Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755576Ab0DCB3h (ORCPT ); Fri, 2 Apr 2010 21:29:37 -0400 Received: from sous-sol.org ([216.99.217.87]:45701 "EHLO x200.localdomain" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1754889Ab0DCB3A (ORCPT ); Fri, 2 Apr 2010 21:29:00 -0400 Message-Id: <20100403012814.543517900@sous-sol.org> User-Agent: quilt/0.47-1 Date: Fri, 02 Apr 2010 18:27:52 -0700 From: Chris Wright To: Joerg Roedel Cc: nhorman@tuxdriver.com, nhorman@redhat.com, vgoyal@redhat.com, hbabu@us.ibm.com, ebiederm@xmission.com, iommu@lists.linux-foundation.org, kexec@lists.infradead.org, linux-kernel@vger.kernel.org Subject: [PATCH 1/4] x86/amd-iommu: enable iommu before attaching devices References: <20100403012751.834020949@sous-sol.org> Content-Disposition: inline; filename=amd-enable-iommu-earlier.patch Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1609 Lines: 51 Hit another kdump problem as reported by Neil Horman. When initializaing the IOMMU, we attach devices to their domains before the IOMMU is fully (re)initialized. Attaching a device will issue some important invalidations. In the context of the newly kexec'd kdump kernel, the IOMMU may have stale cached data from the original kernel. Because we do the attach too early, the invalidation commands are placed in the new command buffer before the IOMMU is updated w/ that buffer. This leaves the stale entries in the kdump context and can renders device unusable. Simply enable the IOMMU before we do the attach. Cc: Neil Horman Cc: Vivek Goyal Signed-off-by: Chris Wright --- arch/x86/kernel/amd_iommu_init.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -1288,6 +1288,8 @@ static int __init amd_iommu_init(void) if (ret) goto free; + enable_iommus(); + if (iommu_pass_through) ret = amd_iommu_init_passthrough(); else @@ -1300,8 +1302,6 @@ static int __init amd_iommu_init(void) amd_iommu_init_notifier(); - enable_iommus(); - if (iommu_pass_through) goto out; @@ -1315,6 +1315,7 @@ out: return ret; free: + disable_iommus(); amd_iommu_uninit_devices(); -- 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/