[Qemu-devel] [PATCH qemu v9 10/13] spapr_pci_vfio: Enable multiple groups per container

[Alexey Kardashevskiy <aik@ozlabs.ru>]

This enables multiple IOMMU groups in one VFIO container which means
that multiple devices from different groups can share the same IOMMU
table (or tables if DDW).

This removes a group id from vfio_container_ioctl(). The kernel support
is required for this; if the host kernel does not have the support,
it will allow only one group per container. The PHB's "iommuid" property
is ignored. The ioctl is called for every container attached to
the address space. At the moment there is just one container anyway.

If there is no container attached to the address space,
vfio_container_do_ioctl() returns -1.

This removes casts to sPAPRPHBVFIOState as none of sPAPRPHBVFIOState
members is accessed here.

Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
---
 hw/ppc/spapr_pci_vfio.c | 17 ++++++-----------
 hw/vfio/common.c        | 20 ++++++--------------
 include/hw/vfio/vfio.h  |  2 +-
 3 files changed, 13 insertions(+), 26 deletions(-)

diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c
index 04ca4cf..fe7d7d1 100644
--- a/hw/ppc/spapr_pci_vfio.c
+++ b/hw/ppc/spapr_pci_vfio.c
@@ -33,7 +33,7 @@ int spapr_phb_vfio_dma_capabilities_update(sPAPRPHBState *sphb)
     struct vfio_iommu_spapr_tce_info info = { .argsz = sizeof(info) };
     int ret;
 
-    ret = vfio_container_ioctl(&sphb->iommu_as, sphb->iommugroupid,
+    ret = vfio_container_ioctl(&sphb->iommu_as,
                                VFIO_IOMMU_SPAPR_TCE_GET_INFO, &info);
     if (ret) {
         return ret;
@@ -59,8 +59,7 @@ void spapr_phb_vfio_eeh_reenable(sPAPRPHBState *sphb)
      * ensures that the contained PCI devices will work properly
      * after reboot.
      */
-    vfio_container_ioctl(&sphb->iommu_as,
-                         sphb->iommugroupid, VFIO_EEH_PE_OP, &op);
+    vfio_container_ioctl(&sphb->iommu_as, VFIO_EEH_PE_OP, &op);
 }
 
 int spapr_phb_vfio_eeh_set_option(sPAPRPHBState *sphb,
@@ -86,8 +85,7 @@ int spapr_phb_vfio_eeh_set_option(sPAPRPHBState *sphb,
         return RTAS_OUT_PARAM_ERROR;
     }
 
-    ret = vfio_container_ioctl(&sphb->iommu_as, sphb->iommugroupid,
-                               VFIO_EEH_PE_OP, &op);
+    ret = vfio_container_ioctl(&sphb->iommu_as, VFIO_EEH_PE_OP, &op);
     if (ret < 0) {
         return RTAS_OUT_HW_ERROR;
     }
@@ -101,8 +99,7 @@ int spapr_phb_vfio_eeh_get_state(sPAPRPHBState *sphb, int *state)
     int ret;
 
     op.op = VFIO_EEH_PE_GET_STATE;
-    ret = vfio_container_ioctl(&sphb->iommu_as, sphb->iommugroupid,
-                               VFIO_EEH_PE_OP, &op);
+    ret = vfio_container_ioctl(&sphb->iommu_as, VFIO_EEH_PE_OP, &op);
     if (ret < 0) {
         return RTAS_OUT_PARAM_ERROR;
     }
@@ -175,8 +172,7 @@ int spapr_phb_vfio_eeh_reset(sPAPRPHBState *sphb, int option)
         return RTAS_OUT_PARAM_ERROR;
     }
 
-    ret = vfio_container_ioctl(&sphb->iommu_as, sphb->iommugroupid,
-                               VFIO_EEH_PE_OP, &op);
+    ret = vfio_container_ioctl(&sphb->iommu_as, VFIO_EEH_PE_OP, &op);
     if (ret < 0) {
         return RTAS_OUT_HW_ERROR;
     }
@@ -190,8 +186,7 @@ int spapr_phb_vfio_eeh_configure(sPAPRPHBState *sphb)
     int ret;
 
     op.op = VFIO_EEH_PE_CONFIGURE;
-    ret = vfio_container_ioctl(&sphb->iommu_as, sphb->iommugroupid,
-                               VFIO_EEH_PE_OP, &op);
+    ret = vfio_container_ioctl(&sphb->iommu_as, VFIO_EEH_PE_OP, &op);
     if (ret < 0) {
         return RTAS_OUT_PARAM_ERROR;
     }
diff --git a/hw/vfio/common.c b/hw/vfio/common.c
index b1045da..89ef37b 100644
--- a/hw/vfio/common.c
+++ b/hw/vfio/common.c
@@ -918,34 +918,26 @@ void vfio_put_base_device(VFIODevice *vbasedev)
     close(vbasedev->fd);
 }
 
-static int vfio_container_do_ioctl(AddressSpace *as, int32_t groupid,
+static int vfio_container_do_ioctl(AddressSpace *as,
                                    int req, void *param)
 {
-    VFIOGroup *group;
     VFIOContainer *container;
     int ret = -1;
+    VFIOAddressSpace *space = vfio_get_address_space(as);
 
-    group = vfio_get_group(groupid, as);
-    if (!group) {
-        error_report("vfio: group %d not registered", groupid);
-        return ret;
-    }
-
-    container = group->container;
-    if (group->container) {
+    QLIST_FOREACH(container, &space->containers, next) {
         ret = ioctl(container->fd, req, param);
         if (ret < 0) {
             error_report("vfio: failed to ioctl %d to container: ret=%d, %s",
                          _IOC_NR(req) - VFIO_BASE, ret, strerror(errno));
+            return -errno;
         }
     }
 
-    vfio_put_group(group);
-
     return ret;
 }
 
-int vfio_container_ioctl(AddressSpace *as, int32_t groupid,
+int vfio_container_ioctl(AddressSpace *as,
                          int req, void *param)
 {
     /* We allow only certain ioctls to the container */
@@ -960,5 +952,5 @@ int vfio_container_ioctl(AddressSpace *as, int32_t groupid,
         return -1;
     }
 
-    return vfio_container_do_ioctl(as, groupid, req, param);
+    return vfio_container_do_ioctl(as, req, param);
 }
diff --git a/include/hw/vfio/vfio.h b/include/hw/vfio/vfio.h
index 0b26cd8..76b5744 100644
--- a/include/hw/vfio/vfio.h
+++ b/include/hw/vfio/vfio.h
@@ -3,7 +3,7 @@
 
 #include "qemu/typedefs.h"
 
-extern int vfio_container_ioctl(AddressSpace *as, int32_t groupid,
+extern int vfio_container_ioctl(AddressSpace *as,
                                 int req, void *param);
 
 #endif
-- 
2.4.0.rc3.8.gfb3e7d5