From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751710Ab2DQSaq (ORCPT ); Tue, 17 Apr 2012 14:30:46 -0400 Received: from smtp.snhosting.dk ([87.238.248.203]:56080 "EHLO smtp.domainteam.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750725Ab2DQSap (ORCPT ); Tue, 17 Apr 2012 14:30:45 -0400 Date: Tue, 17 Apr 2012 20:30:42 +0200 From: Sam Ravnborg To: Yinghai Lu , Tejun Heo Cc: Michal Hocko , linux-mm@kvack.org, LKML Subject: Re: Weirdness in __alloc_bootmem_node_high Message-ID: <20120417183042.GA21051@merkur.ravnborg.org> References: <20120417155502.GE22687@tiehlicka.suse.cz> <20120417173203.GA32482@tiehlicka.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Apr 17, 2012 at 11:07:10AM -0700, Yinghai Lu wrote: > On Tue, Apr 17, 2012 at 10:32 AM, Michal Hocko wrote: > > On Tue 17-04-12 10:12:30, Yinghai Lu wrote: > >> > >> We are not using bootmem with x86 now, so could remove those workaround now. > > > > Could you be more specific about what the workaround is used for? > > Don't bootmem allocating too low to use up all low memory. like for > system with lots of memory for sparse vmemmap. > > when nobootmem.c is used, __alloc_bootmem_node_high is the same as > __alloc_bootmem_node. It would be nice if someone familiar with the memblock/bootmem internals could cleans up the leftovers from the migration of x86 to memblock / nobootmem. This would be less to be confused about when other migrate to use memblock. Sam