From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 14 Mar 2005 08:37:25 -0800 From: "Martin J. Bligh" Subject: Re: ia64 needs to shake memory from quicklists when there is memory pressure. Message-ID: <163690000.1110818244@[10.10.2.4]> In-Reply-To: <20050314162412.GA9117@lnx-holt.americas.sgi.com> References: <20050309170915.GA1583@lnx-holt.americas.sgi.com> <2120000.1110388509@[10.10.2.4]> <20050314162412.GA9117@lnx-holt.americas.sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: owner-linux-mm@kvack.org Return-Path: To: Robin Holt Cc: akpm@osdl.org, tony.luck@intel.com, linux-ia64@vger.kernel.org, linux-mm@kvack.org List-ID: >> > The "ideal" would be to have a node aware slab cache. Since that >> > is probably a long time coming, >> >> Manfred already did one. Perhaps we can get that going again? would >> be useful for more than just this ... > > Is this the kmem_cache_alloc_node() stuff? If so, when I use that Yes, I think so. > for page table allocations, running AIM7 on a 32 processor machine > consistently live-locks the system. Everybody is waiting inside > kmem_cache_alloc_node() cachep->spinlock. > > Is there a different patch I should be looking for? Don't know of one, but maybe we can get Manfred to fix the existing one ;-) M. -- 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: aart@kvack.org