From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: [patch 3/5] slub: Support 4k kmallocs again to compensate for page allocator slowness In-Reply-To: <20080214040313.855530089@sgi.com> Message-ID: From: "Pekka Enberg" MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT Date: Thu, 14 Feb 2008 09:01:49 +0200 (EET) Sender: owner-linux-mm@kvack.org Return-Path: To: clameter@sgi.com Cc: "linux-mm@kvack.org" List-ID: On 2/14/2008, "Christoph Lameter" wrote: > Currently we hand off PAGE_SIZEd kmallocs to the page allocator in the > mistaken belief that the page allocator can handle these allocations > effectively. However, measurements indicate a mininum slowdown by the > factor of 8 (and that is only SMP, NUMA is much worse) vs the slub > fastpath which causes regressions in tbench. Reviewed-by: Pekka Enberg -- 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