Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756594AbaAFWNE (ORCPT ); Mon, 6 Jan 2014 17:13:04 -0500 Received: from mail.linuxfoundation.org ([140.211.169.12]:46969 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755704AbaAFWNC (ORCPT ); Mon, 6 Jan 2014 17:13:02 -0500 Date: Mon, 6 Jan 2014 14:13:00 -0800 From: Andrew Morton To: Raghavendra K T Cc: Fengguang Wu , David Cohen , Al Viro , Damien Ramonda , jack@suse.cz, Linus , linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [RFC PATCH V3] mm readahead: Fix the readahead fail in case of empty numa node Message-Id: <20140106141300.4e1c950d45c614d6c29bdd8f@linux-foundation.org> In-Reply-To: <1389003715-29733-1-git-send-email-raghavendra.kt@linux.vnet.ibm.com> References: <1389003715-29733-1-git-send-email-raghavendra.kt@linux.vnet.ibm.com> X-Mailer: Sylpheed 3.2.0beta5 (GTK+ 2.24.10; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 6 Jan 2014 15:51:55 +0530 Raghavendra K T wrote: > Currently, max_sane_readahead returns zero on the cpu with empty numa node, > fix this by checking for potential empty numa node case during calculation. > We also limit the number of readahead pages to 4k. > > ... > > --- a/mm/readahead.c > +++ b/mm/readahead.c > @@ -237,14 +237,25 @@ int force_page_cache_readahead(struct address_space *mapping, struct file *filp, > return ret; > } > > +#define MAX_REMOTE_READAHEAD 4096UL > /* > * Given a desired number of PAGE_CACHE_SIZE readahead pages, return a > * sensible upper limit. > */ > unsigned long max_sane_readahead(unsigned long nr) > { > - return min(nr, (node_page_state(numa_node_id(), NR_INACTIVE_FILE) > - + node_page_state(numa_node_id(), NR_FREE_PAGES)) / 2); > + unsigned long local_free_page; > + unsigned long sane_nr = min(nr, MAX_REMOTE_READAHEAD); > + > + local_free_page = node_page_state(numa_node_id(), NR_INACTIVE_FILE) > + + node_page_state(numa_node_id(), NR_FREE_PAGES); > + > + /* > + * Readahead onto remote memory is better than no readahead when local > + * numa node does not have memory. We sanitize readahead size depending > + * on free memory in the local node but limiting to 4k pages. > + */ > + return local_free_page ? min(sane_nr, local_free_page / 2) : sane_nr; > } So if the local node has two free pages, we do just one page of readahead. Then the local node has one free page and we do zero pages readahead. Assuming that bug(!) is fixed, the local node now has zero free pages and we suddenly resume doing large readahead. This transition from large readahead to very small readahead then back to large readahead is illogical, surely? -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/