From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <44641A72.8050801@yahoo.com.au> Date: Fri, 12 May 2006 15:17:38 +1000 From: Nick Piggin MIME-Version: 1.0 Subject: Re: [PATCH 0/2][RFC] New version of shared page tables References: <1146671004.24422.20.camel@wildcat.int.mccr.org> <57DF992082E5BD7D36C9D441@[10.1.1.4]> <445FA0CA.4010008@us.ibm.com> <44600F9B.1060207@yahoo.com.au> <446242CB.4090106@us.ibm.com> In-Reply-To: <446242CB.4090106@us.ibm.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Brian Twichell Cc: Hugh Dickins , Dave McCracken , Linux Memory Management , Linux Kernel List-ID: Brian Twichell wrote: > Nick Piggin wrote: >> Of course if it was free performance then we'd want it. The downsides >> are that it >> is a significant complexity for a pretty small (3%) performance gain >> for your apparent >> target workload, which is pretty uncommon among all Linux users. > > > Our performance data demonstrated that the potential gain for the > non-hugepage case is much higher than 3%. The point is, there are hugepages. They were a significant additional complexity but the concession was made because they did provide a large speedup for databases. > >> >> Ignoring the complexity, it is still not free. Sharing data across >> processes adds to >> synchronisation overhead and hurts scalability. Some of these page >> fault scalability >> scenarios have shown to be important enough that we have introduced >> complexity _there_. > > > True, but this needs to be balanced against the fact that pagetable > sharing will reduce the number of page faults when it is achieved. > Let's say you have N processes which touch all the pages in an M page > shared memory region. Without shared pagetables this requires N*M page > faults; if pagetable sharing is achieved, only M pagefaults are required. > >> >> And it seems customers running "out-of-the-box" settings really want >> to start using >> hugepages if they're interested in getting the most performance >> possible, no? > > > My perspective is that, once the customer is required to invoke "echo > XXX > /proc/sys/vm/nr_hugepages" they've left the "out-of-the-box" > domain, and entered the domain of hoping that the number of hugepages is > sufficient, because if it's not, they'll probably need to reboot, which > can be pretty inconvenient for a production transaction-processing > application. I think it is pretty easy to reserve hugepages at bootup. This is what a production transaction processing system will be doing, won't it? Especially if they're performance constrained and hugepages gives them a 30% performance boost. -- SUSE Labs, Novell Inc. Send instant messages to your online friends http://au.messenger.yahoo.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