Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756152AbZJVOCL (ORCPT ); Thu, 22 Oct 2009 10:02:11 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756092AbZJVOCL (ORCPT ); Thu, 22 Oct 2009 10:02:11 -0400 Received: from mga02.intel.com ([134.134.136.20]:29046 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756078AbZJVOCK (ORCPT ); Thu, 22 Oct 2009 10:02:10 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.44,605,1249282800"; d="scan'208";a="460216193" Subject: Re: [RFC Patch] use MTRR for write combining if PAT is not available From: Suresh Siddha Reply-To: Suresh Siddha To: "H. Peter Anvin" Cc: Thomas Schlichter , Jan Beulich , Arjan van de Ven , "dri-devel@lists.sourceforge.net" , Robert Hancock , Henrique de Moraes Holschuh , "jbarnes@virtuousgeek.org" , Jeremy Fitzhardinge , "linux-kernel@vger.kernel.org" , Ingo Molnar , Ingo Molnar , Thomas Gleixner , Thomas Hellstrom , Tejun Heo , "Pallipadi, Venkatesh" , "x86@kernel.org" , Yinghai Lu In-Reply-To: <4AE04C95.6070402@zytor.com> References: <1395763521@web.de> <4AE04C95.6070402@zytor.com> Content-Type: text/plain Organization: Intel Corp Date: Thu, 22 Oct 2009 06:26:53 -0700 Message-Id: <1256218013.2546.9.camel@sbs-t61> Mime-Version: 1.0 X-Mailer: Evolution 2.26.3 (2.26.3-1.fc11) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1682 Lines: 38 On Thu, 2009-10-22 at 05:14 -0700, H. Peter Anvin wrote: > On 10/22/2009 09:08 PM, Thomas Schlichter wrote: > >> > >> I have couple of issues with this patchset still. pci_mmap_page_range() > >> doesn't get called for each fork(). So, we won't be ref counting the > >> mtrr usage properly. > > > > When forking, what happens with the "struct file"? If it is being copied, then the > > processes share the same private data which would be freed during the first > > release(). I think this would be a problem whereever file-private data are used. > > > > So I think it must be shared between the forked processes and some reference > > counting must exist. This reference counting must ensure that release() is only > > called when all processes did close() their file. > > > > And in that case (shared "struct file", one single release() call in the end) this > > implementation should be completely safe... > > > > struct file is shared between forked processes. That is correct. But I am referring to the ref-count getting incremented in Thomas's patch only in the pci_mmap_page_range() which will be called only during first mmap. We need to keep track of the counts of later forks too. For PAT, we keep track of this ref counting in track_pfn_vma_copy(). We shouldn't use different tracking mechanisms for PAT and non-PAT. We should cleanly tap into track_pfn_vma_copy() or extend that to cover this case aswell. thanks, suresh -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/