Skip to contentSkip to conversation

Re-evaluate RMRR info for devices removed from si domain

PATCH Mingarelli, Thomas 1 participant 1 message Mar 29, 2012 2 messages not in the archive IOMMU

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): 20120228190537.10845.31327.sendpatchset-G9O6rAK5tKBoF4rmPqRKaQC/[email protected]

Mingarelli, Thomas

author Tested-by Tony Camuso
#2 lore raw

This patch is being submitted to handle the case where a pci device is placed into the si domain, when booting in iommu passthrough mode, then removed. The RMRR information for such devices need to be re-processed to avoid DMA Read errors due to the Present Bit being cleared in the device's context entry.

Details of testing performed (per Mr. Camuso):

I think I was able to prove that your patch works with si domain devices.

I was able to Device Assign the two Patsburg USB controllers on the DL360G8 to your RH6 VM.

When I did that, I saw them on the VM guest, but they did not appear on the host. This is exactly the behavior you expect for Device-Assigned devices.

Guest ..
[root@rh6 ~]# lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub Bus 002 Device 002: ID 0627:0001 Adomax Technology Co., Ltd Bus 002 Device 003: ID 0000:0000 Bus 001 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse

Host ...
[root@dl380g8 ~]# lsusb
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 003 Device 002: ID 03f0:7029 Hewlett-Packard [root@dl380g8 ~]#

When I powered-down the guest, the USB controllers appeared on the host system again.

[root@dl380g8 ~]# lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 003 Device 002: ID 03f0:7029 Hewlett-Packard Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub Bus 002 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub Bus 001 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse Bus 002 Device 003: ID 0424:2660 Standard Microsystems Corp. Bus 002 Device 004: ID 03f0:3207 Hewlett-Packard 4 GB flash drive
[root@dl380g8 ~]#

Then, I powered-up the guest again, the usb controllers were there on the guest system again.

[root@localhost ~]# lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub Bus 002 Device 002: ID 0627:0001 Adomax Technology Co., Ltd Bus 002 Device 003: ID 0000:0000 Bus 001 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse [root@localhost ~]#

And they had vanished from the host system.

[root@dl380g8 ~]# lsusb
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 003 Device 002: ID 03f0:7029 Hewlett-Packard [root@dl380g8 ~]#

I believe this proves that your patch works correctly with si domain devices.

PATCH FOLLOWS...

drivers/iommu/intel-iommu.c renamed from drivers/iommu/intel-iommu.c.ORIG +20 -2
diff -up linux-3.2.4/drivers/iommu/intel-iommu.c.ORIG linux-3.2.4/drivers/iommu/intel-iommu.c
@@ -2676,7 +2676,8 @@ static int iommu_dummy(struct pci_dev *p
 static int iommu_no_mapping(struct device *dev)
 {
 	struct pci_dev *pdev;
-	int found;
+	struct dmar_rmrr_unit *rmrr;
+	int i, ret, found;
 
 	if (unlikely(dev->bus != &pci_bus_type))
 		return 1;
@@ -2699,7 +2700,24 @@ static int iommu_no_mapping(struct devic
 			 */
 			domain_remove_one_dev_info(si_domain, pdev);
 			printk(KERN_INFO "32bit %s uses non-identity mapping\n",
-			       pci_name(pdev));
+				pci_name(pdev));
+			printk(KERN_INFO "IOMMU: Re-processing RMRR information for device %s:\n",
+				pci_name(pdev));
+			for_each_rmrr_units(rmrr) {
+				for (i = 0; i < rmrr->devices_cnt; i++) {
+					/*
+					 * Here we are just concerned with finding the
+					 * one device that was removed from the
+					 * si_domain and re-evaluating its RMRR info.
+					 */
+					if (rmrr->devices[i] != pdev)
+						continue;
+					ret = iommu_prepare_rmrr_dev(rmrr, pdev);
+					if (ret)
+						printk(KERN_ERR
+							"IOMMU: mapping reserved region failed for device.\n");
+				}
+			}
 			return 0;
 		}
 	} else {