Merge patch series "Connect VFIO to IOMMUFD"

Jason Gunthorpe <[email protected]> says:

==================
This series provides an alternative container layer for VFIO implemented
using iommufd. This is optional, if CONFIG_IOMMUFD is not set then it will
not be compiled in.

At this point iommufd can be injected by passing in a iommfd FD to
VFIO_GROUP_SET_CONTAINER which will use the VFIO compat layer in iommufd
to obtain the compat IOAS and then connect up all the VFIO drivers as
appropriate.

This is temporary stopping point, a following series will provide a way to
directly open a VFIO device FD and directly connect it to IOMMUFD using
native ioctls that can expose the IOMMUFD features like hwpt, future
vPASID and dynamic attachment.

This series, in compat mode, has passed all the qemu tests we have
available, including the test suites for the Intel GVT mdev. Aside from
the temporary limitation with P2P memory this is belived to be fully
compatible with VFIO.

This is on github: https://github.com/jgunthorpe/linux/commits/vfio_iommufd

It requires the iommufd series:

https://lore.kernel.org/r/[email protected]
==================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jason Gunthorpe <[email protected]>

Author: jgunthorpe

drivers/gpu/drm/i915/gvt/kvmgt.c

@@ -1484,6 +1484,9 @@ static const struct vfio_device_ops intel_vgpu_dev_ops = {
 	.mmap		= intel_vgpu_mmap,
 	.ioctl		= intel_vgpu_ioctl,
 	.dma_unmap	= intel_vgpu_dma_unmap,
+	.bind_iommufd	= vfio_iommufd_emulated_bind,
+	.unbind_iommufd = vfio_iommufd_emulated_unbind,
+	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
 };
 
 static int intel_vgpu_probe(struct mdev_device *mdev)

drivers/iommu/iommufd/Kconfig

@@ -12,6 +12,26 @@ config IOMMUFD
 	  If you don't know what to do here, say N.
 
 if IOMMUFD
+config IOMMUFD_VFIO_CONTAINER
+	bool "IOMMUFD provides the VFIO container /dev/vfio/vfio"
+	depends on VFIO && !VFIO_CONTAINER
+	default VFIO && !VFIO_CONTAINER
+	help
+	  IOMMUFD will provide /dev/vfio/vfio instead of VFIO. This relies on
+	  IOMMUFD providing compatibility emulation to give the same ioctls.
+	  It provides an option to build a kernel with legacy VFIO components
+	  removed.
+
+	  IOMMUFD VFIO container emulation is known to lack certain features
+	  of the native VFIO container, such as no-IOMMU support, peer-to-peer
+	  DMA mapping, PPC IOMMU support, as well as other potentially
+	  undiscovered gaps.  This option is currently intended for the
+	  purpose of testing IOMMUFD with unmodified userspace supporting VFIO
+	  and making use of the Type1 VFIO IOMMU backend.  General purpose
+	  enabling of this option is currently discouraged.
+
+	  Unless testing IOMMUFD, say N here.
+
 config IOMMUFD_TEST
 	bool "IOMMU Userspace API Test support"
 	depends on DEBUG_KERNEL

drivers/iommu/iommufd/main.c

@@ -18,13 +18,15 @@
 #include <uapi/linux/iommufd.h>
 #include <linux/iommufd.h>
 
+#include "io_pagetable.h"
 #include "iommufd_private.h"
 #include "iommufd_test.h"
 
 struct iommufd_object_ops {
 	void (*destroy)(struct iommufd_object *obj);
 };
 static const struct iommufd_object_ops iommufd_object_ops[];
+static struct miscdevice vfio_misc_dev;
 
 struct iommufd_object *_iommufd_object_alloc(struct iommufd_ctx *ictx,
 					     size_t size,
@@ -170,6 +172,16 @@ static int iommufd_fops_open(struct inode *inode, struct file *filp)
 	if (!ictx)
 		return -ENOMEM;
 
+	/*
+	 * For compatibility with VFIO when /dev/vfio/vfio is opened we default
+	 * to the same rlimit accounting as vfio uses.
+	 */
+	if (IS_ENABLED(CONFIG_IOMMUFD_VFIO_CONTAINER) &&
+	    filp->private_data == &vfio_misc_dev) {
+		ictx->account_mode = IOPT_PAGES_ACCOUNT_MM;
+		pr_info_once("IOMMUFD is providing /dev/vfio/vfio, not VFIO.\n");
+	}
+
 	xa_init_flags(&ictx->objects, XA_FLAGS_ALLOC1 | XA_FLAGS_ACCOUNT);
 	ictx->file = filp;
 	filp->private_data = ictx;
@@ -400,25 +412,49 @@ static struct miscdevice iommu_misc_dev = {
 	.mode = 0660,
 };
 
+
+static struct miscdevice vfio_misc_dev = {
+	.minor = VFIO_MINOR,
+	.name = "vfio",
+	.fops = &iommufd_fops,
+	.nodename = "vfio/vfio",
+	.mode = 0666,
+};
+
 static int __init iommufd_init(void)
 {
 	int ret;
 
 	ret = misc_register(&iommu_misc_dev);
 	if (ret)
 		return ret;
+
+	if (IS_ENABLED(CONFIG_IOMMUFD_VFIO_CONTAINER)) {
+		ret = misc_register(&vfio_misc_dev);
+		if (ret)
+			goto err_misc;
+	}
 	iommufd_test_init();
 	return 0;
+err_misc:
+	misc_deregister(&iommu_misc_dev);
+	return ret;
 }
 
 static void __exit iommufd_exit(void)
 {
 	iommufd_test_exit();
+	if (IS_ENABLED(CONFIG_IOMMUFD_VFIO_CONTAINER))
+		misc_deregister(&vfio_misc_dev);
 	misc_deregister(&iommu_misc_dev);
 }
 
 module_init(iommufd_init);
 module_exit(iommufd_exit);
 
+#if IS_ENABLED(CONFIG_IOMMUFD_VFIO_CONTAINER)
+MODULE_ALIAS_MISCDEV(VFIO_MINOR);
+MODULE_ALIAS("devname:vfio/vfio");
+#endif
 MODULE_DESCRIPTION("I/O Address Space Management for passthrough devices");
 MODULE_LICENSE("GPL");

drivers/s390/cio/vfio_ccw_ops.c

@@ -588,6 +588,9 @@ static const struct vfio_device_ops vfio_ccw_dev_ops = {
 	.ioctl = vfio_ccw_mdev_ioctl,
 	.request = vfio_ccw_mdev_request,
 	.dma_unmap = vfio_ccw_dma_unmap,
+	.bind_iommufd = vfio_iommufd_emulated_bind,
+	.unbind_iommufd = vfio_iommufd_emulated_unbind,
+	.attach_ioas = vfio_iommufd_emulated_attach_ioas,
 };
 
 struct mdev_driver vfio_ccw_mdev_driver = {

drivers/s390/crypto/vfio_ap_ops.c

@@ -1805,6 +1805,9 @@ static const struct vfio_device_ops vfio_ap_matrix_dev_ops = {
 	.close_device = vfio_ap_mdev_close_device,
 	.ioctl = vfio_ap_mdev_ioctl,
 	.dma_unmap = vfio_ap_mdev_dma_unmap,
+	.bind_iommufd = vfio_iommufd_emulated_bind,
+	.unbind_iommufd = vfio_iommufd_emulated_unbind,
+	.attach_ioas = vfio_iommufd_emulated_attach_ioas,
 };
 
 static struct mdev_driver vfio_ap_matrix_driver = {

drivers/vfio/Kconfig

@@ -2,15 +2,28 @@
 menuconfig VFIO
 	tristate "VFIO Non-Privileged userspace driver framework"
 	select IOMMU_API
-	select VFIO_IOMMU_TYPE1 if MMU && (X86 || S390 || ARM || ARM64)
+	depends on IOMMUFD || !IOMMUFD
 	select INTERVAL_TREE
+	select VFIO_CONTAINER if IOMMUFD=n
 	help
 	  VFIO provides a framework for secure userspace device drivers.
 	  See Documentation/driver-api/vfio.rst for more details.
 
 	  If you don't know what to do here, say N.
 
 if VFIO
+config VFIO_CONTAINER
+	bool "Support for the VFIO container /dev/vfio/vfio"
+	select VFIO_IOMMU_TYPE1 if MMU && (X86 || S390 || ARM || ARM64)
+	default y
+	help
+	  The VFIO container is the classic interface to VFIO for establishing
+	  IOMMU mappings. If N is selected here then IOMMUFD must be used to
+	  manage the mappings.
+
+	  Unless testing IOMMUFD say Y here.
+
+if VFIO_CONTAINER
 config VFIO_IOMMU_TYPE1
 	tristate
 	default n
@@ -20,16 +33,6 @@ config VFIO_IOMMU_SPAPR_TCE
 	depends on SPAPR_TCE_IOMMU
 	default VFIO
 
-config VFIO_SPAPR_EEH
-	tristate
-	depends on EEH && VFIO_IOMMU_SPAPR_TCE
-	default VFIO
-
-config VFIO_VIRQFD
-	tristate
-	select EVENTFD
-	default n
-
 config VFIO_NOIOMMU
 	bool "VFIO No-IOMMU support"
 	help
@@ -43,6 +46,17 @@ config VFIO_NOIOMMU
 	  this mode since there is no IOMMU to provide DMA translation.
 
 	  If you don't know what to do here, say N.
+endif
+
+config VFIO_SPAPR_EEH
+	tristate
+	depends on EEH && VFIO_IOMMU_SPAPR_TCE
+	default VFIO
+
+config VFIO_VIRQFD
+	tristate
+	select EVENTFD
+	default n
 
 source "drivers/vfio/pci/Kconfig"
 source "drivers/vfio/platform/Kconfig"

drivers/vfio/Makefile

@@ -4,8 +4,9 @@ vfio_virqfd-y := virqfd.o
 obj-$(CONFIG_VFIO) += vfio.o
 
 vfio-y += vfio_main.o \
-	  iova_bitmap.o \
-	  container.o
+	  iova_bitmap.o
+vfio-$(CONFIG_IOMMUFD) += iommufd.o
+vfio-$(CONFIG_VFIO_CONTAINER) += container.o
 
 obj-$(CONFIG_VFIO_VIRQFD) += vfio_virqfd.o
 obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o