From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:42816) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1YVZtE-0003WF-0h for qemu-devel@nongnu.org; Wed, 11 Mar 2015 02:13:12 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1YVZt9-00031Y-0g for qemu-devel@nongnu.org; Wed, 11 Mar 2015 02:13:07 -0400 Received: from e23smtp07.au.ibm.com ([202.81.31.140]:58203) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1YVZt8-00031D-DT for qemu-devel@nongnu.org; Wed, 11 Mar 2015 02:13:02 -0400 Received: from /spool/local by e23smtp07.au.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 11 Mar 2015 16:12:58 +1000 From: Gavin Shan Date: Wed, 11 Mar 2015 17:11:52 +1100 Message-Id: <1426054314-19564-1-git-send-email-gwshan@linux.vnet.ibm.com> Subject: [Qemu-devel] [PATCH 1/3] VFIO: Clear stale MSIx table during EEH reset List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: aik@ozlabs.ru, agraf@suse.de, Gavin Shan , alex.williamson@redhat.com, qemu-ppc@nongnu.org, david@gibson.dropbear.id.au The PCI device MSIx table is cleaned out in hardware after EEH PE reset. However, we still hold the stale MSIx entries in QEMU, which should be cleared accordingly. Otherwise, we will run into another (recursive) EEH error and the PCI devices contained in the PE have to be offlined exceptionally. The patch clears stale MSIx table before EEH PE reset so that MSIx table could be restored properly after EEH PE reset. Signed-off-by: Gavin Shan --- hw/vfio/common.c | 6 +++++- hw/vfio/pci.c | 39 +++++++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio.h | 3 ++- 3 files changed, 46 insertions(+), 2 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 148eb53..e3833f4 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -949,8 +949,12 @@ int vfio_container_ioctl(AddressSpace *as, int32_t groupid, switch (req) { case VFIO_CHECK_EXTENSION: case VFIO_IOMMU_SPAPR_TCE_GET_INFO: - case VFIO_EEH_PE_OP: break; + case VFIO_EEH_PE_OP: + if (!vfio_container_eeh_event(as, groupid, param)) { + break; + } + /* fallthru */ default: /* Return an error on unknown requests */ error_report("vfio: unsupported ioctl %X", req); diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index 6b80539..8c4a8cb 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -3319,6 +3319,45 @@ static void vfio_unregister_req_notifier(VFIOPCIDevice *vdev) vdev->req_enabled = false; } +int vfio_container_eeh_event(AddressSpace *as, int32_t groupid, + struct vfio_eeh_pe_op *op) +{ + VFIOGroup *group; + VFIODevice *vbasedev; + VFIOPCIDevice *vdev; + + group = vfio_get_group(groupid, as); + if (!group) { + vfio_put_group(group); + error_report("vfio: group %d not found\n", groupid); + return -1; + } + + switch (op->op) { + case VFIO_EEH_PE_RESET_HOT: + case VFIO_EEH_PE_RESET_FUNDAMENTAL: + /* + * The MSIx table will be cleaned out by reset. We need + * disable it so that it can be reenabled properly. Also, + * the cached MSIx table should be cleared as it's not + * reflecting the contents in hardware. + */ + QLIST_FOREACH(vbasedev, &group->device_list, next) { + vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev); + if (msix_enabled(&vdev->pdev)) { + vfio_disable_msix(vdev); + } + + msix_reset(&vdev->pdev); + } + + break; + } + + vfio_put_group(group); + return 0; +} + static int vfio_initfn(PCIDevice *pdev) { VFIOPCIDevice *vdev = DO_UPCAST(VFIOPCIDevice, pdev, pdev); diff --git a/include/hw/vfio/vfio.h b/include/hw/vfio/vfio.h index 0b26cd8..99528a3 100644 --- a/include/hw/vfio/vfio.h +++ b/include/hw/vfio/vfio.h @@ -5,5 +5,6 @@ extern int vfio_container_ioctl(AddressSpace *as, int32_t groupid, int req, void *param); - +extern int vfio_container_eeh_event(AddressSpace *as, int32_t groupid, + struct vfio_eeh_pe_op *op); #endif -- 1.8.3.2