From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-9.6 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 111CBC33CA9 for ; Mon, 13 Jan 2020 15:38:21 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id C06FE2081E for ; Mon, 13 Jan 2020 15:38:20 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (2048-bit key) header.d=infradead.org header.i=@infradead.org header.b="h5nnknDp" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org C06FE2081E Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=infradead.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 4559F8E0017; Mon, 13 Jan 2020 10:37:57 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 3E29C8E0012; Mon, 13 Jan 2020 10:37:57 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 215318E0018; Mon, 13 Jan 2020 10:37:57 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0197.hostedemail.com [216.40.44.197]) by kanga.kvack.org (Postfix) with ESMTP id DCDA38E0017 for ; Mon, 13 Jan 2020 10:37:56 -0500 (EST) Received: from smtpin23.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay05.hostedemail.com (Postfix) with SMTP id 950E8181AC9CC for ; Mon, 13 Jan 2020 15:37:56 +0000 (UTC) X-FDA: 76373016552.23.step10_7f812e56c3d55 X-HE-Tag: step10_7f812e56c3d55 X-Filterd-Recvd-Size: 6665 Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) by imf50.hostedemail.com (Postfix) with ESMTP for ; Mon, 13 Jan 2020 15:37:56 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=infradead.org; s=bombadil.20170209; h=Content-Transfer-Encoding: MIME-Version:References:In-Reply-To:Message-Id:Date:Subject:Cc:To:From:Sender :Reply-To:Content-Type:Content-ID:Content-Description:Resent-Date:Resent-From :Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Id:List-Help: List-Unsubscribe:List-Subscribe:List-Post:List-Owner:List-Archive; bh=dOAadIMTUupd20Eg2Q5RqFxOIthPxhUDYq6Q/juU65U=; b=h5nnknDpxtg47EUFlq9egYwY2Q BVp+XqvUz3RyOjCxys3SgxE+ritoHjVvgyDZ36NkMrrrwfr2kC2knN1NP0CoGkVdvtuFhj2Yvf4tA VjSYy3z+eL3rL11kVEmD7VKk1EGwVF9GyZ7wc8sscVYOXXEURYqq314o60EJJWx/6ToUW5CoKJjAn DpuvjUPQ5hkFpWzMcnaTNX+I0A70aAjno35EMwqvH8u+pFuUGSYiQjJrf8chJCmwN5w6GYyvu31GU pa2Kqn/JHsmhmGP6MdHRRtR0ZidTpsH5FAet4RtCa4BKvSNH4Xi2reB0reKS+G7/yxIkwXXkq0YYg vYkIgTuw==; Received: from willy by bombadil.infradead.org with local (Exim 4.92.3 #3 (Red Hat Linux)) id 1ir1mr-00075z-5B; Mon, 13 Jan 2020 15:37:53 +0000 From: Matthew Wilcox To: linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org Cc: "Matthew Wilcox (Oracle)" , jlayton@kernel.org, hch@infradead.org Subject: [PATCH 3/8] mm: Use a pagevec for readahead Date: Mon, 13 Jan 2020 07:37:41 -0800 Message-Id: <20200113153746.26654-4-willy@infradead.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20200113153746.26654-1-willy@infradead.org> References: <20200113153746.26654-1-willy@infradead.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: From: "Matthew Wilcox (Oracle)" Instead of using a linked list, use a small array. This does mean we will allocate and then submit for I/O no more than 15 pages at a time (60kB), but we have the block queue plugged so the bios can be combined afterwards. We generally don't readahead more than 256kB anyway, so this is not a huge reduction in efficiency, and we'll make up for it with later patches. Signed-off-by: Matthew Wilcox (Oracle) --- mm/readahead.c | 97 +++++++++++++++++++++++++++----------------------- 1 file changed, 52 insertions(+), 45 deletions(-) diff --git a/mm/readahead.c b/mm/readahead.c index 6bf73ef33b7e..76a70a4406b5 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -113,35 +113,37 @@ int read_cache_pages(struct address_space *mapping,= struct list_head *pages, =20 EXPORT_SYMBOL(read_cache_pages); =20 -static int read_pages(struct address_space *mapping, struct file *filp, - struct list_head *pages, unsigned int nr_pages, gfp_t gfp) +/* + * We ignore I/O errors - they will be handled by the actual consumer of + * the data that we attempted to prefetch. + */ +static unsigned read_pages(struct address_space *mapping, struct file *f= ilp, + struct pagevec *pvec, pgoff_t offset, gfp_t gfp) { - struct blk_plug plug; - unsigned page_idx; - int ret; - - blk_start_plug(&plug); + struct page *page; + unsigned int nr_pages =3D pagevec_count(pvec); =20 if (mapping->a_ops->readpages) { - ret =3D mapping->a_ops->readpages(filp, mapping, pages, nr_pages); - /* Clean up the remaining pages */ - put_pages_list(pages); - goto out; - } + LIST_HEAD(pages); =20 - for (page_idx =3D 0; page_idx < nr_pages; page_idx++) { - struct page *page =3D lru_to_page(pages); - list_del(&page->lru); - if (!add_to_page_cache_lru(page, mapping, page->index, gfp)) - mapping->a_ops->readpage(filp, page); - put_page(page); + pagevec_for_each(pvec, page) { + page->index =3D offset++; + list_add(&page->lru, &pages); + } + mapping->a_ops->readpages(filp, mapping, &pages, nr_pages); + /* Clean up the remaining pages */ + put_pages_list(&pages); + } else { + pagevec_for_each(pvec, page) { + if (!add_to_page_cache_lru(page, mapping, offset++, + gfp)) + mapping->a_ops->readpage(filp, page); + put_page(page); + } } - ret =3D 0; =20 -out: - blk_finish_plug(&plug); - - return ret; + pagevec_reinit(pvec); + return nr_pages; } =20 /* @@ -159,59 +161,64 @@ unsigned long __do_page_cache_readahead(struct addr= ess_space *mapping, struct inode *inode =3D mapping->host; struct page *page; unsigned long end_index; /* The last page we want to read */ - LIST_HEAD(page_pool); + struct pagevec pages; int page_idx; + pgoff_t page_offset =3D offset; unsigned long nr_pages =3D 0; loff_t isize =3D i_size_read(inode); gfp_t gfp_mask =3D readahead_gfp_mask(mapping); + struct blk_plug plug; + + blk_start_plug(&plug); =20 if (isize =3D=3D 0) goto out; =20 end_index =3D ((isize - 1) >> PAGE_SHIFT); + pagevec_init(&pages); =20 /* * Preallocate as many pages as we will need. */ for (page_idx =3D 0; page_idx < nr_to_read; page_idx++) { - pgoff_t page_offset =3D offset + page_idx; + page_offset++; =20 if (page_offset > end_index) break; =20 page =3D xa_load(&mapping->i_pages, page_offset); + + /* + * Page already present? Kick off the current batch of + * contiguous pages before continuing with the next batch. + */ if (page && !xa_is_value(page)) { - /* - * Page already present? Kick off the current batch of - * contiguous pages before continuing with the next - * batch. - */ - if (nr_pages) - read_pages(mapping, filp, &page_pool, nr_pages, - gfp_mask); - nr_pages =3D 0; + unsigned int count =3D pagevec_count(&pages); + + if (count) + nr_pages +=3D read_pages(mapping, filp, &pages, + offset, gfp_mask); + offset =3D page_offset + 1; continue; } =20 page =3D __page_cache_alloc(gfp_mask); if (!page) break; - page->index =3D page_offset; - list_add(&page->lru, &page_pool); + if (pagevec_add(&pages, page) =3D=3D 0) { + nr_pages +=3D read_pages(mapping, filp, &pages, + offset, gfp_mask); + offset =3D page_offset + 1; + } if (page_idx =3D=3D nr_to_read - lookahead_size) SetPageReadahead(page); - nr_pages++; } =20 - /* - * Now start the IO. We ignore I/O errors - if the page is not - * uptodate then the caller will launch readpage again, and - * will then handle the error. - */ - if (nr_pages) - read_pages(mapping, filp, &page_pool, nr_pages, gfp_mask); - BUG_ON(!list_empty(&page_pool)); + if (pagevec_count(&pages)) + nr_pages +=3D read_pages(mapping, filp, &pages, offset, gfp_mask); out: + blk_finish_plug(&plug); + return nr_pages; } =20 --=20 2.24.1