public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Yi Liu <yi.l.liu@intel.com>
To: alex.williamson@redhat.com, jgg@nvidia.com
Cc: kevin.tian@intel.com, cohuck@redhat.com, eric.auger@redhat.com,
	nicolinc@nvidia.com, kvm@vger.kernel.org, mjrosato@linux.ibm.com,
	chao.p.peng@linux.intel.com, yi.l.liu@intel.com,
	yi.y.sun@linux.intel.com, peterx@redhat.com, jasowang@redhat.com
Subject: [RFC 07/12] vfio: Block device access via device fd until device is opened
Date: Mon, 19 Dec 2022 00:47:13 -0800	[thread overview]
Message-ID: <20221219084718.9342-8-yi.l.liu@intel.com> (raw)
In-Reply-To: <20221219084718.9342-1-yi.l.liu@intel.com>

Allow the vfio_device file to be in a state where the device FD is
opened but the device is not opened. This inbetween state is not used
when the device FD is spawned from the group FD, however when we create
the device FD directly by opening a cdev it will be opened in the
blocked state.

In the blocked state, currently only the bind operation is allowed,
other device accesses are not allowed. Completing bind will allow user
to further access the device.

This is implemented by adding a flag in struct vfio_device_file to mark
the blocked state and using a simple smp_load_acquire() to obtain the
flag value and serialize all the device setup with the thread accessing
this device.

Due to this scheme it is not possible to unbind the FD, once it bound it
remains bound until the FD is closed.

Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yi Liu <yi.l.liu@intel.com>
---
 drivers/vfio/vfio.h      |  1 +
 drivers/vfio/vfio_main.c | 34 ++++++++++++++++++++++++++++++++++
 2 files changed, 35 insertions(+)

diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h
index 24ccf9f0ab16..4f29e3a89f64 100644
--- a/drivers/vfio/vfio.h
+++ b/drivers/vfio/vfio.h
@@ -20,6 +20,7 @@ struct vfio_device_file {
 	struct vfio_device *device;
 	struct kvm *kvm;
 	struct iommufd_ctx *iommufd;
+	bool access_granted;
 };
 
 void vfio_device_put_registration(struct vfio_device *device);
diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index 2aa3597abb40..bd1d9621af46 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -444,6 +444,11 @@ int vfio_device_open(struct vfio_device_file *df)
 		}
 	}
 
+	/*
+	 * Paired with smp_load_acquire() in vfio_device_fops::ioctl/
+	 * read/write/mmap
+	 */
+	smp_store_release(&df->access_granted, true);
 	return 0;
 }
 
@@ -452,6 +457,11 @@ void vfio_device_close(struct vfio_device_file *df)
 	struct vfio_device *device = df->device;
 
 	mutex_lock(&device->dev_set->lock);
+	/*
+	 * Paired with smp_load_acquire() in vfio_device_fops::ioctl/
+	 * read/write/mmap
+	 */
+	smp_store_release(&df->access_granted, false);
 	vfio_assert_device_open(device);
 	if (device->open_count == 1)
 		vfio_device_last_close(df);
@@ -968,8 +978,14 @@ static long vfio_device_fops_unl_ioctl(struct file *filep,
 {
 	struct vfio_device_file *df = filep->private_data;
 	struct vfio_device *device = df->device;
+	bool access;
 	int ret;
 
+	/* Paired with smp_store_release() in vfio_device_open/close() */
+	access = smp_load_acquire(&df->access_granted);
+	if (!access)
+		return -EINVAL;
+
 	ret = vfio_device_pm_runtime_get(device);
 	if (ret)
 		return ret;
@@ -996,6 +1012,12 @@ static ssize_t vfio_device_fops_read(struct file *filep, char __user *buf,
 {
 	struct vfio_device_file *df = filep->private_data;
 	struct vfio_device *device = df->device;
+	bool access;
+
+	/* Paired with smp_store_release() in vfio_device_open/close() */
+	access = smp_load_acquire(&df->access_granted);
+	if (!access)
+		return -EINVAL;
 
 	if (unlikely(!device->ops->read))
 		return -EINVAL;
@@ -1009,6 +1031,12 @@ static ssize_t vfio_device_fops_write(struct file *filep,
 {
 	struct vfio_device_file *df = filep->private_data;
 	struct vfio_device *device = df->device;
+	bool access;
+
+	/* Paired with smp_store_release() in vfio_device_open/close() */
+	access = smp_load_acquire(&df->access_granted);
+	if (!access)
+		return -EINVAL;
 
 	if (unlikely(!device->ops->write))
 		return -EINVAL;
@@ -1020,6 +1048,12 @@ static int vfio_device_fops_mmap(struct file *filep, struct vm_area_struct *vma)
 {
 	struct vfio_device_file *df = filep->private_data;
 	struct vfio_device *device = df->device;
+	bool access;
+
+	/* Paired with smp_store_release() in vfio_device_open/close() */
+	access = smp_load_acquire(&df->access_granted);
+	if (!access)
+		return -EINVAL;
 
 	if (unlikely(!device->ops->mmap))
 		return -EINVAL;
-- 
2.34.1


  parent reply	other threads:[~2022-12-19  8:48 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-19  8:47 [RFC 00/12] Add vfio_device cdev for iommufd support Yi Liu
2022-12-19  8:47 ` [RFC 01/12] vfio: Allocate per device file structure Yi Liu
2022-12-21  3:57   ` Tian, Kevin
2022-12-21  6:46     ` Yi Liu
2022-12-19  8:47 ` [RFC 02/12] vfio: Refine vfio file kAPIs Yi Liu
2022-12-19  8:47 ` [RFC 03/12] vfio: Accept vfio device file in the driver facing kAPI Yi Liu
2022-12-21  4:07   ` Tian, Kevin
2022-12-21  7:02     ` Yi Liu
2023-01-04 18:25     ` Jason Gunthorpe
2023-01-09  4:13       ` Tian, Kevin
2022-12-19  8:47 ` [RFC 04/12] kvm/vfio: Rename kvm_vfio_group to prepare for accepting vfio device fd Yi Liu
2022-12-19  8:47 ` [RFC 05/12] kvm/vfio: Accept vfio device file from userspace Yi Liu
2023-01-06 14:32   ` Jason Gunthorpe
2023-01-06 14:46     ` Yi Liu
2023-01-06 14:55       ` Jason Gunthorpe
2023-01-06 15:04         ` Yi Liu
2023-01-06 16:08           ` Jason Gunthorpe
2023-01-09  4:17           ` Tian, Kevin
2023-01-09  4:26             ` Yi Liu
2022-12-19  8:47 ` [RFC 06/12] vfio: Pass struct vfio_device_file * to vfio_device_open/close() Yi Liu
2022-12-21  4:10   ` Tian, Kevin
2022-12-21  7:04     ` Yi Liu
2022-12-19  8:47 ` Yi Liu [this message]
2022-12-21  4:18   ` [RFC 07/12] vfio: Block device access via device fd until device is opened Tian, Kevin
2023-01-04 20:47   ` Jason Gunthorpe
2022-12-19  8:47 ` [RFC 08/12] vfio: Add infrastructure for bind_iommufd and attach Yi Liu
2023-01-09  5:46   ` Tian, Kevin
2023-01-09 13:21     ` Jason Gunthorpe
2023-01-10  2:53       ` Tian, Kevin
2022-12-19  8:47 ` [RFC 09/12] vfio: Make vfio_device_open() exclusive between group path and device cdev path Yi Liu
2023-01-09  6:03   ` Tian, Kevin
2022-12-19  8:47 ` [RFC 10/12] vfio: Add cdev for vfio_device Yi Liu
2023-01-09  6:54   ` Tian, Kevin
2022-12-19  8:47 ` [RFC 11/12] vfio: Add ioctls for device cdev iommufd Yi Liu
2023-01-09  7:47   ` Tian, Kevin
2023-01-09 14:14     ` Jason Gunthorpe
2023-01-09 15:07       ` Yi Liu
2023-01-09 15:12         ` Jason Gunthorpe
2023-01-09 15:20           ` Yi Liu
2023-01-09 14:55     ` Yi Liu
2023-01-10  2:57       ` Tian, Kevin
2022-12-19  8:47 ` [RFC 12/12] vfio: Compile group optionally Yi Liu
2022-12-19  8:51 ` [RFC 00/12] Add vfio_device cdev for iommufd support Yi Liu
2023-01-04 15:23 ` Yi Liu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20221219084718.9342-8-yi.l.liu@intel.com \
    --to=yi.l.liu@intel.com \
    --cc=alex.williamson@redhat.com \
    --cc=chao.p.peng@linux.intel.com \
    --cc=cohuck@redhat.com \
    --cc=eric.auger@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=jgg@nvidia.com \
    --cc=kevin.tian@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=mjrosato@linux.ibm.com \
    --cc=nicolinc@nvidia.com \
    --cc=peterx@redhat.com \
    --cc=yi.y.sun@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox