Re: [PATCH 00/27] vfio-user client

[John Levon <john.levon@nutanix.com>]

On Mon, May 19, 2025 at 02:40:12PM +0200, Cédric Le Goater wrote:

> patches 3-5 seem be ok to merge. They are first on my vfio-next
> candidates.
> 
> patches 1-2 should be reworked on top of the memory_get_xlat_addr()
> changes [1] proposed by Steven.

Now v5 is out I will rebase on top of it.

> The rest (9-27) is vfio-user territory only (I hope). Still, it
> should be merged through the VFIO tree.

It isn't entirely contained within hw/vfio-user/ (there are e.g. some exports of
util functions), but yes essentially. See diff below for illustration.

> Would it be complex to propose a functional test for it ?

We have a libvfio-user we can use, but this would be a problem:

 487   │ 2025-05-19 14:19:20,304: modprobe gpio-pci-idio-16
 488   │ 2025-05-19 14:19:20,306: modprobe: FATAL: Module gpio-pci-idio-16 not found in directory /lib/modules/5.3.7-301.fc31.x86_64

If somebody can help with how to get a suitable test initrd with this module
available, we could at least do basic testing (although no DMA).

regards
john


diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h
index 7a03d24805..80599454d4 100644
--- a/hw/vfio/pci.h
+++ b/hw/vfio/pci.h
@@ -116,6 +116,7 @@ typedef struct VFIOMSIXInfo {
     uint32_t pba_offset;
     unsigned long *pending;
     bool noresize;
+    MemoryRegion *pba_region;
 } VFIOMSIXInfo;
 
 /*
@@ -215,6 +216,20 @@ uint32_t vfio_pci_read_config(PCIDevice *pdev, uint32_t addr, int len);
 void vfio_pci_write_config(PCIDevice *pdev,
                            uint32_t addr, uint32_t val, int len);
 
+void vfio_intx_eoi(VFIODevice *vbasedev);
+Object *vfio_pci_get_object(VFIODevice *vbasedev);
+int vfio_pci_save_config(VFIODevice *vbasedev, QEMUFile *f, Error **errp);
+int vfio_pci_load_config(VFIODevice *vbasedev, QEMUFile *f);
+void vfio_pci_put_device(VFIOPCIDevice *vdev);
+bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp);
+void vfio_teardown_msi(VFIOPCIDevice *vdev);
+void vfio_bars_exit(VFIOPCIDevice *vdev);
+bool vfio_add_capabilities(VFIOPCIDevice *vdev, Error **errp);
+void vfio_register_err_notifier(VFIOPCIDevice *vdev);
+void vfio_register_req_notifier(VFIOPCIDevice *vdev);
+bool vfio_pci_config_setup(VFIOPCIDevice *vdev, Error **errp);
+bool vfio_interrupt_setup(VFIOPCIDevice *vdev, Error **errp);
+
 uint64_t vfio_vga_read(void *opaque, hwaddr addr, unsigned size);
 void vfio_vga_write(void *opaque, hwaddr addr, uint64_t data, unsigned size);
 
diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h
index c18986a621..c854f902ed 100644
--- a/include/hw/vfio/vfio-container-base.h
+++ b/include/hw/vfio/vfio-container-base.h
@@ -109,6 +109,7 @@ vfio_container_get_page_size_mask(const VFIOContainerBase *bcontainer)
 #define TYPE_VFIO_IOMMU_LEGACY TYPE_VFIO_IOMMU "-legacy"
 #define TYPE_VFIO_IOMMU_SPAPR TYPE_VFIO_IOMMU "-spapr"
 #define TYPE_VFIO_IOMMU_IOMMUFD TYPE_VFIO_IOMMU "-iommufd"
+#define TYPE_VFIO_IOMMU_USER TYPE_VFIO_IOMMU "-user"
 
 OBJECT_DECLARE_TYPE(VFIOContainerBase, VFIOIOMMUClass, VFIO_IOMMU)
 
diff --git a/include/hw/vfio/vfio-device.h b/include/hw/vfio/vfio-device.h
index a23ef4ea13..09f1a21bf8 100644
--- a/include/hw/vfio/vfio-device.h
+++ b/include/hw/vfio/vfio-device.h
@@ -46,6 +46,7 @@ typedef struct VFIOMigration VFIOMigration;
 
 typedef struct IOMMUFDBackend IOMMUFDBackend;
 typedef struct VFIOIOASHwpt VFIOIOASHwpt;
+typedef struct VFIOUserProxy VFIOUserProxy;
 
 typedef struct VFIODevice {
     QLIST_ENTRY(VFIODevice) next;
@@ -86,6 +87,7 @@ typedef struct VFIODevice {
     QLIST_ENTRY(VFIODevice) hwpt_next;
     struct vfio_region_info **reginfo;
     int *region_fds;
+    VFIOUserProxy *proxy;
 } VFIODevice;
 
 struct VFIODeviceOps {
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 57ef65dcc9..a92e38f822 100644
--- a/hw/vfio/pci.c
+++ b/hw/vfio/pci.c
@@ -103,7 +103,7 @@ static void vfio_intx_interrupt(void *opaque)
     }
 }
 
-static void vfio_intx_eoi(VFIODevice *vbasedev)
+void vfio_intx_eoi(VFIODevice *vbasedev)
 {
     VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev);
 
@@ -1770,7 +1770,7 @@ static bool vfio_msix_setup(VFIOPCIDevice *vdev, int pos, Error **errp)
     return true;
 }
 
-static void vfio_teardown_msi(VFIOPCIDevice *vdev)
+void vfio_teardown_msi(VFIOPCIDevice *vdev)
 {
     msi_uninit(&vdev->pdev);
 
@@ -1869,7 +1869,7 @@ static void vfio_bars_register(VFIOPCIDevice *vdev)
     }
 }
 
-static void vfio_bars_exit(VFIOPCIDevice *vdev)
+void vfio_bars_exit(VFIOPCIDevice *vdev)
 {
     int i;
 
@@ -2460,7 +2460,7 @@ static void vfio_add_ext_cap(VFIOPCIDevice *vdev)
     g_free(config);
 }
 
-static bool vfio_add_capabilities(VFIOPCIDevice *vdev, Error **errp)
+bool vfio_add_capabilities(VFIOPCIDevice *vdev, Error **errp)
 {
     PCIDevice *pdev = &vdev->pdev;
 
@@ -2629,7 +2629,7 @@ static void vfio_pci_compute_needs_reset(VFIODevice *vbasedev)
     }
 }
 
-static Object *vfio_pci_get_object(VFIODevice *vbasedev)
+Object *vfio_pci_get_object(VFIODevice *vbasedev)
 {
     VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev);
 
@@ -2685,7 +2685,7 @@ static const VMStateDescription vmstate_vfio_pci_config = {
     }
 };
 
-static int vfio_pci_save_config(VFIODevice *vbasedev, QEMUFile *f, Error **errp)
+int vfio_pci_save_config(VFIODevice *vbasedev, QEMUFile *f, Error **errp)
 {
     VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev);
 
@@ -2693,7 +2693,7 @@ static int vfio_pci_save_config(VFIODevice *vbasedev, QEMUFile *f, Error **errp)
                                        errp);
 }
 
-static int vfio_pci_load_config(VFIODevice *vbasedev, QEMUFile *f)
+int vfio_pci_load_config(VFIODevice *vbasedev, QEMUFile *f)
 {
     VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev);
     PCIDevice *pdev = &vdev->pdev;
@@ -2807,7 +2807,7 @@ bool vfio_populate_vga(VFIOPCIDevice *vdev, Error **errp)
     return true;
 }
 
-static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp)
+bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp)
 {
     VFIODevice *vbasedev = &vdev->vbasedev;
     struct vfio_region_info *reg_info = NULL;
@@ -2895,7 +2895,7 @@ static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp)
     return true;
 }
 
-static void vfio_pci_put_device(VFIOPCIDevice *vdev)
+void vfio_pci_put_device(VFIOPCIDevice *vdev)
 {
     vfio_display_finalize(vdev);
     vfio_bars_finalize(vdev);
@@ -2943,7 +2943,7 @@ static void vfio_err_notifier_handler(void *opaque)
  * and continue after disabling error recovery support for the
  * device.
  */
-static void vfio_register_err_notifier(VFIOPCIDevice *vdev)
+void vfio_register_err_notifier(VFIOPCIDevice *vdev)
 {
     Error *err = NULL;
     int32_t fd;
@@ -3002,7 +3002,7 @@ static void vfio_req_notifier_handler(void *opaque)
     }
 }
 
-static void vfio_register_req_notifier(VFIOPCIDevice *vdev)
+void vfio_register_req_notifier(VFIOPCIDevice *vdev)
 {
     struct vfio_irq_info irq_info;
     Error *err = NULL;
@@ -3056,7 +3056,7 @@ static void vfio_unregister_req_notifier(VFIOPCIDevice *vdev)
     vdev->req_enabled = false;
 }
 
-static bool vfio_pci_config_setup(VFIOPCIDevice *vdev, Error **errp)
+bool vfio_pci_config_setup(VFIOPCIDevice *vdev, Error **errp)
 {
     PCIDevice *pdev = &vdev->pdev;
     VFIODevice *vbasedev = &vdev->vbasedev;
@@ -3162,7 +3162,7 @@ static bool vfio_pci_config_setup(VFIOPCIDevice *vdev, Error **errp)
     return true;
 }
 
-static bool vfio_interrupt_setup(VFIOPCIDevice *vdev, Error **errp)
+bool vfio_interrupt_setup(VFIOPCIDevice *vdev, Error **errp)
 {
     PCIDevice *pdev = &vdev->pdev;