[PATCH v2 00/22] vfio: Give VFIO_DEVICE_GET_REGION_INFO its own op

[PATCH v2 00/22] vfio: Give VFIO_DEVICE_GET_REGION_INFO its own op

[Jason Gunthorpe]

There is alot of duplicated code in the drivers for processing
VFIO_DEVICE_GET_REGION_INFO. Introduce a new op get_region_info_caps()
which provides a struct vfio_info_cap and handles the cap chain logic
to write the caps back to userspace and remove all of this duplication
from drivers.

This is done in two steps, the first is a largely mechanical introduction
of the get_region_info(). These patches are best viewed with the diff
option to ignore whitespace (-b) as most of the lines are re-indending
things.

Then drivers are updated to remove the duplicate cap related code. Some
drivers are converted to use vfio_info_add_capability() instead of open
coding a version of it.

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

v2:
 - Rename
    hisi_acc_vfio_get_region -> hisi_acc_vfio_ioctl_get_region
    vfio_fsl_mc_get_region_info -> vfio_fsl_mc_ioctl_get_region_info
    intel_vgpu_get_region_info -> intel_vgpu_ioctl_get_region_info
    mbochs_get_region_info -> mbochs_ioctl_get_region_info
    intel_vgpu_get_region_info -> intel_vgpu_ioctl_get_region_info
    vfio_ccw_mdev_get_region_info -> vfio_ccw_mdev_ioctl_get_region_info
    hisi_acc_vfio_get_region -> hisi_acc_vfio_ioctl_get_region
    vfio_fsl_mc_get_region_info -> vfio_fsl_mc_ioctl_get_region_info
 - Consistently free caps.buf in vfio_get_region_info()
v1: https://patch.msgid.link/r/0-v1-679a6fa27d31+209-vfio_get_region_info_op_jgg@nvidia.com

Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

Jason Gunthorpe (22):
  vfio: Provide a get_region_info op
  vfio/hisi: Convert to the get_region_info op
  vfio/virtio: Convert to the get_region_info op
  vfio/nvgrace: Convert to the get_region_info op
  vfio/pci: Fill in the missing get_region_info ops
  vfio/mtty: Provide a get_region_info op
  vfio/mdpy: Provide a get_region_info op
  vfio/mbochs: Provide a get_region_info op
  vfio/platform: Provide a get_region_info op
  vfio/fsl: Provide a get_region_info op
  vfio/cdx: Provide a get_region_info op
  vfio/ccw: Provide a get_region_info op
  vfio/gvt: Provide a get_region_info op
  vfio: Require drivers to implement get_region_info
  vfio: Add get_region_info_caps op
  vfio/mbochs: Convert mbochs to use vfio_info_add_capability()
  vfio/gvt: Convert to get_region_info_caps
  vfio/ccw: Convert to get_region_info_caps
  vfio/pci: Convert all PCI drivers to get_region_info_caps
  vfio/platform: Convert to get_region_info_caps
  vfio: Move the remaining drivers to get_region_info_caps
  vfio: Remove the get_region_info op

 drivers/gpu/drm/i915/gvt/kvmgt.c              | 272 ++++++++----------
 drivers/s390/cio/vfio_ccw_ops.c               |  47 +--
 drivers/vfio/cdx/main.c                       |  29 +-
 drivers/vfio/fsl-mc/vfio_fsl_mc.c             |  43 ++-
 .../vfio/pci/hisilicon/hisi_acc_vfio_pci.c    |  54 ++--
 drivers/vfio/pci/mlx5/main.c                  |   1 +
 drivers/vfio/pci/nvgrace-gpu/main.c           |  53 +---
 drivers/vfio/pci/pds/vfio_dev.c               |   1 +
 drivers/vfio/pci/qat/main.c                   |   1 +
 drivers/vfio/pci/vfio_pci.c                   |   1 +
 drivers/vfio/pci/vfio_pci_core.c              | 110 +++----
 drivers/vfio/pci/virtio/common.h              |   5 +-
 drivers/vfio/pci/virtio/legacy_io.c           |  38 +--
 drivers/vfio/pci/virtio/main.c                |   5 +-
 drivers/vfio/platform/vfio_amba.c             |   1 +
 drivers/vfio/platform/vfio_platform.c         |   1 +
 drivers/vfio/platform/vfio_platform_common.c  |  40 ++-
 drivers/vfio/platform/vfio_platform_private.h |   3 +
 drivers/vfio/vfio_main.c                      |  49 ++++
 include/linux/vfio.h                          |   4 +
 include/linux/vfio_pci_core.h                 |   3 +
 samples/vfio-mdev/mbochs.c                    |  71 ++---
 samples/vfio-mdev/mdpy.c                      |  34 +--
 samples/vfio-mdev/mtty.c                      |  33 +--
 24 files changed, 368 insertions(+), 531 deletions(-)


base-commit: 211ddde0823f1442e4ad052a2f30f050145ccada
-- 
2.43.0

[PATCH v2 01/22] vfio: Provide a get_region_info op

[Jason Gunthorpe]

Instead of hooking the general ioctl op, have the core code directly
decode VFIO_DEVICE_GET_REGION_INFO and call an op just for it.

This is intended to allow mechanical changes to the drivers to pull their
VFIO_DEVICE_GET_REGION_INFO int oa function. Later patches will improve
the function signature to consolidate more code.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/pci/vfio_pci_core.c | 9 ++++++---
 drivers/vfio/vfio_main.c         | 7 +++++++
 include/linux/vfio.h             | 2 ++
 include/linux/vfio_pci_core.h    | 2 ++
 4 files changed, 17 insertions(+), 3 deletions(-)

diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
index 7dcf5439dedc9d..1dc350003f075c 100644
--- a/drivers/vfio/pci/vfio_pci_core.c
+++ b/drivers/vfio/pci/vfio_pci_core.c
@@ -996,9 +996,11 @@ static int vfio_pci_ioctl_get_info(struct vfio_pci_core_device *vdev,
 	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 }
 
-static int vfio_pci_ioctl_get_region_info(struct vfio_pci_core_device *vdev,
-					  struct vfio_region_info __user *arg)
+int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
+				   struct vfio_region_info __user *arg)
 {
+	struct vfio_pci_core_device *vdev =
+		container_of(core_vdev, struct vfio_pci_core_device, vdev);
 	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
 	struct pci_dev *pdev = vdev->pdev;
 	struct vfio_region_info info;
@@ -1132,6 +1134,7 @@ static int vfio_pci_ioctl_get_region_info(struct vfio_pci_core_device *vdev,
 
 	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 }
+EXPORT_SYMBOL_GPL(vfio_pci_ioctl_get_region_info);
 
 static int vfio_pci_ioctl_get_irq_info(struct vfio_pci_core_device *vdev,
 				       struct vfio_irq_info __user *arg)
@@ -1458,7 +1461,7 @@ long vfio_pci_core_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
 	case VFIO_DEVICE_GET_PCI_HOT_RESET_INFO:
 		return vfio_pci_ioctl_get_pci_hot_reset_info(vdev, uarg);
 	case VFIO_DEVICE_GET_REGION_INFO:
-		return vfio_pci_ioctl_get_region_info(vdev, uarg);
+		return vfio_pci_ioctl_get_region_info(core_vdev, uarg);
 	case VFIO_DEVICE_IOEVENTFD:
 		return vfio_pci_ioctl_ioeventfd(vdev, uarg);
 	case VFIO_DEVICE_PCI_HOT_RESET:
diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index 38c8e9350a60ec..a390163ce706c4 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -1296,7 +1296,14 @@ static long vfio_device_fops_unl_ioctl(struct file *filep,
 		ret = vfio_ioctl_device_feature(device, uptr);
 		break;
 
+	case VFIO_DEVICE_GET_REGION_INFO:
+		if (!device->ops->get_region_info)
+			goto ioctl_fallback;
+		ret = device->ops->get_region_info(device, uptr);
+		break;
+
 	default:
+ioctl_fallback:
 		if (unlikely(!device->ops->ioctl))
 			ret = -EINVAL;
 		else
diff --git a/include/linux/vfio.h b/include/linux/vfio.h
index eb563f538dee51..be5fcf8432e8d5 100644
--- a/include/linux/vfio.h
+++ b/include/linux/vfio.h
@@ -132,6 +132,8 @@ struct vfio_device_ops {
 			 size_t count, loff_t *size);
 	long	(*ioctl)(struct vfio_device *vdev, unsigned int cmd,
 			 unsigned long arg);
+	int	(*get_region_info)(struct vfio_device *vdev,
+				   struct vfio_region_info __user *arg);
 	int	(*mmap)(struct vfio_device *vdev, struct vm_area_struct *vma);
 	void	(*request)(struct vfio_device *vdev, unsigned int count);
 	int	(*match)(struct vfio_device *vdev, char *buf);
diff --git a/include/linux/vfio_pci_core.h b/include/linux/vfio_pci_core.h
index f541044e42a2ad..160bc2e31ece75 100644
--- a/include/linux/vfio_pci_core.h
+++ b/include/linux/vfio_pci_core.h
@@ -115,6 +115,8 @@ long vfio_pci_core_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
 		unsigned long arg);
 int vfio_pci_core_ioctl_feature(struct vfio_device *device, u32 flags,
 				void __user *arg, size_t argsz);
+int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
+				   struct vfio_region_info __user *arg);
 ssize_t vfio_pci_core_read(struct vfio_device *core_vdev, char __user *buf,
 		size_t count, loff_t *ppos);
 ssize_t vfio_pci_core_write(struct vfio_device *core_vdev, const char __user *buf,
-- 
2.43.0

[PATCH v2 02/22] vfio/hisi: Convert to the get_region_info op

[Jason Gunthorpe]

Change the function signature of hisi_acc_vfio_pci_ioctl()
and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 .../vfio/pci/hisilicon/hisi_acc_vfio_pci.c    | 57 +++++++++----------
 1 file changed, 27 insertions(+), 30 deletions(-)

diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
index fde33f54e99ec5..899db4d742a010 100644
--- a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
+++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
@@ -1324,43 +1324,39 @@ static ssize_t hisi_acc_vfio_pci_read(struct vfio_device *core_vdev,
 	return vfio_pci_core_read(core_vdev, buf, new_count, ppos);
 }
 
-static long hisi_acc_vfio_pci_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
-				    unsigned long arg)
+static int hisi_acc_vfio_ioctl_get_region(struct vfio_device *core_vdev,
+					  struct vfio_region_info __user *arg)
 {
-	if (cmd == VFIO_DEVICE_GET_REGION_INFO) {
-		struct vfio_pci_core_device *vdev =
-			container_of(core_vdev, struct vfio_pci_core_device, vdev);
-		struct pci_dev *pdev = vdev->pdev;
-		struct vfio_region_info info;
-		unsigned long minsz;
+	struct vfio_pci_core_device *vdev =
+		container_of(core_vdev, struct vfio_pci_core_device, vdev);
+	struct pci_dev *pdev = vdev->pdev;
+	struct vfio_region_info info;
+	unsigned long minsz;
 
-		minsz = offsetofend(struct vfio_region_info, offset);
+	minsz = offsetofend(struct vfio_region_info, offset);
 
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
 
-		if (info.argsz < minsz)
-			return -EINVAL;
+	if (info.argsz < minsz)
+		return -EINVAL;
 
-		if (info.index == VFIO_PCI_BAR2_REGION_INDEX) {
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+	if (info.index != VFIO_PCI_BAR2_REGION_INDEX)
+		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
 
-			/*
-			 * ACC VF dev BAR2 region consists of both functional
-			 * register space and migration control register space.
-			 * Report only the functional region to Guest.
-			 */
-			info.size = pci_resource_len(pdev, info.index) / 2;
+	info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
 
-			info.flags = VFIO_REGION_INFO_FLAG_READ |
-					VFIO_REGION_INFO_FLAG_WRITE |
-					VFIO_REGION_INFO_FLAG_MMAP;
+	/*
+	 * ACC VF dev BAR2 region consists of both functional
+	 * register space and migration control register space.
+	 * Report only the functional region to Guest.
+	 */
+	info.size = pci_resource_len(pdev, info.index) / 2;
 
-			return copy_to_user((void __user *)arg, &info, minsz) ?
-					    -EFAULT : 0;
-		}
-	}
-	return vfio_pci_core_ioctl(core_vdev, cmd, arg);
+	info.flags = VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE |
+		     VFIO_REGION_INFO_FLAG_MMAP;
+
+	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 }
 
 static int hisi_acc_vf_debug_check(struct seq_file *seq, struct vfio_device *vdev)
@@ -1557,7 +1553,8 @@ static const struct vfio_device_ops hisi_acc_vfio_pci_migrn_ops = {
 	.release = vfio_pci_core_release_dev,
 	.open_device = hisi_acc_vfio_pci_open_device,
 	.close_device = hisi_acc_vfio_pci_close_device,
-	.ioctl = hisi_acc_vfio_pci_ioctl,
+	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = hisi_acc_vfio_ioctl_get_region,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = hisi_acc_vfio_pci_read,
 	.write = hisi_acc_vfio_pci_write,
-- 
2.43.0

[PATCH v2 03/22] vfio/virtio: Convert to the get_region_info op

[Jason Gunthorpe]

Remove virtiovf_vfio_pci_core_ioctl() and change the signature of
virtiovf_pci_ioctl_get_region_info().

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/pci/virtio/common.h    |  4 +---
 drivers/vfio/pci/virtio/legacy_io.c | 20 ++++----------------
 drivers/vfio/pci/virtio/main.c      |  3 ++-
 3 files changed, 7 insertions(+), 20 deletions(-)

diff --git a/drivers/vfio/pci/virtio/common.h b/drivers/vfio/pci/virtio/common.h
index c7d7e27af386e9..a10f2d92cb6238 100644
--- a/drivers/vfio/pci/virtio/common.h
+++ b/drivers/vfio/pci/virtio/common.h
@@ -109,10 +109,8 @@ void virtiovf_migration_reset_done(struct pci_dev *pdev);
 
 #ifdef CONFIG_VIRTIO_VFIO_PCI_ADMIN_LEGACY
 int virtiovf_open_legacy_io(struct virtiovf_pci_core_device *virtvdev);
-long virtiovf_vfio_pci_core_ioctl(struct vfio_device *core_vdev,
-				  unsigned int cmd, unsigned long arg);
 int virtiovf_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				       unsigned int cmd, unsigned long arg);
+				       struct vfio_region_info __user *arg);
 ssize_t virtiovf_pci_core_write(struct vfio_device *core_vdev,
 				const char __user *buf, size_t count,
 				loff_t *ppos);
diff --git a/drivers/vfio/pci/virtio/legacy_io.c b/drivers/vfio/pci/virtio/legacy_io.c
index 832af5ba267c49..d735d5c4bd7775 100644
--- a/drivers/vfio/pci/virtio/legacy_io.c
+++ b/drivers/vfio/pci/virtio/legacy_io.c
@@ -281,15 +281,14 @@ ssize_t virtiovf_pci_core_write(struct vfio_device *core_vdev, const char __user
 }
 
 int virtiovf_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				       unsigned int cmd, unsigned long arg)
+				       struct vfio_region_info __user *arg)
 {
 	struct virtiovf_pci_core_device *virtvdev = container_of(
 		core_vdev, struct virtiovf_pci_core_device, core_device.vdev);
 	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
-	void __user *uarg = (void __user *)arg;
 	struct vfio_region_info info = {};
 
-	if (copy_from_user(&info, uarg, minsz))
+	if (copy_from_user(&info, arg, minsz))
 		return -EFAULT;
 
 	if (info.argsz < minsz)
@@ -301,20 +300,9 @@ int virtiovf_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
 		info.size = virtvdev->bar0_virtual_buf_size;
 		info.flags = VFIO_REGION_INFO_FLAG_READ |
 			     VFIO_REGION_INFO_FLAG_WRITE;
-		return copy_to_user(uarg, &info, minsz) ? -EFAULT : 0;
+		return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 	default:
-		return vfio_pci_core_ioctl(core_vdev, cmd, arg);
-	}
-}
-
-long virtiovf_vfio_pci_core_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
-				  unsigned long arg)
-{
-	switch (cmd) {
-	case VFIO_DEVICE_GET_REGION_INFO:
-		return virtiovf_pci_ioctl_get_region_info(core_vdev, cmd, arg);
-	default:
-		return vfio_pci_core_ioctl(core_vdev, cmd, arg);
+		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
 	}
 }
 
diff --git a/drivers/vfio/pci/virtio/main.c b/drivers/vfio/pci/virtio/main.c
index 8084f3e36a9f70..92b525e52abe4d 100644
--- a/drivers/vfio/pci/virtio/main.c
+++ b/drivers/vfio/pci/virtio/main.c
@@ -108,7 +108,8 @@ static const struct vfio_device_ops virtiovf_vfio_pci_tran_lm_ops = {
 	.release = virtiovf_pci_core_release_dev,
 	.open_device = virtiovf_pci_open_device,
 	.close_device = virtiovf_pci_close_device,
-	.ioctl = virtiovf_vfio_pci_core_ioctl,
+	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = virtiovf_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = virtiovf_pci_core_read,
 	.write = virtiovf_pci_core_write,
-- 
2.43.0

[PATCH v2 04/22] vfio/nvgrace: Convert to the get_region_info op

[Jason Gunthorpe]

Change the signature of nvgrace_gpu_ioctl_get_region_info()

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Ankit Agrawal <ankita@nvidia.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/pci/nvgrace-gpu/main.c | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

diff --git a/drivers/vfio/pci/nvgrace-gpu/main.c b/drivers/vfio/pci/nvgrace-gpu/main.c
index e346392b72f6ab..d3a5253473e002 100644
--- a/drivers/vfio/pci/nvgrace-gpu/main.c
+++ b/drivers/vfio/pci/nvgrace-gpu/main.c
@@ -205,9 +205,9 @@ static int nvgrace_gpu_mmap(struct vfio_device *core_vdev,
 	return 0;
 }
 
-static long
+static int
 nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
-				  unsigned long arg)
+				  struct vfio_region_info __user *arg)
 {
 	struct nvgrace_gpu_pci_core_device *nvdev =
 		container_of(core_vdev, struct nvgrace_gpu_pci_core_device,
@@ -220,7 +220,7 @@ nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
 	u32 size;
 	int ret;
 
-	if (copy_from_user(&info, (void __user *)arg, minsz))
+	if (copy_from_user(&info, arg, minsz))
 		return -EFAULT;
 
 	if (info.argsz < minsz)
@@ -232,8 +232,7 @@ nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
 	 */
 	memregion = nvgrace_gpu_memregion(info.index, nvdev);
 	if (!memregion)
-		return vfio_pci_core_ioctl(core_vdev,
-					   VFIO_DEVICE_GET_REGION_INFO, arg);
+		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
 
 	size = struct_size(sparse, areas, 1);
 
@@ -285,16 +284,13 @@ nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
 		}
 		kfree(caps.buf);
 	}
-	return copy_to_user((void __user *)arg, &info, minsz) ?
-			    -EFAULT : 0;
+	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 }
 
 static long nvgrace_gpu_ioctl(struct vfio_device *core_vdev,
 			      unsigned int cmd, unsigned long arg)
 {
 	switch (cmd) {
-	case VFIO_DEVICE_GET_REGION_INFO:
-		return nvgrace_gpu_ioctl_get_region_info(core_vdev, arg);
 	case VFIO_DEVICE_IOEVENTFD:
 		return -ENOTTY;
 	case VFIO_DEVICE_RESET:
@@ -690,6 +686,7 @@ static const struct vfio_device_ops nvgrace_gpu_pci_ops = {
 	.open_device	= nvgrace_gpu_open_device,
 	.close_device	= nvgrace_gpu_close_device,
 	.ioctl		= nvgrace_gpu_ioctl,
+	.get_region_info = nvgrace_gpu_ioctl_get_region_info,
 	.device_feature	= vfio_pci_core_ioctl_feature,
 	.read		= nvgrace_gpu_read,
 	.write		= nvgrace_gpu_write,
-- 
2.43.0

[PATCH v2 05/22] vfio/pci: Fill in the missing get_region_info ops

[Jason Gunthorpe]

Now that every variant driver provides a get_region_info op remove the
ioctl based dispatch from vfio_pci_core_ioctl().

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c | 1 +
 drivers/vfio/pci/mlx5/main.c                   | 1 +
 drivers/vfio/pci/nvgrace-gpu/main.c            | 1 +
 drivers/vfio/pci/pds/vfio_dev.c                | 1 +
 drivers/vfio/pci/qat/main.c                    | 1 +
 drivers/vfio/pci/vfio_pci.c                    | 1 +
 drivers/vfio/pci/vfio_pci_core.c               | 2 --
 drivers/vfio/pci/virtio/main.c                 | 2 ++
 8 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
index 899db4d742a010..559b12c3d67493 100644
--- a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
+++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
@@ -1574,6 +1574,7 @@ static const struct vfio_device_ops hisi_acc_vfio_pci_ops = {
 	.open_device = hisi_acc_vfio_pci_open_device,
 	.close_device = vfio_pci_core_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/mlx5/main.c b/drivers/vfio/pci/mlx5/main.c
index 7ec47e736a8e5a..b7f941f8047ea4 100644
--- a/drivers/vfio/pci/mlx5/main.c
+++ b/drivers/vfio/pci/mlx5/main.c
@@ -1366,6 +1366,7 @@ static const struct vfio_device_ops mlx5vf_pci_ops = {
 	.open_device = mlx5vf_pci_open_device,
 	.close_device = mlx5vf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/nvgrace-gpu/main.c b/drivers/vfio/pci/nvgrace-gpu/main.c
index d3a5253473e002..cab743a30dc35d 100644
--- a/drivers/vfio/pci/nvgrace-gpu/main.c
+++ b/drivers/vfio/pci/nvgrace-gpu/main.c
@@ -707,6 +707,7 @@ static const struct vfio_device_ops nvgrace_gpu_pci_core_ops = {
 	.open_device	= nvgrace_gpu_open_device,
 	.close_device	= vfio_pci_core_close_device,
 	.ioctl		= vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature	= vfio_pci_core_ioctl_feature,
 	.read		= vfio_pci_core_read,
 	.write		= vfio_pci_core_write,
diff --git a/drivers/vfio/pci/pds/vfio_dev.c b/drivers/vfio/pci/pds/vfio_dev.c
index f3ccb0008f6752..1946bc75d99b49 100644
--- a/drivers/vfio/pci/pds/vfio_dev.c
+++ b/drivers/vfio/pci/pds/vfio_dev.c
@@ -195,6 +195,7 @@ static const struct vfio_device_ops pds_vfio_ops = {
 	.open_device = pds_vfio_open_device,
 	.close_device = pds_vfio_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/qat/main.c b/drivers/vfio/pci/qat/main.c
index a19b68043eb2e7..8452d9c1d11d3a 100644
--- a/drivers/vfio/pci/qat/main.c
+++ b/drivers/vfio/pci/qat/main.c
@@ -609,6 +609,7 @@ static const struct vfio_device_ops qat_vf_pci_ops = {
 	.open_device = qat_vf_pci_open_device,
 	.close_device = qat_vf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
 	.mmap = vfio_pci_core_mmap,
diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index ac10f14417f2f3..2d9122efc10baa 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -132,6 +132,7 @@ static const struct vfio_device_ops vfio_pci_ops = {
 	.open_device	= vfio_pci_open_device,
 	.close_device	= vfio_pci_core_close_device,
 	.ioctl		= vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read		= vfio_pci_core_read,
 	.write		= vfio_pci_core_write,
diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
index 1dc350003f075c..f21d9026068c37 100644
--- a/drivers/vfio/pci/vfio_pci_core.c
+++ b/drivers/vfio/pci/vfio_pci_core.c
@@ -1460,8 +1460,6 @@ long vfio_pci_core_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
 		return vfio_pci_ioctl_get_irq_info(vdev, uarg);
 	case VFIO_DEVICE_GET_PCI_HOT_RESET_INFO:
 		return vfio_pci_ioctl_get_pci_hot_reset_info(vdev, uarg);
-	case VFIO_DEVICE_GET_REGION_INFO:
-		return vfio_pci_ioctl_get_region_info(core_vdev, uarg);
 	case VFIO_DEVICE_IOEVENTFD:
 		return vfio_pci_ioctl_ioeventfd(vdev, uarg);
 	case VFIO_DEVICE_PCI_HOT_RESET:
diff --git a/drivers/vfio/pci/virtio/main.c b/drivers/vfio/pci/virtio/main.c
index 92b525e52abe4d..d68096bc525215 100644
--- a/drivers/vfio/pci/virtio/main.c
+++ b/drivers/vfio/pci/virtio/main.c
@@ -88,6 +88,7 @@ static const struct vfio_device_ops virtiovf_vfio_pci_lm_ops = {
 	.open_device = virtiovf_pci_open_device,
 	.close_device = virtiovf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
@@ -131,6 +132,7 @@ static const struct vfio_device_ops virtiovf_vfio_pci_ops = {
 	.open_device = virtiovf_pci_open_device,
 	.close_device = vfio_pci_core_close_device,
 	.ioctl = vfio_pci_core_ioctl,
+	.get_region_info = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
-- 
2.43.0

[PATCH v2 06/22] vfio/mtty: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of mtty_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 samples/vfio-mdev/mtty.c | 53 ++++++++++++++++++++++------------------
 1 file changed, 29 insertions(+), 24 deletions(-)

diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index 59eefe2fed10ca..b27f9b93471bc5 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -1785,6 +1785,34 @@ static int mtty_get_device_info(struct vfio_device_info *dev_info)
 	return 0;
 }
 
+static int mtty_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info __user *arg)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct vfio_region_info info;
+	void *cap_type = NULL;
+	u16 cap_type_id = 0;
+	unsigned long minsz;
+	int ret;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	ret = mtty_get_region_info(mdev_state, &info, &cap_type_id, &cap_type);
+	if (ret)
+		return ret;
+
+	if (copy_to_user(arg, &info, minsz))
+		return -EFAULT;
+	return 0;
+}
+
 static long mtty_ioctl(struct vfio_device *vdev, unsigned int cmd,
 			unsigned long arg)
 {
@@ -1817,30 +1845,6 @@ static long mtty_ioctl(struct vfio_device *vdev, unsigned int cmd,
 
 		return 0;
 	}
-	case VFIO_DEVICE_GET_REGION_INFO:
-	{
-		struct vfio_region_info info;
-		u16 cap_type_id = 0;
-		void *cap_type = NULL;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		ret = mtty_get_region_info(mdev_state, &info, &cap_type_id,
-					   &cap_type);
-		if (ret)
-			return ret;
-
-		if (copy_to_user((void __user *)arg, &info, minsz))
-			return -EFAULT;
-
-		return 0;
-	}
 
 	case VFIO_DEVICE_GET_IRQ_INFO:
 	{
@@ -1949,6 +1953,7 @@ static const struct vfio_device_ops mtty_dev_ops = {
 	.read = mtty_read,
 	.write = mtty_write,
 	.ioctl = mtty_ioctl,
+	.get_region_info = mtty_ioctl_get_region_info,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
 	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
-- 
2.43.0

[PATCH v2 07/22] vfio/mdpy: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of mdpy_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 samples/vfio-mdev/mdpy.c | 53 ++++++++++++++++++++++------------------
 1 file changed, 29 insertions(+), 24 deletions(-)

diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 8104831ae125be..0c65ed22173862 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -512,6 +512,34 @@ static int mdpy_query_gfx_plane(struct mdev_state *mdev_state,
 	return 0;
 }
 
+static int mdpy_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info __user *arg)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct vfio_region_info info;
+	void *cap_type = NULL;
+	u16 cap_type_id = 0;
+	unsigned long minsz;
+	int ret;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	ret = mdpy_get_region_info(mdev_state, &info, &cap_type_id, &cap_type);
+	if (ret)
+		return ret;
+
+	if (copy_to_user(arg, &info, minsz))
+		return -EFAULT;
+	return 0;
+}
+
 static long mdpy_ioctl(struct vfio_device *vdev, unsigned int cmd,
 		       unsigned long arg)
 {
@@ -544,30 +572,6 @@ static long mdpy_ioctl(struct vfio_device *vdev, unsigned int cmd,
 
 		return 0;
 	}
-	case VFIO_DEVICE_GET_REGION_INFO:
-	{
-		struct vfio_region_info info;
-		u16 cap_type_id = 0;
-		void *cap_type = NULL;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		ret = mdpy_get_region_info(mdev_state, &info, &cap_type_id,
-					   &cap_type);
-		if (ret)
-			return ret;
-
-		if (copy_to_user((void __user *)arg, &info, minsz))
-			return -EFAULT;
-
-		return 0;
-	}
 
 	case VFIO_DEVICE_GET_IRQ_INFO:
 	{
@@ -665,6 +669,7 @@ static const struct vfio_device_ops mdpy_dev_ops = {
 	.read = mdpy_read,
 	.write = mdpy_write,
 	.ioctl = mdpy_ioctl,
+	.get_region_info = mdpy_ioctl_get_region_info,
 	.mmap = mdpy_mmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
-- 
2.43.0

[PATCH v2 08/22] vfio/mbochs: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of mbochs_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 samples/vfio-mdev/mbochs.c | 56 +++++++++++++++++++++-----------------
 1 file changed, 31 insertions(+), 25 deletions(-)

diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index 18623ba666e33f..7f889b31fa2ce2 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -1185,13 +1185,42 @@ static int mbochs_get_gfx_dmabuf(struct mdev_state *mdev_state, u32 id)
 	return dma_buf_fd(dmabuf->buf, 0);
 }
 
+static int mbochs_ioctl_get_region_info(struct vfio_device *vdev,
+					struct vfio_region_info __user *arg)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct vfio_region_info_ext info;
+	unsigned long minsz, outsz;
+	int ret;
+
+	minsz = offsetofend(typeof(info), base.offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	outsz = info.base.argsz;
+	if (outsz < minsz)
+		return -EINVAL;
+	if (outsz > sizeof(info))
+		return -EINVAL;
+
+	ret = mbochs_get_region_info(mdev_state, &info);
+	if (ret)
+		return ret;
+
+	if (copy_to_user(arg, &info, outsz))
+		return -EFAULT;
+	return 0;
+}
+
 static long mbochs_ioctl(struct vfio_device *vdev, unsigned int cmd,
 			 unsigned long arg)
 {
 	struct mdev_state *mdev_state =
 		container_of(vdev, struct mdev_state, vdev);
 	int ret = 0;
-	unsigned long minsz, outsz;
+	unsigned long minsz;
 
 	switch (cmd) {
 	case VFIO_DEVICE_GET_INFO:
@@ -1215,30 +1244,6 @@ static long mbochs_ioctl(struct vfio_device *vdev, unsigned int cmd,
 
 		return 0;
 	}
-	case VFIO_DEVICE_GET_REGION_INFO:
-	{
-		struct vfio_region_info_ext info;
-
-		minsz = offsetofend(typeof(info), base.offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		outsz = info.base.argsz;
-		if (outsz < minsz)
-			return -EINVAL;
-		if (outsz > sizeof(info))
-			return -EINVAL;
-
-		ret = mbochs_get_region_info(mdev_state, &info);
-		if (ret)
-			return ret;
-
-		if (copy_to_user((void __user *)arg, &info, outsz))
-			return -EFAULT;
-
-		return 0;
-	}
 
 	case VFIO_DEVICE_GET_IRQ_INFO:
 	{
@@ -1376,6 +1381,7 @@ static const struct vfio_device_ops mbochs_dev_ops = {
 	.read = mbochs_read,
 	.write = mbochs_write,
 	.ioctl = mbochs_ioctl,
+	.get_region_info = mbochs_ioctl_get_region_info,
 	.mmap = mbochs_mmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
-- 
2.43.0

[PATCH v2 09/22] vfio/platform: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of vfio_platform_ioctl() and re-indent it. Add it to all
platform drivers.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/platform/vfio_amba.c             |  1 +
 drivers/vfio/platform/vfio_platform.c         |  1 +
 drivers/vfio/platform/vfio_platform_common.c  | 50 +++++++++++--------
 drivers/vfio/platform/vfio_platform_private.h |  2 +
 4 files changed, 32 insertions(+), 22 deletions(-)

diff --git a/drivers/vfio/platform/vfio_amba.c b/drivers/vfio/platform/vfio_amba.c
index 9f5c527baa8a36..d600deaf23b6d7 100644
--- a/drivers/vfio/platform/vfio_amba.c
+++ b/drivers/vfio/platform/vfio_amba.c
@@ -115,6 +115,7 @@ static const struct vfio_device_ops vfio_amba_ops = {
 	.open_device	= vfio_platform_open_device,
 	.close_device	= vfio_platform_close_device,
 	.ioctl		= vfio_platform_ioctl,
+	.get_region_info = vfio_platform_ioctl_get_region_info,
 	.read		= vfio_platform_read,
 	.write		= vfio_platform_write,
 	.mmap		= vfio_platform_mmap,
diff --git a/drivers/vfio/platform/vfio_platform.c b/drivers/vfio/platform/vfio_platform.c
index 512533501eb7f3..0e85c914b65105 100644
--- a/drivers/vfio/platform/vfio_platform.c
+++ b/drivers/vfio/platform/vfio_platform.c
@@ -101,6 +101,7 @@ static const struct vfio_device_ops vfio_platform_ops = {
 	.open_device	= vfio_platform_open_device,
 	.close_device	= vfio_platform_close_device,
 	.ioctl		= vfio_platform_ioctl,
+	.get_region_info = vfio_platform_ioctl_get_region_info,
 	.read		= vfio_platform_read,
 	.write		= vfio_platform_write,
 	.mmap		= vfio_platform_mmap,
diff --git a/drivers/vfio/platform/vfio_platform_common.c b/drivers/vfio/platform/vfio_platform_common.c
index 3bf1043cd7957c..3ebd50fb78fbb7 100644
--- a/drivers/vfio/platform/vfio_platform_common.c
+++ b/drivers/vfio/platform/vfio_platform_common.c
@@ -272,6 +272,34 @@ int vfio_platform_open_device(struct vfio_device *core_vdev)
 }
 EXPORT_SYMBOL_GPL(vfio_platform_open_device);
 
+int vfio_platform_ioctl_get_region_info(struct vfio_device *core_vdev,
+					struct vfio_region_info __user *arg)
+{
+	struct vfio_platform_device *vdev =
+		container_of(core_vdev, struct vfio_platform_device, vdev);
+	struct vfio_region_info info;
+	unsigned long minsz;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	if (info.index >= vdev->num_regions)
+		return -EINVAL;
+
+	/* map offset to the physical address  */
+	info.offset = VFIO_PLATFORM_INDEX_TO_OFFSET(info.index);
+	info.size = vdev->regions[info.index].size;
+	info.flags = vdev->regions[info.index].flags;
+
+	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+}
+EXPORT_SYMBOL_GPL(vfio_platform_ioctl_get_region_info);
+
 long vfio_platform_ioctl(struct vfio_device *core_vdev,
 			 unsigned int cmd, unsigned long arg)
 {
@@ -300,28 +328,6 @@ long vfio_platform_ioctl(struct vfio_device *core_vdev,
 		return copy_to_user((void __user *)arg, &info, minsz) ?
 			-EFAULT : 0;
 
-	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO) {
-		struct vfio_region_info info;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		if (info.index >= vdev->num_regions)
-			return -EINVAL;
-
-		/* map offset to the physical address  */
-		info.offset = VFIO_PLATFORM_INDEX_TO_OFFSET(info.index);
-		info.size = vdev->regions[info.index].size;
-		info.flags = vdev->regions[info.index].flags;
-
-		return copy_to_user((void __user *)arg, &info, minsz) ?
-			-EFAULT : 0;
-
 	} else if (cmd == VFIO_DEVICE_GET_IRQ_INFO) {
 		struct vfio_irq_info info;
 
diff --git a/drivers/vfio/platform/vfio_platform_private.h b/drivers/vfio/platform/vfio_platform_private.h
index 8d8fab51684909..a6008320e77bae 100644
--- a/drivers/vfio/platform/vfio_platform_private.h
+++ b/drivers/vfio/platform/vfio_platform_private.h
@@ -85,6 +85,8 @@ int vfio_platform_open_device(struct vfio_device *core_vdev);
 void vfio_platform_close_device(struct vfio_device *core_vdev);
 long vfio_platform_ioctl(struct vfio_device *core_vdev,
 			 unsigned int cmd, unsigned long arg);
+int vfio_platform_ioctl_get_region_info(struct vfio_device *core_vdev,
+					struct vfio_region_info __user *arg);
 ssize_t vfio_platform_read(struct vfio_device *core_vdev,
 			   char __user *buf, size_t count,
 			   loff_t *ppos);
-- 
2.43.0

[PATCH v2 10/22] vfio/fsl: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of vfio_fsl_mc_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/fsl-mc/vfio_fsl_mc.c | 56 ++++++++++++++++++-------------
 1 file changed, 32 insertions(+), 24 deletions(-)

diff --git a/drivers/vfio/fsl-mc/vfio_fsl_mc.c b/drivers/vfio/fsl-mc/vfio_fsl_mc.c
index 76ccbab0e3d642..d38e51a57f0775 100644
--- a/drivers/vfio/fsl-mc/vfio_fsl_mc.c
+++ b/drivers/vfio/fsl-mc/vfio_fsl_mc.c
@@ -117,6 +117,37 @@ static void vfio_fsl_mc_close_device(struct vfio_device *core_vdev)
 	fsl_mc_cleanup_irq_pool(mc_cont);
 }
 
+static int
+vfio_fsl_mc_ioctl_get_region_info(struct vfio_device *core_vdev,
+				  struct vfio_region_info __user *arg)
+{
+	struct vfio_fsl_mc_device *vdev =
+		container_of(core_vdev, struct vfio_fsl_mc_device, vdev);
+	struct fsl_mc_device *mc_dev = vdev->mc_dev;
+	struct vfio_region_info info;
+	unsigned long minsz;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	if (info.index >= mc_dev->obj_desc.region_count)
+		return -EINVAL;
+
+	/* map offset to the physical address  */
+	info.offset = VFIO_FSL_MC_INDEX_TO_OFFSET(info.index);
+	info.size = vdev->regions[info.index].size;
+	info.flags = vdev->regions[info.index].flags;
+
+	if (copy_to_user(arg, &info, minsz))
+		return -EFAULT;
+	return 0;
+}
+
 static long vfio_fsl_mc_ioctl(struct vfio_device *core_vdev,
 			      unsigned int cmd, unsigned long arg)
 {
@@ -149,30 +180,6 @@ static long vfio_fsl_mc_ioctl(struct vfio_device *core_vdev,
 		return copy_to_user((void __user *)arg, &info, minsz) ?
 			-EFAULT : 0;
 	}
-	case VFIO_DEVICE_GET_REGION_INFO:
-	{
-		struct vfio_region_info info;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		if (info.index >= mc_dev->obj_desc.region_count)
-			return -EINVAL;
-
-		/* map offset to the physical address  */
-		info.offset = VFIO_FSL_MC_INDEX_TO_OFFSET(info.index);
-		info.size = vdev->regions[info.index].size;
-		info.flags = vdev->regions[info.index].flags;
-
-		if (copy_to_user((void __user *)arg, &info, minsz))
-			return -EFAULT;
-		return 0;
-	}
 	case VFIO_DEVICE_GET_IRQ_INFO:
 	{
 		struct vfio_irq_info info;
@@ -589,6 +596,7 @@ static const struct vfio_device_ops vfio_fsl_mc_ops = {
 	.open_device	= vfio_fsl_mc_open_device,
 	.close_device	= vfio_fsl_mc_close_device,
 	.ioctl		= vfio_fsl_mc_ioctl,
+	.get_region_info = vfio_fsl_mc_ioctl_get_region_info,
 	.read		= vfio_fsl_mc_read,
 	.write		= vfio_fsl_mc_write,
 	.mmap		= vfio_fsl_mc_mmap,
-- 
2.43.0

[PATCH v2 11/22] vfio/cdx: Provide a get_region_info op

[Jason Gunthorpe]

Change the signature of vfio_cdx_ioctl_get_region_info() and hook it to
the op.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/cdx/main.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/vfio/cdx/main.c b/drivers/vfio/cdx/main.c
index 5dd5f5ad768653..506d849139d35a 100644
--- a/drivers/vfio/cdx/main.c
+++ b/drivers/vfio/cdx/main.c
@@ -129,9 +129,11 @@ static int vfio_cdx_ioctl_get_info(struct vfio_cdx_device *vdev,
 	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
 }
 
-static int vfio_cdx_ioctl_get_region_info(struct vfio_cdx_device *vdev,
+static int vfio_cdx_ioctl_get_region_info(struct vfio_device *core_vdev,
 					  struct vfio_region_info __user *arg)
 {
+	struct vfio_cdx_device *vdev =
+		container_of(core_vdev, struct vfio_cdx_device, vdev);
 	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
 	struct cdx_device *cdx_dev = to_cdx_device(vdev->vdev.dev);
 	struct vfio_region_info info;
@@ -219,8 +221,6 @@ static long vfio_cdx_ioctl(struct vfio_device *core_vdev,
 	switch (cmd) {
 	case VFIO_DEVICE_GET_INFO:
 		return vfio_cdx_ioctl_get_info(vdev, uarg);
-	case VFIO_DEVICE_GET_REGION_INFO:
-		return vfio_cdx_ioctl_get_region_info(vdev, uarg);
 	case VFIO_DEVICE_GET_IRQ_INFO:
 		return vfio_cdx_ioctl_get_irq_info(vdev, uarg);
 	case VFIO_DEVICE_SET_IRQS:
@@ -284,6 +284,7 @@ static const struct vfio_device_ops vfio_cdx_ops = {
 	.open_device	= vfio_cdx_open_device,
 	.close_device	= vfio_cdx_close_device,
 	.ioctl		= vfio_cdx_ioctl,
+	.get_region_info = vfio_cdx_ioctl_get_region_info,
 	.device_feature = vfio_cdx_ioctl_feature,
 	.mmap		= vfio_cdx_mmap,
 	.bind_iommufd	= vfio_iommufd_physical_bind,
-- 
2.43.0

[PATCH v2 12/22] vfio/ccw: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of vfio_ccw_mdev_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/s390/cio/vfio_ccw_ops.c | 44 +++++++++++++++++++--------------
 1 file changed, 26 insertions(+), 18 deletions(-)

diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index ea532a8a4a0c29..6d46e0bc76df15 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -504,6 +504,31 @@ void vfio_ccw_unregister_dev_regions(struct vfio_ccw_private *private)
 	private->region = NULL;
 }
 
+static int
+vfio_ccw_mdev_ioctl_get_region_info(struct vfio_device *vdev,
+				    struct vfio_region_info __user *arg)
+{
+	struct vfio_ccw_private *private =
+		container_of(vdev, struct vfio_ccw_private, vdev);
+	struct vfio_region_info info;
+	unsigned long minsz;
+	int ret;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	ret = vfio_ccw_mdev_get_region_info(private, &info, arg);
+	if (ret)
+		return ret;
+
+	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+}
+
 static ssize_t vfio_ccw_mdev_ioctl(struct vfio_device *vdev,
 				   unsigned int cmd,
 				   unsigned long arg)
@@ -532,24 +557,6 @@ static ssize_t vfio_ccw_mdev_ioctl(struct vfio_device *vdev,
 
 		return copy_to_user((void __user *)arg, &info, minsz) ? -EFAULT : 0;
 	}
-	case VFIO_DEVICE_GET_REGION_INFO:
-	{
-		struct vfio_region_info info;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		ret = vfio_ccw_mdev_get_region_info(private, &info, arg);
-		if (ret)
-			return ret;
-
-		return copy_to_user((void __user *)arg, &info, minsz) ? -EFAULT : 0;
-	}
 	case VFIO_DEVICE_GET_IRQ_INFO:
 	{
 		struct vfio_irq_info info;
@@ -627,6 +634,7 @@ static const struct vfio_device_ops vfio_ccw_dev_ops = {
 	.read = vfio_ccw_mdev_read,
 	.write = vfio_ccw_mdev_write,
 	.ioctl = vfio_ccw_mdev_ioctl,
+	.get_region_info = vfio_ccw_mdev_ioctl_get_region_info,
 	.request = vfio_ccw_mdev_request,
 	.dma_unmap = vfio_ccw_dma_unmap,
 	.bind_iommufd = vfio_iommufd_emulated_bind,
-- 
2.43.0

[PATCH v2 13/22] vfio/gvt: Provide a get_region_info op

[Jason Gunthorpe]

Move it out of intel_vgpu_ioctl() and re-indent it.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/gpu/drm/i915/gvt/kvmgt.c | 301 +++++++++++++++----------------
 1 file changed, 150 insertions(+), 151 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 69830a5c49d3fd..1feb2a28ca5fda 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -1140,6 +1140,155 @@ static int intel_vgpu_set_irqs(struct intel_vgpu *vgpu, u32 flags,
 	return func(vgpu, index, start, count, flags, data);
 }
 
+static int intel_vgpu_ioctl_get_region_info(struct vfio_device *vfio_dev,
+					    struct vfio_region_info __user *arg)
+{
+	struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
+	struct vfio_region_info_cap_sparse_mmap *sparse = NULL;
+	struct intel_vgpu *vgpu = vfio_dev_to_vgpu(vfio_dev);
+	struct vfio_region_info info;
+	unsigned long minsz;
+	int nr_areas = 1;
+	int cap_type_id;
+	unsigned int i;
+	int ret;
+
+	minsz = offsetofend(struct vfio_region_info, offset);
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	switch (info.index) {
+	case VFIO_PCI_CONFIG_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = vgpu->gvt->device_info.cfg_space_size;
+		info.flags = VFIO_REGION_INFO_FLAG_READ |
+			     VFIO_REGION_INFO_FLAG_WRITE;
+		break;
+	case VFIO_PCI_BAR0_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = vgpu->cfg_space.bar[info.index].size;
+		if (!info.size) {
+			info.flags = 0;
+			break;
+		}
+
+		info.flags = VFIO_REGION_INFO_FLAG_READ |
+			     VFIO_REGION_INFO_FLAG_WRITE;
+		break;
+	case VFIO_PCI_BAR1_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = 0;
+		info.flags = 0;
+		break;
+	case VFIO_PCI_BAR2_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.flags = VFIO_REGION_INFO_FLAG_CAPS |
+			     VFIO_REGION_INFO_FLAG_MMAP |
+			     VFIO_REGION_INFO_FLAG_READ |
+			     VFIO_REGION_INFO_FLAG_WRITE;
+		info.size = gvt_aperture_sz(vgpu->gvt);
+
+		sparse = kzalloc(struct_size(sparse, areas, nr_areas),
+				 GFP_KERNEL);
+		if (!sparse)
+			return -ENOMEM;
+
+		sparse->header.id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
+		sparse->header.version = 1;
+		sparse->nr_areas = nr_areas;
+		cap_type_id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
+		sparse->areas[0].offset =
+			PAGE_ALIGN(vgpu_aperture_offset(vgpu));
+		sparse->areas[0].size = vgpu_aperture_sz(vgpu);
+		break;
+
+	case VFIO_PCI_BAR3_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = 0;
+		info.flags = 0;
+
+		gvt_dbg_core("get region info bar:%d\n", info.index);
+		break;
+
+	case VFIO_PCI_ROM_REGION_INDEX:
+	case VFIO_PCI_VGA_REGION_INDEX:
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = 0;
+		info.flags = 0;
+
+		gvt_dbg_core("get region info index:%d\n", info.index);
+		break;
+	default: {
+		struct vfio_region_info_cap_type cap_type = {
+			.header.id = VFIO_REGION_INFO_CAP_TYPE,
+			.header.version = 1
+		};
+
+		if (info.index >= VFIO_PCI_NUM_REGIONS + vgpu->num_regions)
+			return -EINVAL;
+		info.index = array_index_nospec(
+			info.index, VFIO_PCI_NUM_REGIONS + vgpu->num_regions);
+
+		i = info.index - VFIO_PCI_NUM_REGIONS;
+
+		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+		info.size = vgpu->region[i].size;
+		info.flags = vgpu->region[i].flags;
+
+		cap_type.type = vgpu->region[i].type;
+		cap_type.subtype = vgpu->region[i].subtype;
+
+		ret = vfio_info_add_capability(&caps, &cap_type.header,
+					       sizeof(cap_type));
+		if (ret)
+			return ret;
+	}
+	}
+
+	if ((info.flags & VFIO_REGION_INFO_FLAG_CAPS) && sparse) {
+		switch (cap_type_id) {
+		case VFIO_REGION_INFO_CAP_SPARSE_MMAP:
+			ret = vfio_info_add_capability(
+				&caps, &sparse->header,
+				struct_size(sparse, areas, sparse->nr_areas));
+			if (ret) {
+				kfree(sparse);
+				return ret;
+			}
+			break;
+		default:
+			kfree(sparse);
+			return -EINVAL;
+		}
+	}
+
+	if (caps.size) {
+		info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
+		if (info.argsz < sizeof(info) + caps.size) {
+			info.argsz = sizeof(info) + caps.size;
+			info.cap_offset = 0;
+		} else {
+			vfio_info_cap_shift(&caps, sizeof(info));
+			if (copy_to_user((void __user *)arg + sizeof(info),
+					 caps.buf, caps.size)) {
+				kfree(caps.buf);
+				kfree(sparse);
+				return -EFAULT;
+			}
+			info.cap_offset = sizeof(info);
+		}
+
+		kfree(caps.buf);
+	}
+
+	kfree(sparse);
+	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+}
+
 static long intel_vgpu_ioctl(struct vfio_device *vfio_dev, unsigned int cmd,
 			     unsigned long arg)
 {
@@ -1168,157 +1317,6 @@ static long intel_vgpu_ioctl(struct vfio_device *vfio_dev, unsigned int cmd,
 		return copy_to_user((void __user *)arg, &info, minsz) ?
 			-EFAULT : 0;
 
-	} else if (cmd == VFIO_DEVICE_GET_REGION_INFO) {
-		struct vfio_region_info info;
-		struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
-		unsigned int i;
-		int ret;
-		struct vfio_region_info_cap_sparse_mmap *sparse = NULL;
-		int nr_areas = 1;
-		int cap_type_id;
-
-		minsz = offsetofend(struct vfio_region_info, offset);
-
-		if (copy_from_user(&info, (void __user *)arg, minsz))
-			return -EFAULT;
-
-		if (info.argsz < minsz)
-			return -EINVAL;
-
-		switch (info.index) {
-		case VFIO_PCI_CONFIG_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = vgpu->gvt->device_info.cfg_space_size;
-			info.flags = VFIO_REGION_INFO_FLAG_READ |
-				     VFIO_REGION_INFO_FLAG_WRITE;
-			break;
-		case VFIO_PCI_BAR0_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = vgpu->cfg_space.bar[info.index].size;
-			if (!info.size) {
-				info.flags = 0;
-				break;
-			}
-
-			info.flags = VFIO_REGION_INFO_FLAG_READ |
-				     VFIO_REGION_INFO_FLAG_WRITE;
-			break;
-		case VFIO_PCI_BAR1_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = 0;
-			info.flags = 0;
-			break;
-		case VFIO_PCI_BAR2_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.flags = VFIO_REGION_INFO_FLAG_CAPS |
-					VFIO_REGION_INFO_FLAG_MMAP |
-					VFIO_REGION_INFO_FLAG_READ |
-					VFIO_REGION_INFO_FLAG_WRITE;
-			info.size = gvt_aperture_sz(vgpu->gvt);
-
-			sparse = kzalloc(struct_size(sparse, areas, nr_areas),
-					 GFP_KERNEL);
-			if (!sparse)
-				return -ENOMEM;
-
-			sparse->header.id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
-			sparse->header.version = 1;
-			sparse->nr_areas = nr_areas;
-			cap_type_id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
-			sparse->areas[0].offset =
-					PAGE_ALIGN(vgpu_aperture_offset(vgpu));
-			sparse->areas[0].size = vgpu_aperture_sz(vgpu);
-			break;
-
-		case VFIO_PCI_BAR3_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = 0;
-			info.flags = 0;
-
-			gvt_dbg_core("get region info bar:%d\n", info.index);
-			break;
-
-		case VFIO_PCI_ROM_REGION_INDEX:
-		case VFIO_PCI_VGA_REGION_INDEX:
-			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = 0;
-			info.flags = 0;
-
-			gvt_dbg_core("get region info index:%d\n", info.index);
-			break;
-		default:
-			{
-				struct vfio_region_info_cap_type cap_type = {
-					.header.id = VFIO_REGION_INFO_CAP_TYPE,
-					.header.version = 1 };
-
-				if (info.index >= VFIO_PCI_NUM_REGIONS +
-						vgpu->num_regions)
-					return -EINVAL;
-				info.index =
-					array_index_nospec(info.index,
-							VFIO_PCI_NUM_REGIONS +
-							vgpu->num_regions);
-
-				i = info.index - VFIO_PCI_NUM_REGIONS;
-
-				info.offset =
-					VFIO_PCI_INDEX_TO_OFFSET(info.index);
-				info.size = vgpu->region[i].size;
-				info.flags = vgpu->region[i].flags;
-
-				cap_type.type = vgpu->region[i].type;
-				cap_type.subtype = vgpu->region[i].subtype;
-
-				ret = vfio_info_add_capability(&caps,
-							&cap_type.header,
-							sizeof(cap_type));
-				if (ret)
-					return ret;
-			}
-		}
-
-		if ((info.flags & VFIO_REGION_INFO_FLAG_CAPS) && sparse) {
-			switch (cap_type_id) {
-			case VFIO_REGION_INFO_CAP_SPARSE_MMAP:
-				ret = vfio_info_add_capability(&caps,
-					&sparse->header,
-					struct_size(sparse, areas,
-						    sparse->nr_areas));
-				if (ret) {
-					kfree(sparse);
-					return ret;
-				}
-				break;
-			default:
-				kfree(sparse);
-				return -EINVAL;
-			}
-		}
-
-		if (caps.size) {
-			info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
-			if (info.argsz < sizeof(info) + caps.size) {
-				info.argsz = sizeof(info) + caps.size;
-				info.cap_offset = 0;
-			} else {
-				vfio_info_cap_shift(&caps, sizeof(info));
-				if (copy_to_user((void __user *)arg +
-						  sizeof(info), caps.buf,
-						  caps.size)) {
-					kfree(caps.buf);
-					kfree(sparse);
-					return -EFAULT;
-				}
-				info.cap_offset = sizeof(info);
-			}
-
-			kfree(caps.buf);
-		}
-
-		kfree(sparse);
-		return copy_to_user((void __user *)arg, &info, minsz) ?
-			-EFAULT : 0;
 	} else if (cmd == VFIO_DEVICE_GET_IRQ_INFO) {
 		struct vfio_irq_info info;
 
@@ -1475,6 +1473,7 @@ static const struct vfio_device_ops intel_vgpu_dev_ops = {
 	.write		= intel_vgpu_write,
 	.mmap		= intel_vgpu_mmap,
 	.ioctl		= intel_vgpu_ioctl,
+	.get_region_info = intel_vgpu_ioctl_get_region_info,
 	.dma_unmap	= intel_vgpu_dma_unmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd = vfio_iommufd_emulated_unbind,
-- 
2.43.0

[PATCH v2 14/22] vfio: Require drivers to implement get_region_info

[Jason Gunthorpe]

Remove the fallback through the ioctl callback, no drivers use this now.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/vfio_main.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index a390163ce706c4..f056e82ba35075 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -1297,13 +1297,13 @@ static long vfio_device_fops_unl_ioctl(struct file *filep,
 		break;
 
 	case VFIO_DEVICE_GET_REGION_INFO:
-		if (!device->ops->get_region_info)
-			goto ioctl_fallback;
-		ret = device->ops->get_region_info(device, uptr);
+		if (unlikely(!device->ops->get_region_info))
+			ret = -EINVAL;
+		else
+			ret = device->ops->get_region_info(device, uptr);
 		break;
 
 	default:
-ioctl_fallback:
 		if (unlikely(!device->ops->ioctl))
 			ret = -EINVAL;
 		else
-- 
2.43.0

[PATCH v2 15/22] vfio: Add get_region_info_caps op

[Jason Gunthorpe]

This op does the copy to/from user for the info and can return back
a cap chain through a vfio_info_cap * result.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/vfio_main.c | 56 +++++++++++++++++++++++++++++++++++++---
 include/linux/vfio.h     |  4 +++
 2 files changed, 56 insertions(+), 4 deletions(-)

diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index f056e82ba35075..48d034aede46fc 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -1259,6 +1259,57 @@ static int vfio_ioctl_device_feature(struct vfio_device *device,
 	}
 }
 
+static long vfio_get_region_info(struct vfio_device *device,
+				 struct vfio_region_info __user *arg)
+{
+	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
+	struct vfio_region_info info = {};
+	struct vfio_info_cap caps = {};
+	int ret;
+
+	if (copy_from_user(&info, arg, minsz))
+		return -EFAULT;
+	if (info.argsz < minsz)
+		return -EINVAL;
+
+	if (device->ops->get_region_info_caps) {
+		ret = device->ops->get_region_info_caps(device, &info, &caps);
+		if (ret)
+			goto out_free;
+
+		if (caps.size) {
+			info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
+			if (info.argsz < sizeof(info) + caps.size) {
+				info.argsz = sizeof(info) + caps.size;
+				info.cap_offset = 0;
+			} else {
+				vfio_info_cap_shift(&caps, sizeof(info));
+				if (copy_to_user(arg + 1, caps.buf,
+						 caps.size)) {
+					ret = -EFAULT;
+					goto out_free;
+				}
+				info.cap_offset = sizeof(info);
+			}
+		}
+
+		if (copy_to_user(arg, &info, minsz)) {
+			ret = -EFAULT;
+			goto out_free;
+		}
+	} else if (device->ops->get_region_info) {
+		ret = device->ops->get_region_info(device, arg);
+		if (ret)
+			return ret;
+	} else {
+		return -EINVAL;
+	}
+
+out_free:
+	kfree(caps.buf);
+	return ret;
+}
+
 static long vfio_device_fops_unl_ioctl(struct file *filep,
 				       unsigned int cmd, unsigned long arg)
 {
@@ -1297,10 +1348,7 @@ static long vfio_device_fops_unl_ioctl(struct file *filep,
 		break;
 
 	case VFIO_DEVICE_GET_REGION_INFO:
-		if (unlikely(!device->ops->get_region_info))
-			ret = -EINVAL;
-		else
-			ret = device->ops->get_region_info(device, uptr);
+		ret = vfio_get_region_info(device, uptr);
 		break;
 
 	default:
diff --git a/include/linux/vfio.h b/include/linux/vfio.h
index be5fcf8432e8d5..6311ddc837701d 100644
--- a/include/linux/vfio.h
+++ b/include/linux/vfio.h
@@ -21,6 +21,7 @@ struct kvm;
 struct iommufd_ctx;
 struct iommufd_device;
 struct iommufd_access;
+struct vfio_info_cap;
 
 /*
  * VFIO devices can be placed in a set, this allows all devices to share this
@@ -134,6 +135,9 @@ struct vfio_device_ops {
 			 unsigned long arg);
 	int	(*get_region_info)(struct vfio_device *vdev,
 				   struct vfio_region_info __user *arg);
+	int	(*get_region_info_caps)(struct vfio_device *vdev,
+					struct vfio_region_info *info,
+					struct vfio_info_cap *caps);
 	int	(*mmap)(struct vfio_device *vdev, struct vm_area_struct *vma);
 	void	(*request)(struct vfio_device *vdev, unsigned int count);
 	int	(*match)(struct vfio_device *vdev, char *buf);
-- 
2.43.0

[PATCH v2 16/22] vfio/mbochs: Convert mbochs to use vfio_info_add_capability()

[Jason Gunthorpe]

This driver open codes the cap chain manipulations. Instead use
vfio_info_add_capability() and the get_region_info_caps() op.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 samples/vfio-mdev/mbochs.c | 75 ++++++++++++--------------------------
 1 file changed, 23 insertions(+), 52 deletions(-)

diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index 7f889b31fa2ce2..64ea19253ee3ad 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -143,11 +143,6 @@ static struct mdev_parent mbochs_parent;
 static atomic_t mbochs_avail_mbytes;
 static const struct vfio_device_ops mbochs_dev_ops;
 
-struct vfio_region_info_ext {
-	struct vfio_region_info          base;
-	struct vfio_region_info_cap_type type;
-};
-
 struct mbochs_mode {
 	u32 drm_format;
 	u32 bytepp;
@@ -1033,10 +1028,12 @@ static int mbochs_dmabuf_export(struct mbochs_dmabuf *dmabuf)
 	return 0;
 }
 
-static int mbochs_get_region_info(struct mdev_state *mdev_state,
-				  struct vfio_region_info_ext *ext)
+static int mbochs_ioctl_get_region_info(struct vfio_device *vdev,
+					struct vfio_region_info *region_info,
+					struct vfio_info_cap *caps)
 {
-	struct vfio_region_info *region_info = &ext->base;
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
 
 	if (region_info->index >= MBOCHS_NUM_REGIONS)
 		return -EINVAL;
@@ -1061,20 +1058,23 @@ static int mbochs_get_region_info(struct mdev_state *mdev_state,
 		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ  |
 				       VFIO_REGION_INFO_FLAG_WRITE);
 		break;
-	case MBOCHS_EDID_REGION_INDEX:
-		ext->base.argsz = sizeof(*ext);
-		ext->base.offset = MBOCHS_EDID_OFFSET;
-		ext->base.size = MBOCHS_EDID_SIZE;
-		ext->base.flags = (VFIO_REGION_INFO_FLAG_READ  |
-				   VFIO_REGION_INFO_FLAG_WRITE |
-				   VFIO_REGION_INFO_FLAG_CAPS);
-		ext->base.cap_offset = offsetof(typeof(*ext), type);
-		ext->type.header.id = VFIO_REGION_INFO_CAP_TYPE;
-		ext->type.header.version = 1;
-		ext->type.header.next = 0;
-		ext->type.type = VFIO_REGION_TYPE_GFX;
-		ext->type.subtype = VFIO_REGION_SUBTYPE_GFX_EDID;
-		break;
+	case MBOCHS_EDID_REGION_INDEX: {
+		struct vfio_region_info_cap_type cap_type = {
+			.header.id = VFIO_REGION_INFO_CAP_TYPE,
+			.header.version = 1,
+			.type = VFIO_REGION_TYPE_GFX,
+			.subtype = VFIO_REGION_SUBTYPE_GFX_EDID,
+		};
+
+		region_info->offset = MBOCHS_EDID_OFFSET;
+		region_info->size = MBOCHS_EDID_SIZE;
+		region_info->flags = (VFIO_REGION_INFO_FLAG_READ |
+				      VFIO_REGION_INFO_FLAG_WRITE |
+				      VFIO_REGION_INFO_FLAG_CAPS);
+
+		return vfio_info_add_capability(caps, &cap_type.header,
+						sizeof(cap_type));
+	}
 	default:
 		region_info->size   = 0;
 		region_info->offset = 0;
@@ -1185,35 +1185,6 @@ static int mbochs_get_gfx_dmabuf(struct mdev_state *mdev_state, u32 id)
 	return dma_buf_fd(dmabuf->buf, 0);
 }
 
-static int mbochs_ioctl_get_region_info(struct vfio_device *vdev,
-					struct vfio_region_info __user *arg)
-{
-	struct mdev_state *mdev_state =
-		container_of(vdev, struct mdev_state, vdev);
-	struct vfio_region_info_ext info;
-	unsigned long minsz, outsz;
-	int ret;
-
-	minsz = offsetofend(typeof(info), base.offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	outsz = info.base.argsz;
-	if (outsz < minsz)
-		return -EINVAL;
-	if (outsz > sizeof(info))
-		return -EINVAL;
-
-	ret = mbochs_get_region_info(mdev_state, &info);
-	if (ret)
-		return ret;
-
-	if (copy_to_user(arg, &info, outsz))
-		return -EFAULT;
-	return 0;
-}
-
 static long mbochs_ioctl(struct vfio_device *vdev, unsigned int cmd,
 			 unsigned long arg)
 {
@@ -1381,7 +1352,7 @@ static const struct vfio_device_ops mbochs_dev_ops = {
 	.read = mbochs_read,
 	.write = mbochs_write,
 	.ioctl = mbochs_ioctl,
-	.get_region_info = mbochs_ioctl_get_region_info,
+	.get_region_info_caps = mbochs_ioctl_get_region_info,
 	.mmap = mbochs_mmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
-- 
2.43.0

[PATCH v2 17/22] vfio/gvt: Convert to get_region_info_caps

[Jason Gunthorpe]

Remove the duplicate code and change info to a pointer.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/gpu/drm/i915/gvt/kvmgt.c | 113 ++++++++++++-------------------
 1 file changed, 42 insertions(+), 71 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 1feb2a28ca5fda..96d23717684f7e 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -1141,56 +1141,46 @@ static int intel_vgpu_set_irqs(struct intel_vgpu *vgpu, u32 flags,
 }
 
 static int intel_vgpu_ioctl_get_region_info(struct vfio_device *vfio_dev,
-					    struct vfio_region_info __user *arg)
+					    struct vfio_region_info *info,
+					    struct vfio_info_cap *caps)
 {
-	struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
 	struct vfio_region_info_cap_sparse_mmap *sparse = NULL;
 	struct intel_vgpu *vgpu = vfio_dev_to_vgpu(vfio_dev);
-	struct vfio_region_info info;
-	unsigned long minsz;
 	int nr_areas = 1;
 	int cap_type_id;
 	unsigned int i;
 	int ret;
 
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	switch (info.index) {
+	switch (info->index) {
 	case VFIO_PCI_CONFIG_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = vgpu->gvt->device_info.cfg_space_size;
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = vgpu->gvt->device_info.cfg_space_size;
+		info->flags = VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
 		break;
 	case VFIO_PCI_BAR0_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = vgpu->cfg_space.bar[info.index].size;
-		if (!info.size) {
-			info.flags = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = vgpu->cfg_space.bar[info->index].size;
+		if (!info->size) {
+			info->flags = 0;
 			break;
 		}
 
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
+		info->flags = VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
 		break;
 	case VFIO_PCI_BAR1_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = 0;
-		info.flags = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = 0;
+		info->flags = 0;
 		break;
 	case VFIO_PCI_BAR2_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.flags = VFIO_REGION_INFO_FLAG_CAPS |
-			     VFIO_REGION_INFO_FLAG_MMAP |
-			     VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
-		info.size = gvt_aperture_sz(vgpu->gvt);
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->flags = VFIO_REGION_INFO_FLAG_CAPS |
+			      VFIO_REGION_INFO_FLAG_MMAP |
+			      VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
+		info->size = gvt_aperture_sz(vgpu->gvt);
 
 		sparse = kzalloc(struct_size(sparse, areas, nr_areas),
 				 GFP_KERNEL);
@@ -1207,20 +1197,20 @@ static int intel_vgpu_ioctl_get_region_info(struct vfio_device *vfio_dev,
 		break;
 
 	case VFIO_PCI_BAR3_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = 0;
-		info.flags = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = 0;
+		info->flags = 0;
 
-		gvt_dbg_core("get region info bar:%d\n", info.index);
+		gvt_dbg_core("get region info bar:%d\n", info->index);
 		break;
 
 	case VFIO_PCI_ROM_REGION_INDEX:
 	case VFIO_PCI_VGA_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = 0;
-		info.flags = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = 0;
+		info->flags = 0;
 
-		gvt_dbg_core("get region info index:%d\n", info.index);
+		gvt_dbg_core("get region info index:%d\n", info->index);
 		break;
 	default: {
 		struct vfio_region_info_cap_type cap_type = {
@@ -1228,32 +1218,32 @@ static int intel_vgpu_ioctl_get_region_info(struct vfio_device *vfio_dev,
 			.header.version = 1
 		};
 
-		if (info.index >= VFIO_PCI_NUM_REGIONS + vgpu->num_regions)
+		if (info->index >= VFIO_PCI_NUM_REGIONS + vgpu->num_regions)
 			return -EINVAL;
-		info.index = array_index_nospec(
-			info.index, VFIO_PCI_NUM_REGIONS + vgpu->num_regions);
+		info->index = array_index_nospec(
+			info->index, VFIO_PCI_NUM_REGIONS + vgpu->num_regions);
 
-		i = info.index - VFIO_PCI_NUM_REGIONS;
+		i = info->index - VFIO_PCI_NUM_REGIONS;
 
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = vgpu->region[i].size;
-		info.flags = vgpu->region[i].flags;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = vgpu->region[i].size;
+		info->flags = vgpu->region[i].flags;
 
 		cap_type.type = vgpu->region[i].type;
 		cap_type.subtype = vgpu->region[i].subtype;
 
-		ret = vfio_info_add_capability(&caps, &cap_type.header,
+		ret = vfio_info_add_capability(caps, &cap_type.header,
 					       sizeof(cap_type));
 		if (ret)
 			return ret;
 	}
 	}
 
-	if ((info.flags & VFIO_REGION_INFO_FLAG_CAPS) && sparse) {
+	if ((info->flags & VFIO_REGION_INFO_FLAG_CAPS) && sparse) {
 		switch (cap_type_id) {
 		case VFIO_REGION_INFO_CAP_SPARSE_MMAP:
 			ret = vfio_info_add_capability(
-				&caps, &sparse->header,
+				caps, &sparse->header,
 				struct_size(sparse, areas, sparse->nr_areas));
 			if (ret) {
 				kfree(sparse);
@@ -1266,27 +1256,8 @@ static int intel_vgpu_ioctl_get_region_info(struct vfio_device *vfio_dev,
 		}
 	}
 
-	if (caps.size) {
-		info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
-		if (info.argsz < sizeof(info) + caps.size) {
-			info.argsz = sizeof(info) + caps.size;
-			info.cap_offset = 0;
-		} else {
-			vfio_info_cap_shift(&caps, sizeof(info));
-			if (copy_to_user((void __user *)arg + sizeof(info),
-					 caps.buf, caps.size)) {
-				kfree(caps.buf);
-				kfree(sparse);
-				return -EFAULT;
-			}
-			info.cap_offset = sizeof(info);
-		}
-
-		kfree(caps.buf);
-	}
-
 	kfree(sparse);
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	return 0;
 }
 
 static long intel_vgpu_ioctl(struct vfio_device *vfio_dev, unsigned int cmd,
@@ -1473,7 +1444,7 @@ static const struct vfio_device_ops intel_vgpu_dev_ops = {
 	.write		= intel_vgpu_write,
 	.mmap		= intel_vgpu_mmap,
 	.ioctl		= intel_vgpu_ioctl,
-	.get_region_info = intel_vgpu_ioctl_get_region_info,
+	.get_region_info_caps = intel_vgpu_ioctl_get_region_info,
 	.dma_unmap	= intel_vgpu_dma_unmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd = vfio_iommufd_emulated_unbind,
-- 
2.43.0

[PATCH v2 18/22] vfio/ccw: Convert to get_region_info_caps

[Jason Gunthorpe]

Remove the duplicate code and flatten the call chain.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/s390/cio/vfio_ccw_ops.c | 55 +++++----------------------------
 1 file changed, 7 insertions(+), 48 deletions(-)

diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index 6d46e0bc76df15..a596f6013019cb 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -313,10 +313,12 @@ static int vfio_ccw_mdev_get_device_info(struct vfio_ccw_private *private,
 	return 0;
 }
 
-static int vfio_ccw_mdev_get_region_info(struct vfio_ccw_private *private,
-					 struct vfio_region_info *info,
-					 unsigned long arg)
+static int vfio_ccw_mdev_ioctl_get_region_info(struct vfio_device *vdev,
+					       struct vfio_region_info *info,
+					       struct vfio_info_cap *caps)
 {
+	struct vfio_ccw_private *private =
+		container_of(vdev, struct vfio_ccw_private, vdev);
 	int i;
 
 	switch (info->index) {
@@ -328,7 +330,6 @@ static int vfio_ccw_mdev_get_region_info(struct vfio_ccw_private *private,
 		return 0;
 	default: /* all other regions are handled via capability chain */
 	{
-		struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
 		struct vfio_region_info_cap_type cap_type = {
 			.header.id = VFIO_REGION_INFO_CAP_TYPE,
 			.header.version = 1 };
@@ -351,27 +352,10 @@ static int vfio_ccw_mdev_get_region_info(struct vfio_ccw_private *private,
 		cap_type.type = private->region[i].type;
 		cap_type.subtype = private->region[i].subtype;
 
-		ret = vfio_info_add_capability(&caps, &cap_type.header,
+		ret = vfio_info_add_capability(caps, &cap_type.header,
 					       sizeof(cap_type));
 		if (ret)
 			return ret;
-
-		info->flags |= VFIO_REGION_INFO_FLAG_CAPS;
-		if (info->argsz < sizeof(*info) + caps.size) {
-			info->argsz = sizeof(*info) + caps.size;
-			info->cap_offset = 0;
-		} else {
-			vfio_info_cap_shift(&caps, sizeof(*info));
-			if (copy_to_user((void __user *)arg + sizeof(*info),
-					 caps.buf, caps.size)) {
-				kfree(caps.buf);
-				return -EFAULT;
-			}
-			info->cap_offset = sizeof(*info);
-		}
-
-		kfree(caps.buf);
-
 	}
 	}
 	return 0;
@@ -504,31 +488,6 @@ void vfio_ccw_unregister_dev_regions(struct vfio_ccw_private *private)
 	private->region = NULL;
 }
 
-static int
-vfio_ccw_mdev_ioctl_get_region_info(struct vfio_device *vdev,
-				    struct vfio_region_info __user *arg)
-{
-	struct vfio_ccw_private *private =
-		container_of(vdev, struct vfio_ccw_private, vdev);
-	struct vfio_region_info info;
-	unsigned long minsz;
-	int ret;
-
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	ret = vfio_ccw_mdev_get_region_info(private, &info, arg);
-	if (ret)
-		return ret;
-
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
-}
-
 static ssize_t vfio_ccw_mdev_ioctl(struct vfio_device *vdev,
 				   unsigned int cmd,
 				   unsigned long arg)
@@ -634,7 +593,7 @@ static const struct vfio_device_ops vfio_ccw_dev_ops = {
 	.read = vfio_ccw_mdev_read,
 	.write = vfio_ccw_mdev_write,
 	.ioctl = vfio_ccw_mdev_ioctl,
-	.get_region_info = vfio_ccw_mdev_ioctl_get_region_info,
+	.get_region_info_caps = vfio_ccw_mdev_ioctl_get_region_info,
 	.request = vfio_ccw_mdev_request,
 	.dma_unmap = vfio_ccw_dma_unmap,
 	.bind_iommufd = vfio_iommufd_emulated_bind,
-- 
2.43.0

[PATCH v2 19/22] vfio/pci: Convert all PCI drivers to get_region_info_caps

[Jason Gunthorpe]

Since the core function signature changes it has to flow up to all
drivers.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 .../vfio/pci/hisilicon/hisi_acc_vfio_pci.c    |  30 ++---
 drivers/vfio/pci/mlx5/main.c                  |   2 +-
 drivers/vfio/pci/nvgrace-gpu/main.c           |  51 ++-------
 drivers/vfio/pci/pds/vfio_dev.c               |   2 +-
 drivers/vfio/pci/qat/main.c                   |   2 +-
 drivers/vfio/pci/vfio_pci.c                   |   2 +-
 drivers/vfio/pci/vfio_pci_core.c              | 103 +++++++-----------
 drivers/vfio/pci/virtio/common.h              |   3 +-
 drivers/vfio/pci/virtio/legacy_io.c           |  26 ++---
 drivers/vfio/pci/virtio/main.c                |   6 +-
 include/linux/vfio_pci_core.h                 |   3 +-
 11 files changed, 80 insertions(+), 150 deletions(-)

diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
index 559b12c3d67493..8915eb4943e46a 100644
--- a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
+++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
@@ -1325,38 +1325,28 @@ static ssize_t hisi_acc_vfio_pci_read(struct vfio_device *core_vdev,
 }
 
 static int hisi_acc_vfio_ioctl_get_region(struct vfio_device *core_vdev,
-					  struct vfio_region_info __user *arg)
+					  struct vfio_region_info *info,
+					  struct vfio_info_cap *caps)
 {
 	struct vfio_pci_core_device *vdev =
 		container_of(core_vdev, struct vfio_pci_core_device, vdev);
 	struct pci_dev *pdev = vdev->pdev;
-	struct vfio_region_info info;
-	unsigned long minsz;
 
-	minsz = offsetofend(struct vfio_region_info, offset);
+	if (info->index != VFIO_PCI_BAR2_REGION_INDEX)
+		return vfio_pci_ioctl_get_region_info(core_vdev, info, caps);
 
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	if (info.index != VFIO_PCI_BAR2_REGION_INDEX)
-		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
-
-	info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+	info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
 
 	/*
 	 * ACC VF dev BAR2 region consists of both functional
 	 * register space and migration control register space.
 	 * Report only the functional region to Guest.
 	 */
-	info.size = pci_resource_len(pdev, info.index) / 2;
+	info->size = pci_resource_len(pdev, info->index) / 2;
 
-	info.flags = VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE |
+	info->flags = VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE |
 		     VFIO_REGION_INFO_FLAG_MMAP;
-
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	return 0;
 }
 
 static int hisi_acc_vf_debug_check(struct seq_file *seq, struct vfio_device *vdev)
@@ -1554,7 +1544,7 @@ static const struct vfio_device_ops hisi_acc_vfio_pci_migrn_ops = {
 	.open_device = hisi_acc_vfio_pci_open_device,
 	.close_device = hisi_acc_vfio_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = hisi_acc_vfio_ioctl_get_region,
+	.get_region_info_caps = hisi_acc_vfio_ioctl_get_region,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = hisi_acc_vfio_pci_read,
 	.write = hisi_acc_vfio_pci_write,
@@ -1574,7 +1564,7 @@ static const struct vfio_device_ops hisi_acc_vfio_pci_ops = {
 	.open_device = hisi_acc_vfio_pci_open_device,
 	.close_device = vfio_pci_core_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/mlx5/main.c b/drivers/vfio/pci/mlx5/main.c
index b7f941f8047ea4..9c5970411d07a1 100644
--- a/drivers/vfio/pci/mlx5/main.c
+++ b/drivers/vfio/pci/mlx5/main.c
@@ -1366,7 +1366,7 @@ static const struct vfio_device_ops mlx5vf_pci_ops = {
 	.open_device = mlx5vf_pci_open_device,
 	.close_device = mlx5vf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/nvgrace-gpu/main.c b/drivers/vfio/pci/nvgrace-gpu/main.c
index cab743a30dc35d..5a6f77d5c81e0f 100644
--- a/drivers/vfio/pci/nvgrace-gpu/main.c
+++ b/drivers/vfio/pci/nvgrace-gpu/main.c
@@ -205,34 +205,25 @@ static int nvgrace_gpu_mmap(struct vfio_device *core_vdev,
 	return 0;
 }
 
-static int
-nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
-				  struct vfio_region_info __user *arg)
+static int nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
+					     struct vfio_region_info *info,
+					     struct vfio_info_cap *caps)
 {
 	struct nvgrace_gpu_pci_core_device *nvdev =
 		container_of(core_vdev, struct nvgrace_gpu_pci_core_device,
 			     core_device.vdev);
-	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
-	struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
 	struct vfio_region_info_cap_sparse_mmap *sparse;
-	struct vfio_region_info info;
 	struct mem_region *memregion;
 	u32 size;
 	int ret;
 
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
 	/*
 	 * Request to determine the BAR region information. Send the
 	 * GPU memory information.
 	 */
-	memregion = nvgrace_gpu_memregion(info.index, nvdev);
+	memregion = nvgrace_gpu_memregion(info->index, nvdev);
 	if (!memregion)
-		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
+		return vfio_pci_ioctl_get_region_info(core_vdev, info, caps);
 
 	size = struct_size(sparse, areas, 1);
 
@@ -251,40 +242,22 @@ nvgrace_gpu_ioctl_get_region_info(struct vfio_device *core_vdev,
 	sparse->header.id = VFIO_REGION_INFO_CAP_SPARSE_MMAP;
 	sparse->header.version = 1;
 
-	ret = vfio_info_add_capability(&caps, &sparse->header, size);
+	ret = vfio_info_add_capability(caps, &sparse->header, size);
 	kfree(sparse);
 	if (ret)
 		return ret;
 
-	info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
+	info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
 	/*
 	 * The region memory size may not be power-of-2 aligned.
 	 * Given that the memory is a BAR and may not be
 	 * aligned, roundup to the next power-of-2.
 	 */
-	info.size = memregion->bar_size;
-	info.flags = VFIO_REGION_INFO_FLAG_READ |
+	info->size = memregion->bar_size;
+	info->flags = VFIO_REGION_INFO_FLAG_READ |
 		     VFIO_REGION_INFO_FLAG_WRITE |
 		     VFIO_REGION_INFO_FLAG_MMAP;
-
-	if (caps.size) {
-		info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
-		if (info.argsz < sizeof(info) + caps.size) {
-			info.argsz = sizeof(info) + caps.size;
-			info.cap_offset = 0;
-		} else {
-			vfio_info_cap_shift(&caps, sizeof(info));
-			if (copy_to_user((void __user *)arg +
-					 sizeof(info), caps.buf,
-					 caps.size)) {
-				kfree(caps.buf);
-				return -EFAULT;
-			}
-			info.cap_offset = sizeof(info);
-		}
-		kfree(caps.buf);
-	}
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	return 0;
 }
 
 static long nvgrace_gpu_ioctl(struct vfio_device *core_vdev,
@@ -686,7 +659,7 @@ static const struct vfio_device_ops nvgrace_gpu_pci_ops = {
 	.open_device	= nvgrace_gpu_open_device,
 	.close_device	= nvgrace_gpu_close_device,
 	.ioctl		= nvgrace_gpu_ioctl,
-	.get_region_info = nvgrace_gpu_ioctl_get_region_info,
+	.get_region_info_caps = nvgrace_gpu_ioctl_get_region_info,
 	.device_feature	= vfio_pci_core_ioctl_feature,
 	.read		= nvgrace_gpu_read,
 	.write		= nvgrace_gpu_write,
@@ -707,7 +680,7 @@ static const struct vfio_device_ops nvgrace_gpu_pci_core_ops = {
 	.open_device	= nvgrace_gpu_open_device,
 	.close_device	= vfio_pci_core_close_device,
 	.ioctl		= vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature	= vfio_pci_core_ioctl_feature,
 	.read		= vfio_pci_core_read,
 	.write		= vfio_pci_core_write,
diff --git a/drivers/vfio/pci/pds/vfio_dev.c b/drivers/vfio/pci/pds/vfio_dev.c
index 1946bc75d99b49..be103c74e96957 100644
--- a/drivers/vfio/pci/pds/vfio_dev.c
+++ b/drivers/vfio/pci/pds/vfio_dev.c
@@ -195,7 +195,7 @@ static const struct vfio_device_ops pds_vfio_ops = {
 	.open_device = pds_vfio_open_device,
 	.close_device = pds_vfio_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/drivers/vfio/pci/qat/main.c b/drivers/vfio/pci/qat/main.c
index 8452d9c1d11d3a..8fbdf7c6d666e1 100644
--- a/drivers/vfio/pci/qat/main.c
+++ b/drivers/vfio/pci/qat/main.c
@@ -609,7 +609,7 @@ static const struct vfio_device_ops qat_vf_pci_ops = {
 	.open_device = qat_vf_pci_open_device,
 	.close_device = qat_vf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
 	.mmap = vfio_pci_core_mmap,
diff --git a/drivers/vfio/pci/vfio_pci.c b/drivers/vfio/pci/vfio_pci.c
index 2d9122efc10baa..a3e49d42c771bc 100644
--- a/drivers/vfio/pci/vfio_pci.c
+++ b/drivers/vfio/pci/vfio_pci.c
@@ -132,7 +132,7 @@ static const struct vfio_device_ops vfio_pci_ops = {
 	.open_device	= vfio_pci_open_device,
 	.close_device	= vfio_pci_core_close_device,
 	.ioctl		= vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read		= vfio_pci_core_read,
 	.write		= vfio_pci_core_write,
diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
index f21d9026068c37..57c0766fb9f80f 100644
--- a/drivers/vfio/pci/vfio_pci_core.c
+++ b/drivers/vfio/pci/vfio_pci_core.c
@@ -997,43 +997,35 @@ static int vfio_pci_ioctl_get_info(struct vfio_pci_core_device *vdev,
 }
 
 int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				   struct vfio_region_info __user *arg)
+				   struct vfio_region_info *info,
+				   struct vfio_info_cap *caps)
 {
 	struct vfio_pci_core_device *vdev =
 		container_of(core_vdev, struct vfio_pci_core_device, vdev);
-	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
 	struct pci_dev *pdev = vdev->pdev;
-	struct vfio_region_info info;
-	struct vfio_info_cap caps = { .buf = NULL, .size = 0 };
 	int i, ret;
 
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	switch (info.index) {
+	switch (info->index) {
 	case VFIO_PCI_CONFIG_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = pdev->cfg_size;
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = pdev->cfg_size;
+		info->flags = VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
 		break;
 	case VFIO_PCI_BAR0_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = pci_resource_len(pdev, info.index);
-		if (!info.size) {
-			info.flags = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = pci_resource_len(pdev, info->index);
+		if (!info->size) {
+			info->flags = 0;
 			break;
 		}
 
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
-		if (vdev->bar_mmap_supported[info.index]) {
-			info.flags |= VFIO_REGION_INFO_FLAG_MMAP;
-			if (info.index == vdev->msix_bar) {
-				ret = msix_mmappable_cap(vdev, &caps);
+		info->flags = VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
+		if (vdev->bar_mmap_supported[info->index]) {
+			info->flags |= VFIO_REGION_INFO_FLAG_MMAP;
+			if (info->index == vdev->msix_bar) {
+				ret = msix_mmappable_cap(vdev, caps);
 				if (ret)
 					return ret;
 			}
@@ -1045,9 +1037,9 @@ int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
 		size_t size;
 		u16 cmd;
 
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.flags = 0;
-		info.size = 0;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->flags = 0;
+		info->size = 0;
 
 		if (pci_resource_start(pdev, PCI_ROM_RESOURCE)) {
 			/*
@@ -1057,16 +1049,17 @@ int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
 			cmd = vfio_pci_memory_lock_and_enable(vdev);
 			io = pci_map_rom(pdev, &size);
 			if (io) {
-				info.flags = VFIO_REGION_INFO_FLAG_READ;
+				info->flags = VFIO_REGION_INFO_FLAG_READ;
 				/* Report the BAR size, not the ROM size. */
-				info.size = pci_resource_len(pdev, PCI_ROM_RESOURCE);
+				info->size = pci_resource_len(pdev,
+							      PCI_ROM_RESOURCE);
 				pci_unmap_rom(pdev, io);
 			}
 			vfio_pci_memory_unlock_and_restore(vdev, cmd);
 		} else if (pdev->rom && pdev->romlen) {
-			info.flags = VFIO_REGION_INFO_FLAG_READ;
+			info->flags = VFIO_REGION_INFO_FLAG_READ;
 			/* Report BAR size as power of two. */
-			info.size = roundup_pow_of_two(pdev->romlen);
+			info->size = roundup_pow_of_two(pdev->romlen);
 		}
 
 		break;
@@ -1075,10 +1068,10 @@ int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
 		if (!vdev->has_vga)
 			return -EINVAL;
 
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = 0xc0000;
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = 0xc0000;
+		info->flags = VFIO_REGION_INFO_FLAG_READ |
+			      VFIO_REGION_INFO_FLAG_WRITE;
 
 		break;
 	default: {
@@ -1087,52 +1080,34 @@ int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
 			.header.version = 1
 		};
 
-		if (info.index >= VFIO_PCI_NUM_REGIONS + vdev->num_regions)
+		if (info->index >= VFIO_PCI_NUM_REGIONS + vdev->num_regions)
 			return -EINVAL;
-		info.index = array_index_nospec(
-			info.index, VFIO_PCI_NUM_REGIONS + vdev->num_regions);
+		info->index = array_index_nospec(
+			info->index, VFIO_PCI_NUM_REGIONS + vdev->num_regions);
 
-		i = info.index - VFIO_PCI_NUM_REGIONS;
+		i = info->index - VFIO_PCI_NUM_REGIONS;
 
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = vdev->region[i].size;
-		info.flags = vdev->region[i].flags;
+		info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+		info->size = vdev->region[i].size;
+		info->flags = vdev->region[i].flags;
 
 		cap_type.type = vdev->region[i].type;
 		cap_type.subtype = vdev->region[i].subtype;
 
-		ret = vfio_info_add_capability(&caps, &cap_type.header,
+		ret = vfio_info_add_capability(caps, &cap_type.header,
 					       sizeof(cap_type));
 		if (ret)
 			return ret;
 
 		if (vdev->region[i].ops->add_capability) {
 			ret = vdev->region[i].ops->add_capability(
-				vdev, &vdev->region[i], &caps);
+				vdev, &vdev->region[i], caps);
 			if (ret)
 				return ret;
 		}
 	}
 	}
-
-	if (caps.size) {
-		info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
-		if (info.argsz < sizeof(info) + caps.size) {
-			info.argsz = sizeof(info) + caps.size;
-			info.cap_offset = 0;
-		} else {
-			vfio_info_cap_shift(&caps, sizeof(info));
-			if (copy_to_user(arg + 1, caps.buf, caps.size)) {
-				kfree(caps.buf);
-				return -EFAULT;
-			}
-			info.cap_offset = sizeof(*arg);
-		}
-
-		kfree(caps.buf);
-	}
-
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(vfio_pci_ioctl_get_region_info);
 
diff --git a/drivers/vfio/pci/virtio/common.h b/drivers/vfio/pci/virtio/common.h
index a10f2d92cb6238..cb3d5e57d3a3e9 100644
--- a/drivers/vfio/pci/virtio/common.h
+++ b/drivers/vfio/pci/virtio/common.h
@@ -110,7 +110,8 @@ void virtiovf_migration_reset_done(struct pci_dev *pdev);
 #ifdef CONFIG_VIRTIO_VFIO_PCI_ADMIN_LEGACY
 int virtiovf_open_legacy_io(struct virtiovf_pci_core_device *virtvdev);
 int virtiovf_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				       struct vfio_region_info __user *arg);
+				       struct vfio_region_info *info,
+				       struct vfio_info_cap *caps);
 ssize_t virtiovf_pci_core_write(struct vfio_device *core_vdev,
 				const char __user *buf, size_t count,
 				loff_t *ppos);
diff --git a/drivers/vfio/pci/virtio/legacy_io.c b/drivers/vfio/pci/virtio/legacy_io.c
index d735d5c4bd7775..1ed349a556291b 100644
--- a/drivers/vfio/pci/virtio/legacy_io.c
+++ b/drivers/vfio/pci/virtio/legacy_io.c
@@ -281,29 +281,19 @@ ssize_t virtiovf_pci_core_write(struct vfio_device *core_vdev, const char __user
 }
 
 int virtiovf_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				       struct vfio_region_info __user *arg)
+				       struct vfio_region_info *info,
+				       struct vfio_info_cap *caps)
 {
 	struct virtiovf_pci_core_device *virtvdev = container_of(
 		core_vdev, struct virtiovf_pci_core_device, core_device.vdev);
-	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
-	struct vfio_region_info info = {};
 
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
+	if (info->index != VFIO_PCI_BAR0_REGION_INDEX)
+		return vfio_pci_ioctl_get_region_info(core_vdev, info, caps);
 
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	switch (info.index) {
-	case VFIO_PCI_BAR0_REGION_INDEX:
-		info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-		info.size = virtvdev->bar0_virtual_buf_size;
-		info.flags = VFIO_REGION_INFO_FLAG_READ |
-			     VFIO_REGION_INFO_FLAG_WRITE;
-		return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
-	default:
-		return vfio_pci_ioctl_get_region_info(core_vdev, arg);
-	}
+	info->offset = VFIO_PCI_INDEX_TO_OFFSET(info->index);
+	info->size = virtvdev->bar0_virtual_buf_size;
+	info->flags = VFIO_REGION_INFO_FLAG_READ | VFIO_REGION_INFO_FLAG_WRITE;
+	return 0;
 }
 
 static int virtiovf_set_notify_addr(struct virtiovf_pci_core_device *virtvdev)
diff --git a/drivers/vfio/pci/virtio/main.c b/drivers/vfio/pci/virtio/main.c
index d68096bc525215..d2e5cbca13c857 100644
--- a/drivers/vfio/pci/virtio/main.c
+++ b/drivers/vfio/pci/virtio/main.c
@@ -88,7 +88,7 @@ static const struct vfio_device_ops virtiovf_vfio_pci_lm_ops = {
 	.open_device = virtiovf_pci_open_device,
 	.close_device = virtiovf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
@@ -110,7 +110,7 @@ static const struct vfio_device_ops virtiovf_vfio_pci_tran_lm_ops = {
 	.open_device = virtiovf_pci_open_device,
 	.close_device = virtiovf_pci_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = virtiovf_pci_ioctl_get_region_info,
+	.get_region_info_caps = virtiovf_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = virtiovf_pci_core_read,
 	.write = virtiovf_pci_core_write,
@@ -132,7 +132,7 @@ static const struct vfio_device_ops virtiovf_vfio_pci_ops = {
 	.open_device = virtiovf_pci_open_device,
 	.close_device = vfio_pci_core_close_device,
 	.ioctl = vfio_pci_core_ioctl,
-	.get_region_info = vfio_pci_ioctl_get_region_info,
+	.get_region_info_caps = vfio_pci_ioctl_get_region_info,
 	.device_feature = vfio_pci_core_ioctl_feature,
 	.read = vfio_pci_core_read,
 	.write = vfio_pci_core_write,
diff --git a/include/linux/vfio_pci_core.h b/include/linux/vfio_pci_core.h
index 160bc2e31ece75..e74f94c17fbeb6 100644
--- a/include/linux/vfio_pci_core.h
+++ b/include/linux/vfio_pci_core.h
@@ -116,7 +116,8 @@ long vfio_pci_core_ioctl(struct vfio_device *core_vdev, unsigned int cmd,
 int vfio_pci_core_ioctl_feature(struct vfio_device *device, u32 flags,
 				void __user *arg, size_t argsz);
 int vfio_pci_ioctl_get_region_info(struct vfio_device *core_vdev,
-				   struct vfio_region_info __user *arg);
+				   struct vfio_region_info *info,
+				   struct vfio_info_cap *caps);
 ssize_t vfio_pci_core_read(struct vfio_device *core_vdev, char __user *buf,
 		size_t count, loff_t *ppos);
 ssize_t vfio_pci_core_write(struct vfio_device *core_vdev, const char __user *buf,
-- 
2.43.0

[PATCH v2 20/22] vfio/platform: Convert to get_region_info_caps

[Jason Gunthorpe]

Remove the duplicate code and change info to a pointer. caps are not used.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/platform/vfio_amba.c             |  2 +-
 drivers/vfio/platform/vfio_platform.c         |  2 +-
 drivers/vfio/platform/vfio_platform_common.c  | 24 ++++++-------------
 drivers/vfio/platform/vfio_platform_private.h |  3 ++-
 4 files changed, 11 insertions(+), 20 deletions(-)

diff --git a/drivers/vfio/platform/vfio_amba.c b/drivers/vfio/platform/vfio_amba.c
index d600deaf23b6d7..fa754f203b2dfc 100644
--- a/drivers/vfio/platform/vfio_amba.c
+++ b/drivers/vfio/platform/vfio_amba.c
@@ -115,7 +115,7 @@ static const struct vfio_device_ops vfio_amba_ops = {
 	.open_device	= vfio_platform_open_device,
 	.close_device	= vfio_platform_close_device,
 	.ioctl		= vfio_platform_ioctl,
-	.get_region_info = vfio_platform_ioctl_get_region_info,
+	.get_region_info_caps = vfio_platform_ioctl_get_region_info,
 	.read		= vfio_platform_read,
 	.write		= vfio_platform_write,
 	.mmap		= vfio_platform_mmap,
diff --git a/drivers/vfio/platform/vfio_platform.c b/drivers/vfio/platform/vfio_platform.c
index 0e85c914b65105..a4d3ace3e02dda 100644
--- a/drivers/vfio/platform/vfio_platform.c
+++ b/drivers/vfio/platform/vfio_platform.c
@@ -101,7 +101,7 @@ static const struct vfio_device_ops vfio_platform_ops = {
 	.open_device	= vfio_platform_open_device,
 	.close_device	= vfio_platform_close_device,
 	.ioctl		= vfio_platform_ioctl,
-	.get_region_info = vfio_platform_ioctl_get_region_info,
+	.get_region_info_caps = vfio_platform_ioctl_get_region_info,
 	.read		= vfio_platform_read,
 	.write		= vfio_platform_write,
 	.mmap		= vfio_platform_mmap,
diff --git a/drivers/vfio/platform/vfio_platform_common.c b/drivers/vfio/platform/vfio_platform_common.c
index 3ebd50fb78fbb7..c2990b7e900fa5 100644
--- a/drivers/vfio/platform/vfio_platform_common.c
+++ b/drivers/vfio/platform/vfio_platform_common.c
@@ -273,30 +273,20 @@ int vfio_platform_open_device(struct vfio_device *core_vdev)
 EXPORT_SYMBOL_GPL(vfio_platform_open_device);
 
 int vfio_platform_ioctl_get_region_info(struct vfio_device *core_vdev,
-					struct vfio_region_info __user *arg)
+					struct vfio_region_info *info,
+					struct vfio_info_cap *caps)
 {
 	struct vfio_platform_device *vdev =
 		container_of(core_vdev, struct vfio_platform_device, vdev);
-	struct vfio_region_info info;
-	unsigned long minsz;
 
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	if (info.index >= vdev->num_regions)
+	if (info->index >= vdev->num_regions)
 		return -EINVAL;
 
 	/* map offset to the physical address  */
-	info.offset = VFIO_PLATFORM_INDEX_TO_OFFSET(info.index);
-	info.size = vdev->regions[info.index].size;
-	info.flags = vdev->regions[info.index].flags;
-
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	info->offset = VFIO_PLATFORM_INDEX_TO_OFFSET(info->index);
+	info->size = vdev->regions[info->index].size;
+	info->flags = vdev->regions[info->index].flags;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(vfio_platform_ioctl_get_region_info);
 
diff --git a/drivers/vfio/platform/vfio_platform_private.h b/drivers/vfio/platform/vfio_platform_private.h
index a6008320e77bae..05084212a76eb6 100644
--- a/drivers/vfio/platform/vfio_platform_private.h
+++ b/drivers/vfio/platform/vfio_platform_private.h
@@ -86,7 +86,8 @@ void vfio_platform_close_device(struct vfio_device *core_vdev);
 long vfio_platform_ioctl(struct vfio_device *core_vdev,
 			 unsigned int cmd, unsigned long arg);
 int vfio_platform_ioctl_get_region_info(struct vfio_device *core_vdev,
-					struct vfio_region_info __user *arg);
+					struct vfio_region_info *info,
+					struct vfio_info_cap *caps);
 ssize_t vfio_platform_read(struct vfio_device *core_vdev,
 			   char __user *buf, size_t count,
 			   loff_t *ppos);
-- 
2.43.0

[PATCH v2 21/22] vfio: Move the remaining drivers to get_region_info_caps

[Jason Gunthorpe]

Remove the duplicate code and change info to a pointer. caps are not used.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Pranjal Shrivastava <praan@google.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/cdx/main.c           | 24 +++++++------------
 drivers/vfio/fsl-mc/vfio_fsl_mc.c | 29 +++++++----------------
 samples/vfio-mdev/mdpy.c          | 39 ++++++-------------------------
 samples/vfio-mdev/mtty.c          | 38 +++++-------------------------
 4 files changed, 29 insertions(+), 101 deletions(-)

diff --git a/drivers/vfio/cdx/main.c b/drivers/vfio/cdx/main.c
index 506d849139d35a..253031b86b60a7 100644
--- a/drivers/vfio/cdx/main.c
+++ b/drivers/vfio/cdx/main.c
@@ -130,29 +130,21 @@ static int vfio_cdx_ioctl_get_info(struct vfio_cdx_device *vdev,
 }
 
 static int vfio_cdx_ioctl_get_region_info(struct vfio_device *core_vdev,
-					  struct vfio_region_info __user *arg)
+					  struct vfio_region_info *info,
+					  struct vfio_info_cap *caps)
 {
 	struct vfio_cdx_device *vdev =
 		container_of(core_vdev, struct vfio_cdx_device, vdev);
-	unsigned long minsz = offsetofend(struct vfio_region_info, offset);
 	struct cdx_device *cdx_dev = to_cdx_device(vdev->vdev.dev);
-	struct vfio_region_info info;
 
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	if (info.index >= cdx_dev->res_count)
+	if (info->index >= cdx_dev->res_count)
 		return -EINVAL;
 
 	/* map offset to the physical address */
-	info.offset = vfio_cdx_index_to_offset(info.index);
-	info.size = vdev->regions[info.index].size;
-	info.flags = vdev->regions[info.index].flags;
-
-	return copy_to_user(arg, &info, minsz) ? -EFAULT : 0;
+	info->offset = vfio_cdx_index_to_offset(info->index);
+	info->size = vdev->regions[info->index].size;
+	info->flags = vdev->regions[info->index].flags;
+	return 0;
 }
 
 static int vfio_cdx_ioctl_get_irq_info(struct vfio_cdx_device *vdev,
@@ -284,7 +276,7 @@ static const struct vfio_device_ops vfio_cdx_ops = {
 	.open_device	= vfio_cdx_open_device,
 	.close_device	= vfio_cdx_close_device,
 	.ioctl		= vfio_cdx_ioctl,
-	.get_region_info = vfio_cdx_ioctl_get_region_info,
+	.get_region_info_caps = vfio_cdx_ioctl_get_region_info,
 	.device_feature = vfio_cdx_ioctl_feature,
 	.mmap		= vfio_cdx_mmap,
 	.bind_iommufd	= vfio_iommufd_physical_bind,
diff --git a/drivers/vfio/fsl-mc/vfio_fsl_mc.c b/drivers/vfio/fsl-mc/vfio_fsl_mc.c
index d38e51a57f0775..ba47100f28c1d0 100644
--- a/drivers/vfio/fsl-mc/vfio_fsl_mc.c
+++ b/drivers/vfio/fsl-mc/vfio_fsl_mc.c
@@ -117,34 +117,21 @@ static void vfio_fsl_mc_close_device(struct vfio_device *core_vdev)
 	fsl_mc_cleanup_irq_pool(mc_cont);
 }
 
-static int
-vfio_fsl_mc_ioctl_get_region_info(struct vfio_device *core_vdev,
-				  struct vfio_region_info __user *arg)
+static int vfio_fsl_mc_ioctl_get_region_info(struct vfio_device *core_vdev,
+					     struct vfio_region_info *info,
+					     struct vfio_info_cap *caps)
 {
 	struct vfio_fsl_mc_device *vdev =
 		container_of(core_vdev, struct vfio_fsl_mc_device, vdev);
 	struct fsl_mc_device *mc_dev = vdev->mc_dev;
-	struct vfio_region_info info;
-	unsigned long minsz;
 
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	if (info.index >= mc_dev->obj_desc.region_count)
+	if (info->index >= mc_dev->obj_desc.region_count)
 		return -EINVAL;
 
 	/* map offset to the physical address  */
-	info.offset = VFIO_FSL_MC_INDEX_TO_OFFSET(info.index);
-	info.size = vdev->regions[info.index].size;
-	info.flags = vdev->regions[info.index].flags;
-
-	if (copy_to_user(arg, &info, minsz))
-		return -EFAULT;
+	info->offset = VFIO_FSL_MC_INDEX_TO_OFFSET(info->index);
+	info->size = vdev->regions[info->index].size;
+	info->flags = vdev->regions[info->index].flags;
 	return 0;
 }
 
@@ -596,7 +583,7 @@ static const struct vfio_device_ops vfio_fsl_mc_ops = {
 	.open_device	= vfio_fsl_mc_open_device,
 	.close_device	= vfio_fsl_mc_close_device,
 	.ioctl		= vfio_fsl_mc_ioctl,
-	.get_region_info = vfio_fsl_mc_ioctl_get_region_info,
+	.get_region_info_caps = vfio_fsl_mc_ioctl_get_region_info,
 	.read		= vfio_fsl_mc_read,
 	.write		= vfio_fsl_mc_write,
 	.mmap		= vfio_fsl_mc_mmap,
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 0c65ed22173862..0759bd68edca0d 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -435,10 +435,13 @@ static int mdpy_mmap(struct vfio_device *vdev, struct vm_area_struct *vma)
 	return remap_vmalloc_range(vma, mdev_state->memblk, 0);
 }
 
-static int mdpy_get_region_info(struct mdev_state *mdev_state,
-				struct vfio_region_info *region_info,
-				u16 *cap_type_id, void **cap_type)
+static int mdpy_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info *region_info,
+				      struct vfio_info_cap *caps)
 {
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
 	if (region_info->index >= VFIO_PCI_NUM_REGIONS &&
 	    region_info->index != MDPY_DISPLAY_REGION)
 		return -EINVAL;
@@ -512,34 +515,6 @@ static int mdpy_query_gfx_plane(struct mdev_state *mdev_state,
 	return 0;
 }
 
-static int mdpy_ioctl_get_region_info(struct vfio_device *vdev,
-				      struct vfio_region_info __user *arg)
-{
-	struct mdev_state *mdev_state =
-		container_of(vdev, struct mdev_state, vdev);
-	struct vfio_region_info info;
-	void *cap_type = NULL;
-	u16 cap_type_id = 0;
-	unsigned long minsz;
-	int ret;
-
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	ret = mdpy_get_region_info(mdev_state, &info, &cap_type_id, &cap_type);
-	if (ret)
-		return ret;
-
-	if (copy_to_user(arg, &info, minsz))
-		return -EFAULT;
-	return 0;
-}
-
 static long mdpy_ioctl(struct vfio_device *vdev, unsigned int cmd,
 		       unsigned long arg)
 {
@@ -669,7 +644,7 @@ static const struct vfio_device_ops mdpy_dev_ops = {
 	.read = mdpy_read,
 	.write = mdpy_write,
 	.ioctl = mdpy_ioctl,
-	.get_region_info = mdpy_ioctl_get_region_info,
+	.get_region_info_caps = mdpy_ioctl_get_region_info,
 	.mmap = mdpy_mmap,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index b27f9b93471bc5..3e029d0cba1ea8 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -1717,10 +1717,12 @@ static int mtty_set_irqs(struct mdev_state *mdev_state, uint32_t flags,
 	return ret;
 }
 
-static int mtty_get_region_info(struct mdev_state *mdev_state,
-			 struct vfio_region_info *region_info,
-			 u16 *cap_type_id, void **cap_type)
+static int mtty_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info *region_info,
+				      struct vfio_info_cap *caps)
 {
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
 	unsigned int size = 0;
 	u32 bar_index;
 
@@ -1785,34 +1787,6 @@ static int mtty_get_device_info(struct vfio_device_info *dev_info)
 	return 0;
 }
 
-static int mtty_ioctl_get_region_info(struct vfio_device *vdev,
-				      struct vfio_region_info __user *arg)
-{
-	struct mdev_state *mdev_state =
-		container_of(vdev, struct mdev_state, vdev);
-	struct vfio_region_info info;
-	void *cap_type = NULL;
-	u16 cap_type_id = 0;
-	unsigned long minsz;
-	int ret;
-
-	minsz = offsetofend(struct vfio_region_info, offset);
-
-	if (copy_from_user(&info, arg, minsz))
-		return -EFAULT;
-
-	if (info.argsz < minsz)
-		return -EINVAL;
-
-	ret = mtty_get_region_info(mdev_state, &info, &cap_type_id, &cap_type);
-	if (ret)
-		return ret;
-
-	if (copy_to_user(arg, &info, minsz))
-		return -EFAULT;
-	return 0;
-}
-
 static long mtty_ioctl(struct vfio_device *vdev, unsigned int cmd,
 			unsigned long arg)
 {
@@ -1953,7 +1927,7 @@ static const struct vfio_device_ops mtty_dev_ops = {
 	.read = mtty_read,
 	.write = mtty_write,
 	.ioctl = mtty_ioctl,
-	.get_region_info = mtty_ioctl_get_region_info,
+	.get_region_info_caps = mtty_ioctl_get_region_info,
 	.bind_iommufd	= vfio_iommufd_emulated_bind,
 	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
 	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
-- 
2.43.0

[PATCH v2 22/22] vfio: Remove the get_region_info op

[Jason Gunthorpe]

No driver uses it now, all are using get_region_info_caps().

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/vfio/vfio_main.c | 48 ++++++++++++++++++----------------------
 include/linux/vfio.h     |  2 --
 2 files changed, 21 insertions(+), 29 deletions(-)

diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index 48d034aede46fc..b8fe1a75e48a0b 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -1267,42 +1267,36 @@ static long vfio_get_region_info(struct vfio_device *device,
 	struct vfio_info_cap caps = {};
 	int ret;
 
+	if (unlikely(!device->ops->get_region_info_caps))
+		return -EINVAL;
+
 	if (copy_from_user(&info, arg, minsz))
 		return -EFAULT;
 	if (info.argsz < minsz)
 		return -EINVAL;
 
-	if (device->ops->get_region_info_caps) {
-		ret = device->ops->get_region_info_caps(device, &info, &caps);
-		if (ret)
-			goto out_free;
+	ret = device->ops->get_region_info_caps(device, &info, &caps);
+	if (ret)
+		goto out_free;
 
-		if (caps.size) {
-			info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
-			if (info.argsz < sizeof(info) + caps.size) {
-				info.argsz = sizeof(info) + caps.size;
-				info.cap_offset = 0;
-			} else {
-				vfio_info_cap_shift(&caps, sizeof(info));
-				if (copy_to_user(arg + 1, caps.buf,
-						 caps.size)) {
-					ret = -EFAULT;
-					goto out_free;
-				}
-				info.cap_offset = sizeof(info);
+	if (caps.size) {
+		info.flags |= VFIO_REGION_INFO_FLAG_CAPS;
+		if (info.argsz < sizeof(info) + caps.size) {
+			info.argsz = sizeof(info) + caps.size;
+			info.cap_offset = 0;
+		} else {
+			vfio_info_cap_shift(&caps, sizeof(info));
+			if (copy_to_user(arg + 1, caps.buf, caps.size)) {
+				ret = -EFAULT;
+				goto out_free;
 			}
+			info.cap_offset = sizeof(info);
 		}
+	}
 
-		if (copy_to_user(arg, &info, minsz)) {
-			ret = -EFAULT;
-			goto out_free;
-		}
-	} else if (device->ops->get_region_info) {
-		ret = device->ops->get_region_info(device, arg);
-		if (ret)
-			return ret;
-	} else {
-		return -EINVAL;
+	if (copy_to_user(arg, &info, minsz)){
+		ret = -EFAULT;
+		goto out_free;
 	}
 
 out_free:
diff --git a/include/linux/vfio.h b/include/linux/vfio.h
index 6311ddc837701d..8e1ddb48b9b54e 100644
--- a/include/linux/vfio.h
+++ b/include/linux/vfio.h
@@ -133,8 +133,6 @@ struct vfio_device_ops {
 			 size_t count, loff_t *size);
 	long	(*ioctl)(struct vfio_device *vdev, unsigned int cmd,
 			 unsigned long arg);
-	int	(*get_region_info)(struct vfio_device *vdev,
-				   struct vfio_region_info __user *arg);
 	int	(*get_region_info_caps)(struct vfio_device *vdev,
 					struct vfio_region_info *info,
 					struct vfio_info_cap *caps);
-- 
2.43.0

Re: [PATCH v2 12/22] vfio/ccw: Provide a get_region_info op

[Eric Farman]

On Fri, 2025-11-07 at 13:41 -0400, Jason Gunthorpe wrote:
> Move it out of vfio_ccw_mdev_ioctl() and re-indent it.
> 
> Reviewed-by: Kevin Tian <kevin.tian@intel.com>
> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
> ---
>  drivers/s390/cio/vfio_ccw_ops.c | 44 +++++++++++++++++++--------------
>  1 file changed, 26 insertions(+), 18 deletions(-)

Reviewed-by: Eric Farman <farman@linux.ibm.com>

Re: [PATCH v2 18/22] vfio/ccw: Convert to get_region_info_caps

[Eric Farman]

On Fri, 2025-11-07 at 13:41 -0400, Jason Gunthorpe wrote:
> Remove the duplicate code and flatten the call chain.
> 
> Reviewed-by: Kevin Tian <kevin.tian@intel.com>
> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
> ---
>  drivers/s390/cio/vfio_ccw_ops.c | 55 +++++----------------------------
>  1 file changed, 7 insertions(+), 48 deletions(-)

Nice.

Reviewed-by: Eric Farman <farman@linux.ibm.com>