[PATCH v8 4/8] s390x/pci: don't fence interpreted devices without MSI-X

[Matthew Rosato <mjrosato@xxxxxxxxxxxxx>]

Lack of MSI-X support is not an issue for interpreted passthrough
devices, so let's let these in.  This will allow, for example, ISM
devices to be passed through -- but only when interpretation is
available and being used.

Reviewed-by: Thomas Huth <thuth@xxxxxxxxxx>
Reviewed-by: Pierre Morel <pmorel@xxxxxxxxxxxxx>
Signed-off-by: Matthew Rosato <mjrosato@xxxxxxxxxxxxx>
---
 hw/s390x/s390-pci-bus.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/hw/s390x/s390-pci-bus.c b/hw/s390x/s390-pci-bus.c
index 156051e6e9..816d17af99 100644
--- a/hw/s390x/s390-pci-bus.c
+++ b/hw/s390x/s390-pci-bus.c
@@ -881,6 +881,10 @@ static int s390_pci_msix_init(S390PCIBusDevice *pbdev)
 
 static void s390_pci_msix_free(S390PCIBusDevice *pbdev)
 {
+    if (pbdev->msix.entries == 0) {
+        return;
+    }
+
     memory_region_del_subregion(&pbdev->iommu->mr, &pbdev->msix_notify_mr);
     object_unparent(OBJECT(&pbdev->msix_notify_mr));
 }
@@ -1093,7 +1097,7 @@ static void s390_pcihost_plug(HotplugHandler *hotplug_dev, DeviceState *dev,
             pbdev->interp = false;
         }
 
-        if (s390_pci_msix_init(pbdev)) {
+        if (s390_pci_msix_init(pbdev) && !pbdev->interp) {
             error_setg(errp, "MSI-X support is mandatory "
                        "in the S390 architecture");
             return;
-- 
2.37.2