From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 16 May 2005 14:06:14 -0700 From: "Martin J. Bligh" Reply-To: "Martin J. Bligh" Subject: Re: [PATCH] Factor in buddy allocator alignment requirements in node memory alignment Message-ID: <737010000.1116277574@flay> In-Reply-To: References: <1116274451.1005.106.camel@localhost> <1116276439.1005.110.camel@localhost> 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: christoph , Dave Hansen Cc: linux-mm , shai@scalex86.org, Linux Kernel Mailing List List-ID: --On Monday, May 16, 2005 12:55:39 -0700 christoph wrote: > On Mon, 16 May 2005, Dave Hansen wrote: > >> > Because the buddy allocator is complaining about wrongly allocated zones! >> >> Just because it complains doesn't mean that anything is actually >> wrong :) >> >> Do you know which pieces of code actually break if the alignment doesn't >> meet what that warning says? > > I have seen nothing break but 4 MB allocations f.e. will not be allocated > on a 4MB boundary with a 2 MB zone alignment. The page allocator always > returnes properly aligned pages but 4MB allocations are an exception? > > Some present or future hardware device or some other code may find that > surprising and crash. Now that it's fixed it is meant to notice that the start of the zone is not aligned, and not key off that, but the aligment itself ... the start and end roundoff bits shouldn't throw the rest out of alignment, as long as we do the calculation sensibly for how we regroup buddies. 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