From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chao Yu Subject: [PATCH 4/5] f2fs: relocate readahead codes in readdir() Date: Fri, 13 Oct 2017 18:01:35 +0800 Message-ID: <20171013100136.35212-4-yuchao0@huawei.com> References: <20171013100136.35212-1-yuchao0@huawei.com> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: <20171013100136.35212-1-yuchao0@huawei.com> Sender: linux-kernel-owner@vger.kernel.org To: jaegeuk@kernel.org Cc: linux-f2fs-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org, chao@kernel.org, Chao Yu List-Id: linux-f2fs-devel.lists.sourceforge.net Previously, for large directory, we just do readahead only once in readdir(), readdir()'s performance may drop when traversing latter blocks. In order to avoid this, relocate readahead codes to covering all traverse flow. Signed-off-by: Chao Yu --- fs/f2fs/dir.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 1ebd206a4085..14646440b662 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -870,11 +870,6 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx) goto out_free; } - /* readahead for multi pages of dir */ - if (npages - n > 1 && !ra_has_index(ra, n)) - page_cache_sync_readahead(inode->i_mapping, ra, file, n, - min(npages - n, (pgoff_t)MAX_DIR_RA_PAGES)); - for (; n < npages; n++) { /* allow readdir() to be interrupted */ @@ -884,6 +879,11 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx) } cond_resched(); + /* readahead for multi pages of dir */ + if (npages - n > 1 && !ra_has_index(ra, n)) + page_cache_sync_readahead(inode->i_mapping, ra, file, n, + min(npages - n, (pgoff_t)MAX_DIR_RA_PAGES)); + dentry_page = get_lock_data_page(inode, n, false); if (IS_ERR(dentry_page)) { err = PTR_ERR(dentry_page); -- 2.13.1.388.g69e6b9b4f4a9