From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qt1-f197.google.com (mail-qt1-f197.google.com [209.85.160.197]) by kanga.kvack.org (Postfix) with ESMTP id 751EC6B000A for ; Fri, 5 Oct 2018 10:55:52 -0400 (EDT) Received: by mail-qt1-f197.google.com with SMTP id s1-v6so12533947qte.19 for ; Fri, 05 Oct 2018 07:55:52 -0700 (PDT) Received: from mx1.redhat.com (mx1.redhat.com. [209.132.183.28]) by mx.google.com with ESMTPS id c24-v6si5746501qtc.275.2018.10.05.07.55.51 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 05 Oct 2018 07:55:51 -0700 (PDT) Subject: Re: [PATCH v4 0/3] fs/dcache: Track # of negative dentries References: <1536773742-32687-1-git-send-email-longman@redhat.com> From: Waiman Long Message-ID: <7054d7fe-5385-c743-b2ea-595e5988799f@redhat.com> Date: Fri, 5 Oct 2018 10:55:47 -0400 MIME-Version: 1.0 In-Reply-To: <1536773742-32687-1-git-send-email-longman@redhat.com> Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Content-Language: en-US Sender: owner-linux-mm@kvack.org List-ID: To: Alexander Viro , Jonathan Corbet , Andrew Morton Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, "Luis R. Rodriguez" , Kees Cook , Linus Torvalds , Jan Kara , "Paul E. McKenney" , Ingo Molnar , Miklos Szeredi , Matthew Wilcox , Larry Woodman , James Bottomley , "Wangkai (Kevin C)" , Michal Hocko On 09/12/2018 01:35 PM, Waiman Long wrote: > v3->v4: > - Drop patch 4 as it is just a minor optimization. > - Add a cc:stable tag to patch 1. > - Clean up some comments in patch 3. > > v2->v3: > - With confirmation that the dummy array in dentry_stat structure > was never a replacement of a previously used field, patch 3 is now > reverted back to use one of dummy field as the negative dentry count > instead of adding a new field. > > v1->v2: > - Clarify what the new nr_dentry_negative per-cpu counter is tracking > and open-code the increment and decrement as suggested by Dave Chinner. > - Append the new nr_dentry_negative count as the 7th element of dentry-state > instead of replacing one of the dummy entries. > - Remove patch "fs/dcache: Make negative dentries easier to be > reclaimed" for now as I need more time to think about what > to do with it. > - Add 2 more patches to address issues found while reviewing the > dentry code. > - Add another patch to change the conditional branch of > nr_dentry_negative accounting to conditional move so as to reduce > the performance impact of the accounting code. > > This patchset addresses 2 issues found in the dentry code and adds a > new nr_dentry_negative per-cpu counter to track the total number of > negative dentries in all the LRU lists. > > Patch 1 fixes a bug in the accounting of nr_dentry_unused in > shrink_dcache_sb(). > > Patch 2 removes the ____cacheline_aligned_in_smp tag from super_block > LRU lists. > > Patch 3 adds the new nr_dentry_negative per-cpu counter. > > Various filesystem related tests were run and no statistically > significant changes in performance outside of the possible noise range > was observed. > > Waiman Long (3): > fs/dcache: Fix incorrect nr_dentry_unused accounting in > shrink_dcache_sb() > fs: Don't need to put list_lru into its own cacheline > fs/dcache: Track & report number of negative dentries > > Documentation/sysctl/fs.txt | 26 ++++++++++++++++---------- > fs/dcache.c | 38 +++++++++++++++++++++++++++++++++----- > include/linux/dcache.h | 7 ++++--- > include/linux/fs.h | 9 +++++---- > 4 files changed, 58 insertions(+), 22 deletions(-) > Any comments on these patches. The first one actually is a bug fix. Cheers, Longman