From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932818Ab0KBHbZ (ORCPT ); Tue, 2 Nov 2010 03:31:25 -0400 Received: from mga02.intel.com ([134.134.136.20]:51130 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932713Ab0KBHbQ (ORCPT ); Tue, 2 Nov 2010 03:31:16 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.58,279,1286175600"; d="scan'208";a="673319015" From: Sheng Yang Organization: Intel Opensource Technology Center To: Jan Kiszka Subject: Re: [PATCH] intel-iommu: Fix use after release during device attach Date: Tue, 2 Nov 2010 15:31:22 +0800 User-Agent: KMail/1.13.5 (Linux/2.6.35-22-generic; KDE/4.5.1; x86_64; ; ) Cc: Linux Kernel Mailing List , kvm , Avi Kivity , Marcelo Tosatti , iommu@lists.linux-foundation.org, David Woodhouse References: <4CCFB84F.6050102@web.de> In-Reply-To: <4CCFB84F.6050102@web.de> MIME-Version: 1.0 Content-Type: Text/Plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Message-Id: <201011021531.22886.sheng@linux.intel.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tuesday 02 November 2010 15:05:51 Jan Kiszka wrote: > From: Jan Kiszka > > Obtail the new pgd pointer before releasing the page containing this > value. > > Signed-off-by: Jan Kiszka > --- > > Who is taking care of this? The kvm tree? > > drivers/pci/intel-iommu.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c > index 4789f8e..35463dd 100644 > --- a/drivers/pci/intel-iommu.c > +++ b/drivers/pci/intel-iommu.c > @@ -3627,9 +3627,9 @@ static int intel_iommu_attach_device(struct > iommu_domain *domain, > > pte = dmar_domain->pgd; > if (dma_pte_present(pte)) { > - free_pgtable_page(dmar_domain->pgd); > dmar_domain->pgd = (struct dma_pte *) > phys_to_virt(dma_pte_addr(pte)); > + free_pgtable_page(pte); > } > dmar_domain->agaw--; > } Reviewed-by: Sheng Yang CC iommu mailing list and David. OK, Jan, I got your meaning now. And it's not the exactly swap. :) I think the old code is safe, seems it's broken(exposed) by: commit 1a8bd481bfba30515b54368d90a915db3faf302f Author: David Woodhouse Date: Tue Aug 10 01:38:53 2010 +0100 intel-iommu: Fix 32-bit build warning with __cmpxchg() drivers/pci/intel-iommu.c: In function 'dma_pte_addr': drivers/pci/intel-iommu.c:239: warning: passing argument 1 of '__cmpxchg64' from incompatible pointer typ It seems that __cmpxchg64() now cares about the type of its pointer argument, so give it a (uint64_t *) instead of a pointer to a structure which contains only that. Signed-off-by: David Woodhouse diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index c9171be..603cdc0 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -236,7 +236,7 @@ static inline u64 dma_pte_addr(struct dma_pte *pte) return pte->val & VTD_PAGE_MASK; #else /* Must have a full atomic 64-bit read */ - return __cmpxchg64(pte, 0ULL, 0ULL) & VTD_PAGE_MASK; + return __cmpxchg64(&pte->val, 0ULL, 0ULL) & VTD_PAGE_MASK; #endif } Seems here is the only affected code? -- regards Yang, Sheng