From patchwork Thu Jul 12 02:13:51 2012 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: liubo X-Patchwork-Id: 1186441 Return-Path: X-Original-To: patchwork-linux-btrfs@patchwork.kernel.org Delivered-To: patchwork-process-083081@patchwork2.kernel.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by patchwork2.kernel.org (Postfix) with ESMTP id 45C04DF25A for ; Thu, 12 Jul 2012 02:41:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756733Ab2GLClN (ORCPT ); Wed, 11 Jul 2012 22:41:13 -0400 Received: from cn.fujitsu.com ([222.73.24.84]:11385 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1756504Ab2GLClL (ORCPT ); Wed, 11 Jul 2012 22:41:11 -0400 X-IronPort-AV: E=Sophos;i="4.77,572,1336320000"; d="scan'208";a="5377531" Received: from unknown (HELO tang.cn.fujitsu.com) ([10.167.250.3]) by song.cn.fujitsu.com with ESMTP; 12 Jul 2012 10:40:09 +0800 Received: from fnstmail02.fnst.cn.fujitsu.com (tang.cn.fujitsu.com [127.0.0.1]) by tang.cn.fujitsu.com (8.14.3/8.13.1) with ESMTP id q6C22me7021974; Thu, 12 Jul 2012 10:03:00 +0800 Received: from localhost.localdomain ([10.167.225.27]) by fnstmail02.fnst.cn.fujitsu.com (Lotus Domino Release 8.5.3) with ESMTP id 2012071210024083-659318 ; Thu, 12 Jul 2012 10:02:40 +0800 From: Liu Bo To: Cc: Subject: [PATCH v2] Btrfs: improve multi-thread buffer read Date: Thu, 12 Jul 2012 10:13:51 +0800 Message-Id: <1342059231-20301-1-git-send-email-liubo2009@cn.fujitsu.com> X-Mailer: git-send-email 1.6.5.2 X-MIMETrack: Itemize by SMTP Server on mailserver/fnst(Release 8.5.3|September 15, 2011) at 2012/07/12 10:02:40, Serialize by Router on mailserver/fnst(Release 8.5.3|September 15, 2011) at 2012/07/12 10:02:56, Serialize complete at 2012/07/12 10:02:56 Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org While testing with my buffer read fio jobs[1], I find that btrfs does not perform well enough. Here is a scenario in fio jobs: We have 4 threads, "t1 t2 t3 t4", starting to buffer read a same file, and all of them will race on add_to_page_cache_lru(), and if one thread successfully puts its page into the page cache, it takes the responsibility to read the page's data. And what's more, reading a page needs a period of time to finish, in which other threads can slide in and process rest pages: t1 t2 t3 t4 add Page1 read Page1 add Page2 | read Page2 add Page3 | | read Page3 add Page4 | | | read Page4 -----|------------|-----------|-----------|-------- v v v v bio bio bio bio Now we have four bios, each of which holds only one page since we need to maintain consecutive pages in bio. Thus, we can end up with far more bios than we need. Here we're going to a) delay the real read-page section and b) try to put more pages into page cache. With that said, we can make each bio hold more pages and reduce the number of bios we need. Here is some numbers taken from fio results: w/o patch w patch ------------- -------- --------------- READ: 745MB/s +32% 987MB/s [1]: [global] group_reporting thread numjobs=4 bs=32k rw=read ioengine=sync directory=/mnt/btrfs/ [READ] filename=foobar size=2000M invalidate=1 Signed-off-by: Liu Bo --- v1->v2: if we fail to make a allocation, just fall back to the old way to read page. fs/btrfs/extent_io.c | 41 +++++++++++++++++++++++++++++++++++++++-- 1 files changed, 39 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 01c21b6..5c8ab6c 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3549,6 +3549,11 @@ int extent_writepages(struct extent_io_tree *tree, return ret; } +struct pagelst { + struct page *page; + struct list_head lst; +}; + int extent_readpages(struct extent_io_tree *tree, struct address_space *mapping, struct list_head *pages, unsigned nr_pages, @@ -3557,19 +3562,51 @@ int extent_readpages(struct extent_io_tree *tree, struct bio *bio = NULL; unsigned page_idx; unsigned long bio_flags = 0; + LIST_HEAD(page_pool); + struct pagelst *pagelst = NULL; for (page_idx = 0; page_idx < nr_pages; page_idx++) { struct page *page = list_entry(pages->prev, struct page, lru); + bool delay_read = true; prefetchw(&page->flags); list_del(&page->lru); + + if (!pagelst) + pagelst = kmalloc(sizeof(*pagelst), GFP_NOFS); + if (!pagelst) + delay_read = false; + if (!add_to_page_cache_lru(page, mapping, page->index, GFP_NOFS)) { - __extent_read_full_page(tree, page, get_extent, - &bio, 0, &bio_flags); + if (delay_read) { + pagelst->page = page; + list_add(&pagelst->lst, &page_pool); + page_cache_get(page); + pagelst = NULL; + } else { + __extent_read_full_page(tree, page, get_extent, + &bio, 0, &bio_flags); + } } page_cache_release(page); } + + while (!list_empty(&page_pool)) { + struct page *page; + + pagelst = list_entry(page_pool.prev, struct pagelst, lst); + page = pagelst->page; + + prefetchw(&page->flags); + __extent_read_full_page(tree, page, get_extent, + &bio, 0, &bio_flags); + + page_cache_release(page); + list_del(&pagelst->lst); + kfree(pagelst); + } + BUG_ON(!list_empty(&page_pool)); BUG_ON(!list_empty(pages)); if (bio) return submit_one_bio(READ, bio, 0, bio_flags);