From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mathieu Desnoyers Subject: Re: [PATCH v7 09/16] SUNRPC/cache: use new hashtable implementation Date: Mon, 29 Oct 2012 11:41:04 -0400 Message-ID: <20121029154104.GA18542@Krystal> References: <1351450948-15618-1-git-send-email-levinsasha928@gmail.com> <1351450948-15618-9-git-send-email-levinsasha928@gmail.com> <20121029124229.GC11733@Krystal> <20121029151343.GA17722@Krystal> <20121029151653.GC9502@fieldses.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20121029151653.GC9502-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dev-bounces-yBygre7rU0TnMu66kgdUjQ@public.gmane.org Errors-To: dev-bounces-yBygre7rU0TnMu66kgdUjQ@public.gmane.org To: "J. Bruce Fields" Cc: snitzer-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, neilb-l3A5Bk7waGM@public.gmane.org, fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, Trond.Myklebust-HgOvQuBEEgTQT0dZR+AlfA@public.gmane.org, paul.gortmaker-CWA4WttNNZF54TAoqtyWWQ@public.gmane.org, dm-devel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, agk-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, aarcange-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, rds-devel-N0ozoZBvEnrZJqsBc5GL+g@public.gmane.org, eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, venkat.x.venkatsubra-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org, ccaulfie-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, mingo-X9Un+BFzKDI@public.gmane.org, dev-yBygre7rU0TnMu66kgdUjQ@public.gmane.org, ericvh-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, josh-iaAMLnmF4UmaiuxdJuQwMA@public.gmane.org, rostedt-nx8X9YLhiw1AfugRpC6u6w@public.gmane.org, lw-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org, teigland-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, Sasha Levin , axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org, linux-nfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, edumazet-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org, netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, ejt-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org, tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org, akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, Linus Torvalds , davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org List-Id: dm-devel.ids * J. Bruce Fields (bfields-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org) wrote: > On Mon, Oct 29, 2012 at 11:13:43AM -0400, Mathieu Desnoyers wrote: > > * Linus Torvalds (torvalds-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org) wrote: > > > On Mon, Oct 29, 2012 at 5:42 AM, Mathieu Desnoyers > > > wrote: > > > > > > > > So defining e.g.: > > > > > > > > #include > > > > > > > > #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(BITS_PER_LONG)) > > > > > > > > would keep the intended behavior in all cases: use one page for the hash > > > > array. > > > > > > Well, since that wasn't true before either because of the long-time > > > bug you point out, clearly the page size isn't all that important. I > > > think it's more important to have small and simple code, and "9" is > > > certainly that, compared to playing ilog2 games with not-so-obvious > > > things. > > > > > > Because there's no reason to believe that '9' is in any way a worse > > > random number than something page-shift-related, is there? And getting > > > away from *previous* overly-complicated size calculations that had > > > been broken because they were too complicated and random, sounds like > > > a good idea. > > > > Good point. I agree that unless we really care about the precise number > > of TLB entries and cache lines used by this hash table, we might want to > > stay away from page-size and pointer-size based calculation. > > > > It might not hurt to explain this in the patch changelog though. > > I'd also be happy to take that as a separate patch now. FYIW: I've made a nice boo-boo above. It should have been: #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(sizeof(struct hlist_head))) Because we happen to have a memory indexed in bytes, not in bits. I guess this goes a long way proving Linus' point about virtues of trivial code. ;-) Thanks, Mathieu -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: linux-nfs-owner@vger.kernel.org Received: from mail.openrapids.net ([64.15.138.104]:58667 "EHLO blackscsi.openrapids.net" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1759364Ab2J2PlH (ORCPT ); Mon, 29 Oct 2012 11:41:07 -0400 Date: Mon, 29 Oct 2012 11:41:04 -0400 From: Mathieu Desnoyers To: "J. Bruce Fields" Cc: Linus Torvalds , Sasha Levin , tj@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, paul.gortmaker@windriver.com, davem@davemloft.net, rostedt@goodmis.org, mingo@elte.hu, ebiederm@xmission.com, aarcange@redhat.com, ericvh@gmail.com, netdev@vger.kernel.org, josh@joshtriplett.org, eric.dumazet@gmail.com, axboe@kernel.dk, agk@redhat.com, dm-devel@redhat.com, neilb@suse.de, ccaulfie@redhat.com, teigland@redhat.com, Trond.Myklebust@netapp.com, fweisbec@gmail.com, jesse@nicira.com, venkat.x.venkatsubra@oracle.com, ejt@redhat.com, snitzer@redhat.com, edumazet@google.com, linux-nfs@vger.kernel.org, dev@openvswitch.org, rds-devel@oss.oracle.com, lw@cn.fujitsu.com Subject: Re: [PATCH v7 09/16] SUNRPC/cache: use new hashtable implementation Message-ID: <20121029154104.GA18542@Krystal> References: <1351450948-15618-1-git-send-email-levinsasha928@gmail.com> <1351450948-15618-9-git-send-email-levinsasha928@gmail.com> <20121029124229.GC11733@Krystal> <20121029151343.GA17722@Krystal> <20121029151653.GC9502@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20121029151653.GC9502@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: * J. Bruce Fields (bfields@fieldses.org) wrote: > On Mon, Oct 29, 2012 at 11:13:43AM -0400, Mathieu Desnoyers wrote: > > * Linus Torvalds (torvalds@linux-foundation.org) wrote: > > > On Mon, Oct 29, 2012 at 5:42 AM, Mathieu Desnoyers > > > wrote: > > > > > > > > So defining e.g.: > > > > > > > > #include > > > > > > > > #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(BITS_PER_LONG)) > > > > > > > > would keep the intended behavior in all cases: use one page for the hash > > > > array. > > > > > > Well, since that wasn't true before either because of the long-time > > > bug you point out, clearly the page size isn't all that important. I > > > think it's more important to have small and simple code, and "9" is > > > certainly that, compared to playing ilog2 games with not-so-obvious > > > things. > > > > > > Because there's no reason to believe that '9' is in any way a worse > > > random number than something page-shift-related, is there? And getting > > > away from *previous* overly-complicated size calculations that had > > > been broken because they were too complicated and random, sounds like > > > a good idea. > > > > Good point. I agree that unless we really care about the precise number > > of TLB entries and cache lines used by this hash table, we might want to > > stay away from page-size and pointer-size based calculation. > > > > It might not hurt to explain this in the patch changelog though. > > I'd also be happy to take that as a separate patch now. FYIW: I've made a nice boo-boo above. It should have been: #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(sizeof(struct hlist_head))) Because we happen to have a memory indexed in bytes, not in bits. I guess this goes a long way proving Linus' point about virtues of trivial code. ;-) Thanks, Mathieu -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx164.postini.com [74.125.245.164]) by kanga.kvack.org (Postfix) with SMTP id EDBD76B0072 for ; Mon, 29 Oct 2012 11:41:06 -0400 (EDT) Date: Mon, 29 Oct 2012 11:41:04 -0400 From: Mathieu Desnoyers Subject: Re: [PATCH v7 09/16] SUNRPC/cache: use new hashtable implementation Message-ID: <20121029154104.GA18542@Krystal> References: <1351450948-15618-1-git-send-email-levinsasha928@gmail.com> <1351450948-15618-9-git-send-email-levinsasha928@gmail.com> <20121029124229.GC11733@Krystal> <20121029151343.GA17722@Krystal> <20121029151653.GC9502@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20121029151653.GC9502@fieldses.org> Sender: owner-linux-mm@kvack.org List-ID: To: "J. Bruce Fields" Cc: Linus Torvalds , Sasha Levin , tj@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, paul.gortmaker@windriver.com, davem@davemloft.net, rostedt@goodmis.org, mingo@elte.hu, ebiederm@xmission.com, aarcange@redhat.com, ericvh@gmail.com, netdev@vger.kernel.org, josh@joshtriplett.org, eric.dumazet@gmail.com, axboe@kernel.dk, agk@redhat.com, dm-devel@redhat.com, neilb@suse.de, ccaulfie@redhat.com, teigland@redhat.com, Trond.Myklebust@netapp.com, fweisbec@gmail.com, jesse@nicira.com, venkat.x.venkatsubra@oracle.com, ejt@redhat.com, snitzer@redhat.com, edumazet@google.com, linux-nfs@vger.kernel.org, dev@openvswitch.org, rds-devel@oss.oracle.com, lw@cn.fujitsu.com * J. Bruce Fields (bfields@fieldses.org) wrote: > On Mon, Oct 29, 2012 at 11:13:43AM -0400, Mathieu Desnoyers wrote: > > * Linus Torvalds (torvalds@linux-foundation.org) wrote: > > > On Mon, Oct 29, 2012 at 5:42 AM, Mathieu Desnoyers > > > wrote: > > > > > > > > So defining e.g.: > > > > > > > > #include > > > > > > > > #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(BITS_PER_LONG)) > > > > > > > > would keep the intended behavior in all cases: use one page for the hash > > > > array. > > > > > > Well, since that wasn't true before either because of the long-time > > > bug you point out, clearly the page size isn't all that important. I > > > think it's more important to have small and simple code, and "9" is > > > certainly that, compared to playing ilog2 games with not-so-obvious > > > things. > > > > > > Because there's no reason to believe that '9' is in any way a worse > > > random number than something page-shift-related, is there? And getting > > > away from *previous* overly-complicated size calculations that had > > > been broken because they were too complicated and random, sounds like > > > a good idea. > > > > Good point. I agree that unless we really care about the precise number > > of TLB entries and cache lines used by this hash table, we might want to > > stay away from page-size and pointer-size based calculation. > > > > It might not hurt to explain this in the patch changelog though. > > I'd also be happy to take that as a separate patch now. FYIW: I've made a nice boo-boo above. It should have been: #define DFR_HASH_BITS (PAGE_SHIFT - ilog2(sizeof(struct hlist_head))) Because we happen to have a memory indexed in bytes, not in bits. I guess this goes a long way proving Linus' point about virtues of trivial code. ;-) Thanks, Mathieu -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org