From: Davidlohr Bueso <davidlohr.bueso@hp.com> To: Andrew Morton <akpm@linux-foundation.org> Cc: David Gibson <david@gibson.dropbear.id.au>, Hugh Dickins <hughd@google.com>, Rik van Riel <riel@redhat.com>, Michel Lespinasse <walken@google.com>, Mel Gorman <mgorman@suse.de>, Konstantin Khlebnikov <khlebnikov@openvz.org>, Michal Hocko <mhocko@suse.cz>, "AneeshKumarK.V" <aneesh.kumar@linux.vnet.ibm.com>, KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>, Hillf Danton <dhillf@gmail.com>, linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>, Anton Blanchard <anton@samba.org> Subject: Re: [PATCH] mm/hugetlb: per-vma instantiation mutexes Date: Mon, 15 Jul 2013 17:12:31 -0700 [thread overview] Message-ID: <1373933551.4622.12.camel@buesod1.americas.hpqcorp.net> (raw) In-Reply-To: <20130715160802.9d0cdc0ee012b5e119317a98@linux-foundation.org> On Mon, 2013-07-15 at 16:08 -0700, Andrew Morton wrote: > On Mon, 15 Jul 2013 17:24:32 +1000 David Gibson <david@gibson.dropbear.id.au> wrote: > > > I have previously proposed a correct method of improving scalability, > > although it doesn't eliminate the lock. That's to use a set of hashed > > mutexes. > > Yep - hashing the mutexes is an obvious and nicely localized way of > improving this. It's a tweak, not a design change. > > The changelog should describe the choice of the hash key with great > precision, please. It's important and is the first thing which > reviewers and readers will zoom in on. > > Should the individual mutexes be cacheline aligned? Depends on the > acquisition frequency, I guess. Please let's work through that. In my test cases, involving different RDBMS, I'm getting around 114k acquisitions. > > Let's not damage uniprocesor kernels too much. AFACIT the main offender > here is fault_mutex_hash(), which is the world's most obfuscated "return > 0;". I guess we could add an ifndef CONFIG_SMP check to the function and return 0 right away. That would eliminate any overhead in fault_mutex_hash(). > > > It wasn't merged before, but I don't recall the reasons > > why. So I've forward ported the patch (will send once everyone agrees that the matter is settled), including the changes Anton Blanchard added a exactly two years ago: https://lkml.org/lkml/2011/7/15/31 My tests show that the number of lock contentions drops from ~11k to around 500. So this approach alleviates a lot of the bottleneck. I've also ran it against libhugetlbfs without any regressions. Thanks, Davidlohr
WARNING: multiple messages have this Message-ID (diff)
From: Davidlohr Bueso <davidlohr.bueso@hp.com> To: Andrew Morton <akpm@linux-foundation.org> Cc: David Gibson <david@gibson.dropbear.id.au>, Hugh Dickins <hughd@google.com>, Rik van Riel <riel@redhat.com>, Michel Lespinasse <walken@google.com>, Mel Gorman <mgorman@suse.de>, Konstantin Khlebnikov <khlebnikov@openvz.org>, Michal Hocko <mhocko@suse.cz>, "AneeshKumarK.V" <aneesh.kumar@linux.vnet.ibm.com>, KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>, Hillf Danton <dhillf@gmail.com>, linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>, Anton Blanchard <anton@samba.org> Subject: Re: [PATCH] mm/hugetlb: per-vma instantiation mutexes Date: Mon, 15 Jul 2013 17:12:31 -0700 [thread overview] Message-ID: <1373933551.4622.12.camel@buesod1.americas.hpqcorp.net> (raw) In-Reply-To: <20130715160802.9d0cdc0ee012b5e119317a98@linux-foundation.org> On Mon, 2013-07-15 at 16:08 -0700, Andrew Morton wrote: > On Mon, 15 Jul 2013 17:24:32 +1000 David Gibson <david@gibson.dropbear.id.au> wrote: > > > I have previously proposed a correct method of improving scalability, > > although it doesn't eliminate the lock. That's to use a set of hashed > > mutexes. > > Yep - hashing the mutexes is an obvious and nicely localized way of > improving this. It's a tweak, not a design change. > > The changelog should describe the choice of the hash key with great > precision, please. It's important and is the first thing which > reviewers and readers will zoom in on. > > Should the individual mutexes be cacheline aligned? Depends on the > acquisition frequency, I guess. Please let's work through that. In my test cases, involving different RDBMS, I'm getting around 114k acquisitions. > > Let's not damage uniprocesor kernels too much. AFACIT the main offender > here is fault_mutex_hash(), which is the world's most obfuscated "return > 0;". I guess we could add an ifndef CONFIG_SMP check to the function and return 0 right away. That would eliminate any overhead in fault_mutex_hash(). > > > It wasn't merged before, but I don't recall the reasons > > why. So I've forward ported the patch (will send once everyone agrees that the matter is settled), including the changes Anton Blanchard added a exactly two years ago: https://lkml.org/lkml/2011/7/15/31 My tests show that the number of lock contentions drops from ~11k to around 500. So this approach alleviates a lot of the bottleneck. I've also ran it against libhugetlbfs without any regressions. Thanks, Davidlohr -- 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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2013-07-16 0:12 UTC|newest] Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top 2013-07-12 23:28 [PATCH] mm/hugetlb: per-vma instantiation mutexes Davidlohr Bueso 2013-07-12 23:28 ` Davidlohr Bueso 2013-07-13 0:54 ` Hugh Dickins 2013-07-13 0:54 ` Hugh Dickins 2013-07-15 3:16 ` Davidlohr Bueso 2013-07-15 3:16 ` Davidlohr Bueso 2013-07-15 7:24 ` David Gibson 2013-07-15 23:08 ` Andrew Morton 2013-07-15 23:08 ` Andrew Morton 2013-07-16 0:12 ` Davidlohr Bueso [this message] 2013-07-16 0:12 ` Davidlohr Bueso 2013-07-16 8:00 ` David Gibson 2013-07-17 19:50 ` [PATCH] hugepage: allow parallelization of the hugepage fault path Davidlohr Bueso 2013-07-17 19:50 ` Davidlohr Bueso 2013-07-18 8:42 ` Joonsoo Kim 2013-07-18 8:42 ` Joonsoo Kim 2013-07-19 7:14 ` David Gibson 2013-07-19 21:24 ` Davidlohr Bueso 2013-07-19 21:24 ` Davidlohr Bueso 2013-07-22 0:59 ` Joonsoo Kim 2013-07-22 0:59 ` Joonsoo Kim 2013-07-18 9:07 ` Joonsoo Kim 2013-07-18 9:07 ` Joonsoo Kim 2013-07-19 0:19 ` Davidlohr Bueso 2013-07-19 0:19 ` Davidlohr Bueso 2013-07-19 0:35 ` Davidlohr Bueso 2013-07-19 0:35 ` Davidlohr Bueso 2013-07-23 7:04 ` Hush Bensen 2013-07-23 7:04 ` Hush Bensen 2013-07-23 6:55 ` Hush Bensen 2013-07-23 6:55 ` Hush Bensen 2013-07-16 1:51 ` [PATCH] mm/hugetlb: per-vma instantiation mutexes Rik van Riel 2013-07-16 1:51 ` Rik van Riel 2013-07-16 5:34 ` Joonsoo Kim 2013-07-16 5:34 ` Joonsoo Kim 2013-07-16 10:01 ` David Gibson 2013-07-18 6:50 ` Joonsoo Kim 2013-07-18 6:50 ` Joonsoo Kim 2013-07-16 8:20 ` David Gibson 2013-07-15 4:18 ` Konstantin Khlebnikov 2013-07-15 4:18 ` Konstantin Khlebnikov
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=1373933551.4622.12.camel@buesod1.americas.hpqcorp.net \ --to=davidlohr.bueso@hp.com \ --cc=akpm@linux-foundation.org \ --cc=aneesh.kumar@linux.vnet.ibm.com \ --cc=anton@samba.org \ --cc=david@gibson.dropbear.id.au \ --cc=dhillf@gmail.com \ --cc=hughd@google.com \ --cc=kamezawa.hiroyu@jp.fujitsu.com \ --cc=khlebnikov@openvz.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=mgorman@suse.de \ --cc=mhocko@suse.cz \ --cc=riel@redhat.com \ --cc=walken@google.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.