From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756841AbcAZOJp (ORCPT ); Tue, 26 Jan 2016 09:09:45 -0500 Received: from outbound-smtp05.blacknight.com ([81.17.249.38]:32998 "EHLO outbound-smtp05.blacknight.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756331AbcAZOJi (ORCPT ); Tue, 26 Jan 2016 09:09:38 -0500 From: Mel Gorman To: Andrew Morton Cc: Hugh Dickins , Jan Kara , Linux-FSDevel , Linux-MM , LKML , Mel Gorman Subject: [PATCH 0/2] Avoid unnecessary page locks in the generic read path v2r1 Date: Tue, 26 Jan 2016 14:09:28 +0000 Message-Id: <1453817370-10399-1-git-send-email-mgorman@techsingularity.net> X-Mailer: git-send-email 2.6.4 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Changelog since V1 o Use lock_page_killable consistently do_generic_file_read (jan) A long time ago there was an attempt to merge a patch that reduced the cost of unlock_page by avoiding the page_waitqueue lookup if there were no waiters. It was rejected on the grounds of complexity but it was pointed out that the read paths call lock_page unnecessarily. This series reduces the number of calls to lock_page when multiple processes read data in at the same time. mm/filemap.c | 90 ++++++++++++++++++++++++++++++++++++++++-------------------- 1 file changed, 60 insertions(+), 30 deletions(-) -- 2.6.4