plug in generic_file_aio_write(). CC: Jens Axboe CC: Li Shaohua Signed-off-by: Wu Fengguang --- mm/filemap.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- linux-next.orig/mm/filemap.c 2012-01-31 18:23:52.000000000 +0800 +++ linux-next/mm/filemap.c 2012-01-31 18:25:38.000000000 +0800 @@ -2267,6 +2267,7 @@ generic_file_direct_write(struct kiocb * struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; + struct blk_plug plug; ssize_t written; size_t write_len; pgoff_t end; @@ -2301,7 +2302,9 @@ generic_file_direct_write(struct kiocb * } } + blk_start_plug(&plug); written = mapping->a_ops->direct_IO(WRITE, iocb, iov, pos, *nr_segs); + blk_finish_plug(&plug); /* * Finally, try again to invalidate clean pages which might have been @@ -2610,13 +2613,11 @@ ssize_t generic_file_aio_write(struct ki { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; - struct blk_plug plug; ssize_t ret; BUG_ON(iocb->ki_pos != pos); mutex_lock(&inode->i_mutex); - blk_start_plug(&plug); ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); mutex_unlock(&inode->i_mutex); @@ -2627,7 +2628,6 @@ ssize_t generic_file_aio_write(struct ki if (err < 0 && ret > 0) ret = err; } - blk_finish_plug(&plug); return ret; } EXPORT_SYMBOL(generic_file_aio_write);