virtio-iommu: Default to bypass during boot

Currently the virtio-iommu device must be programmed before it allows
DMA from any PCI device. This can make the VM entirely unusable when a
virtio-iommu driver isn't present, for example in a bootloader that
loads the OS from storage.

Similarly to the other vIOMMU implementations, default to DMA bypassing
the IOMMU during boot. Add a "boot-bypass" property, defaulting to true,
that lets users change this behavior.

Replace the VIRTIO_IOMMU_F_BYPASS feature, which didn't support bypass
before feature negotiation, with VIRTIO_IOMMU_F_BYPASS_CONFIG.

We add the bypass field to the migration stream without introducing
subsections, based on the assumption that this virtio-iommu device isn't
being used in production enough to require cross-version migration at
the moment (all previous version required workarounds since they didn't
support ACPI and boot-bypass).

Reviewed-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Message-Id: <20220214124356.872985-3-jean-philippe@linaro.org>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

2 files changed, 55 insertions, 9 deletions

diff --git a/hw/virtio/trace-events b/hw/virtio/trace-events
index f7ad6be5fb..a5102eac9e 100644
--- a/hw/virtio/trace-events
+++ b/hw/virtio/trace-events
@@ -89,9 +89,11 @@ virtio_mmio_setting_irq(int level) "virtio_mmio setting IRQ %d"
 
 # virtio-iommu.c
 virtio_iommu_device_reset(void) "reset!"
+virtio_iommu_system_reset(void) "system reset!"
 virtio_iommu_get_features(uint64_t features) "device supports features=0x%"PRIx64
 virtio_iommu_device_status(uint8_t status) "driver status = %d"
-virtio_iommu_get_config(uint64_t page_size_mask, uint64_t start, uint64_t end, uint32_t domain_start, uint32_t domain_end, uint32_t probe_size) "page_size_mask=0x%"PRIx64" input range start=0x%"PRIx64" input range end=0x%"PRIx64" domain range start=%d domain range end=%d probe_size=0x%x"
+virtio_iommu_get_config(uint64_t page_size_mask, uint64_t start, uint64_t end, uint32_t domain_start, uint32_t domain_end, uint32_t probe_size, uint8_t bypass) "page_size_mask=0x%"PRIx64" input range start=0x%"PRIx64" input range end=0x%"PRIx64" domain range start=%d domain range end=%d probe_size=0x%x bypass=0x%x"
+virtio_iommu_set_config(uint8_t bypass) "bypass=0x%x"
 virtio_iommu_attach(uint32_t domain_id, uint32_t ep_id) "domain=%d endpoint=%d"
 virtio_iommu_detach(uint32_t domain_id, uint32_t ep_id) "domain=%d endpoint=%d"
 virtio_iommu_map(uint32_t domain_id, uint64_t virt_start, uint64_t virt_end, uint64_t phys_start, uint32_t flags) "domain=%d virt_start=0x%"PRIx64" virt_end=0x%"PRIx64 " phys_start=0x%"PRIx64" flags=%d"
diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c
index aa9c16a17b..4ca36db4ac 100644
--- a/hw/virtio/virtio-iommu.c
+++ b/hw/virtio/virtio-iommu.c
@@ -24,6 +24,7 @@
 #include "hw/qdev-properties.h"
 #include "hw/virtio/virtio.h"
 #include "sysemu/kvm.h"
+#include "sysemu/reset.h"
 #include "qapi/error.h"
 #include "qemu/error-report.h"
 #include "trace.h"
@@ -728,8 +729,7 @@ static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr,
         .perm = IOMMU_NONE,
     };
 
-    bypass_allowed = virtio_vdev_has_feature(&s->parent_obj,
-                                             VIRTIO_IOMMU_F_BYPASS);
+    bypass_allowed = s->config.bypass;
 
     sid = virtio_iommu_get_bdf(sdev);
 
@@ -831,13 +831,37 @@ static void virtio_iommu_get_config(VirtIODevice *vdev, uint8_t *config_data)
     out_config->domain_range.start = cpu_to_le32(dev_config->domain_range.start);
     out_config->domain_range.end = cpu_to_le32(dev_config->domain_range.end);
     out_config->probe_size = cpu_to_le32(dev_config->probe_size);
+    out_config->bypass = dev_config->bypass;
 
     trace_virtio_iommu_get_config(dev_config->page_size_mask,
                                   dev_config->input_range.start,
                                   dev_config->input_range.end,
                                   dev_config->domain_range.start,
                                   dev_config->domain_range.end,
-                                  dev_config->probe_size);
+                                  dev_config->probe_size,
+                                  dev_config->bypass);
+}
+
+static void virtio_iommu_set_config(VirtIODevice *vdev,
+                                    const uint8_t *config_data)
+{
+    VirtIOIOMMU *dev = VIRTIO_IOMMU(vdev);
+    struct virtio_iommu_config *dev_config = &dev->config;
+    const struct virtio_iommu_config *in_config = (void *)config_data;
+
+    if (in_config->bypass != dev_config->bypass) {
+        if (!virtio_vdev_has_feature(vdev, VIRTIO_IOMMU_F_BYPASS_CONFIG)) {
+            virtio_error(vdev, "cannot set config.bypass");
+            return;
+        } else if (in_config->bypass != 0 && in_config->bypass != 1) {
+            virtio_error(vdev, "invalid config.bypass value '%u'",
+                         in_config->bypass);
+            return;
+        }
+        dev_config->bypass = in_config->bypass;
+    }
+
+    trace_virtio_iommu_set_config(in_config->bypass);
 }
 
 static uint64_t virtio_iommu_get_features(VirtIODevice *vdev, uint64_t f,
@@ -963,6 +987,19 @@ static int virtio_iommu_set_page_size_mask(IOMMUMemoryRegion *mr,
     return 0;
 }
 
+static void virtio_iommu_system_reset(void *opaque)
+{
+    VirtIOIOMMU *s = opaque;
+
+    trace_virtio_iommu_system_reset();
+
+    /*
+     * config.bypass is sticky across device reset, but should be restored on
+     * system reset
+     */
+    s->config.bypass = s->boot_bypass;
+}
+
 static void virtio_iommu_device_realize(DeviceState *dev, Error **errp)
 {
     VirtIODevice *vdev = VIRTIO_DEVICE(dev);
@@ -988,9 +1025,9 @@ static void virtio_iommu_device_realize(DeviceState *dev, Error **errp)
     virtio_add_feature(&s->features, VIRTIO_IOMMU_F_INPUT_RANGE);
     virtio_add_feature(&s->features, VIRTIO_IOMMU_F_DOMAIN_RANGE);
     virtio_add_feature(&s->features, VIRTIO_IOMMU_F_MAP_UNMAP);
-    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_BYPASS);
     virtio_add_feature(&s->features, VIRTIO_IOMMU_F_MMIO);
     virtio_add_feature(&s->features, VIRTIO_IOMMU_F_PROBE);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_BYPASS_CONFIG);
 
     qemu_mutex_init(&s->mutex);
 
@@ -1001,6 +1038,8 @@ static void virtio_iommu_device_realize(DeviceState *dev, Error **errp)
     } else {
         error_setg(errp, "VIRTIO-IOMMU is not attached to any PCI bus!");
     }
+
+    qemu_register_reset(virtio_iommu_system_reset, s);
 }
 
 static void virtio_iommu_device_unrealize(DeviceState *dev)
@@ -1008,6 +1047,8 @@ static void virtio_iommu_device_unrealize(DeviceState *dev)
     VirtIODevice *vdev = VIRTIO_DEVICE(dev);
     VirtIOIOMMU *s = VIRTIO_IOMMU(dev);
 
+    qemu_unregister_reset(virtio_iommu_system_reset, s);
+
     g_hash_table_destroy(s->as_by_busptr);
     if (s->domains) {
         g_tree_destroy(s->domains);
@@ -1141,21 +1182,22 @@ static int iommu_post_load(void *opaque, int version_id)
 
 static const VMStateDescription vmstate_virtio_iommu_device = {
     .name = "virtio-iommu-device",
-    .minimum_version_id = 1,
-    .version_id = 1,
+    .minimum_version_id = 2,
+    .version_id = 2,
     .post_load = iommu_post_load,
     .fields = (VMStateField[]) {
         VMSTATE_GTREE_DIRECT_KEY_V(domains, VirtIOIOMMU, 1,
                                    &vmstate_domain, VirtIOIOMMUDomain),
+        VMSTATE_UINT8_V(config.bypass, VirtIOIOMMU, 2),
         VMSTATE_END_OF_LIST()
     },
 };
 
 static const VMStateDescription vmstate_virtio_iommu = {
     .name = "virtio-iommu",
-    .minimum_version_id = 1,
+    .minimum_version_id = 2,
     .priority = MIG_PRI_IOMMU,
-    .version_id = 1,
+    .version_id = 2,
     .fields = (VMStateField[]) {
         VMSTATE_VIRTIO_DEVICE,
         VMSTATE_END_OF_LIST()
@@ -1164,6 +1206,7 @@ static const VMStateDescription vmstate_virtio_iommu = {
 
 static Property virtio_iommu_properties[] = {
     DEFINE_PROP_LINK("primary-bus", VirtIOIOMMU, primary_bus, "PCI", PCIBus *),
+    DEFINE_PROP_BOOL("boot-bypass", VirtIOIOMMU, boot_bypass, true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
@@ -1180,6 +1223,7 @@ static void virtio_iommu_class_init(ObjectClass *klass, void *data)
     vdc->unrealize = virtio_iommu_device_unrealize;
     vdc->reset = virtio_iommu_device_reset;
     vdc->get_config = virtio_iommu_get_config;
+    vdc->set_config = virtio_iommu_set_config;
     vdc->get_features = virtio_iommu_get_features;
     vdc->set_status = virtio_iommu_set_status;
     vdc->vmsd = &vmstate_virtio_iommu_device;