From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Lameter Subject: Re: [PATCH] alloc_percpu() fails to allocate percpu data Date: Wed, 27 Feb 2008 13:56:55 -0800 (PST) Message-ID: References: <47BDBC23.10605@cosmosbay.com> <20080227122451.6a3b5565.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Cc: dada1@cosmosbay.com, davem@davemloft.net, linux-kernel@vger.kernel.org, netdev@vger.kernel.org, yanmin_zhang@linux.intel.com, travis@sgi.com To: Andrew Morton Return-path: Received: from netops-testserver-3-out.sgi.com ([192.48.171.28]:49526 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1756897AbYB0V47 (ORCPT ); Wed, 27 Feb 2008 16:56:59 -0500 In-Reply-To: <20080227122451.6a3b5565.akpm@linux-foundation.org> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 27 Feb 2008, Andrew Morton wrote: > On Wed, 27 Feb 2008 11:59:32 -0800 (PST) > Christoph Lameter wrote: > > > Any decision made on what to do about this one? Mike or I can > > repost the per cpu allocator against mm? The fix by Eric could be used > > in the interim for 2.6.24? > > > > I suppose I'll merge Eric's patch when I've tested it fully (well, as fully > as I test stuff). Urgh. You too? > It'd be nice to get that cache_line_size()/L1_CACHE_BYTES/L1_CACHE_ALIGN() > mess sorted out. If it's a mess - I _think_ it is? Well I tried it when slub went first in and it did not go well. The issue is that x86 detects the cache line size on bootup. Thus cache_line_size(). Most of the other arch have compile time cache line sizes. Thus L1_CACHE_BYTES. So L1_CACHE_BYTES is the maximum value that cache_line_size() can take. What I was attempting to do is to make x86 have one compile time cache line size L1_CACHE_BYTES. That raised objections because space was wasted.