From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from d01relay04.pok.ibm.com (d01relay04.pok.ibm.com [9.56.227.236]) by e5.ny.us.ibm.com (8.13.8/8.12.11) with ESMTP id k7U0Kto7024623 for ; Tue, 29 Aug 2006 20:20:55 -0400 Received: from d01av04.pok.ibm.com (d01av04.pok.ibm.com [9.56.224.64]) by d01relay04.pok.ibm.com (8.13.6/8.13.6/NCO v8.1.1) with ESMTP id k7U0Ktej247744 for ; Tue, 29 Aug 2006 20:20:55 -0400 Received: from d01av04.pok.ibm.com (loopback [127.0.0.1]) by d01av04.pok.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id k7U0KtfZ026709 for ; Tue, 29 Aug 2006 20:20:55 -0400 Date: Tue, 29 Aug 2006 17:21:10 -0700 From: Nishanth Aravamudan Subject: Re: libnuma interleaving oddness Message-ID: <20060830002110.GZ5195@us.ibm.com> References: <20060829231545.GY5195@us.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: owner-linux-mm@kvack.org Return-Path: To: Christoph Lameter Cc: ak@suse.de, linux-mm@kvack.org, linuxppc-dev@ozlabs.org, lnxninja@us.ibm.com, agl@us.ibm.com List-ID: On 29.08.2006 [16:57:35 -0700], Christoph Lameter wrote: > On Tue, 29 Aug 2006, Nishanth Aravamudan wrote: > > > I don't know if this is a libnuma bug (I extracted out the code from > > libnuma, it looked sane; and even reimplemented it in libhugetlbfs > > for testing purposes, but got the same results) or a NUMA kernel bug > > (mbind is some hairy code...) or a ppc64 bug or maybe not a bug at > > all. Regardless, I'm getting somewhat inconsistent behavior. I can > > provide more debugging output, or whatever is requested, but I > > wasn't sure what to include. I'm hoping someone has heard of or seen > > something similar? > > Are you setting the tasks allocation policy before the allocation or > do you set a vma based policy? The vma based policies will only work > for anonymous pages. The order is (with necessary params filled in): p = mmap( , newsize, RW, PRIVATE, unlinked_hugetlbfs_heap_fd, ); numa_interleave_memory(p, newsize); mlock(p, newsize); /* causes all the hugepages to be faulted in */ munlock(p,newsize); >>From what I gathered from the numa manpages, the interleave policy should take effect on the mlock, as that is "fault-time" in this context. We're forcing the fault, that is. Does that answer your question? Sorry if I'm unclear, I'm a bit of a newbie to the VM. Thanks, Nish -- Nishanth Aravamudan IBM Linux Technology Center -- 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