From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Wed, 28 May 2008 11:13:35 +0200 From: Nick Piggin Subject: Re: [patch 03/23] hugetlb: modular state Message-ID: <20080528091335.GD2630@wotan.suse.de> References: <20080525142317.965503000@nick.local0.net> <20080525143452.408189000@nick.local0.net> <1211920687.12036.11.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1211920687.12036.11.camel@localhost.localdomain> Sender: owner-linux-mm@kvack.org Return-Path: To: Adam Litke Cc: linux-mm@kvack.org, kniht@us.ibm.com, andi@firstfloor.org, nacc@us.ibm.com, abh@cray.com, joachim.deguara@amd.com, Andi Kleen List-ID: On Tue, May 27, 2008 at 03:38:07PM -0500, Adam Litke wrote: > Phew. At last I made it to the end of this one :) It seems okay to me > though. Have you done any performance testing on this patch series yet? > I don't expect the hstate structure to introduce any measurable > performance degradation, but it would be nice to have some numbers to > back up that educated guess. Haven't seen any noticable performance differences, but I don't know that I'm doing particularly interesting testing. Would be nice to get some results with HPC or databases or something that actually test the code. I'd say with HUGE_MAX_HSTATE == 1, the compiler _should_ be able to constant fold much of it away. There would be a few more pointer dereferences (eg to get hstate from inode or vma)... if that _really_ matters, we could special case HUGE_MAX_HSTATE == 1 in some places to bring performance back up. > > Acked-by: Adam Litke > > On Mon, 2008-05-26 at 00:23 +1000, npiggin@suse.de wrote: > > plain text document attachment (hugetlb-modular-state.patch) > > Large, but rather mechanical patch that converts most of the hugetlb.c > > globals into structure members and passes them around. -- 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