summaryrefslogtreecommitdiff
path: root/Documentation/driver-api
diff options
context:
space:
mode:
Diffstat (limited to 'Documentation/driver-api')
-rw-r--r--Documentation/driver-api/basics.rst27
-rw-r--r--Documentation/driver-api/infrastructure.rst18
-rw-r--r--Documentation/driver-api/interconnect.rst25
-rw-r--r--Documentation/driver-api/media/cec-core.rst44
-rw-r--r--Documentation/driver-api/media/dtv-common.rst9
-rw-r--r--Documentation/driver-api/media/v4l2-cci.rst5
-rw-r--r--Documentation/driver-api/media/v4l2-core.rst1
-rw-r--r--Documentation/driver-api/media/v4l2-subdev.rst110
-rw-r--r--Documentation/driver-api/s390-drivers.rst4
-rw-r--r--Documentation/driver-api/tty/tty_buffer.rst7
-rw-r--r--Documentation/driver-api/usb/usb.rst9
-rw-r--r--Documentation/driver-api/vfio.rst147
12 files changed, 329 insertions, 77 deletions
diff --git a/Documentation/driver-api/basics.rst b/Documentation/driver-api/basics.rst
index 7671b531ba1a..d78b7c328ff7 100644
--- a/Documentation/driver-api/basics.rst
+++ b/Documentation/driver-api/basics.rst
@@ -15,8 +15,8 @@ Driver device table
:no-identifiers: pci_device_id
-Delaying, scheduling, and timer routines
-----------------------------------------
+Delaying and scheduling routines
+--------------------------------
.. kernel-doc:: include/linux/sched.h
:internal:
@@ -33,16 +33,16 @@ Delaying, scheduling, and timer routines
.. kernel-doc:: include/linux/completion.h
:internal:
-.. kernel-doc:: kernel/time/timer.c
- :export:
-
-Wait queues and Wake events
----------------------------
+Time and timer routines
+-----------------------
-.. kernel-doc:: include/linux/wait.h
+.. kernel-doc:: include/linux/jiffies.h
:internal:
-.. kernel-doc:: kernel/sched/wait.c
+.. kernel-doc:: kernel/time/time.c
+ :export:
+
+.. kernel-doc:: kernel/time/timer.c
:export:
High-resolution timers
@@ -57,6 +57,15 @@ High-resolution timers
.. kernel-doc:: kernel/time/hrtimer.c
:export:
+Wait queues and Wake events
+---------------------------
+
+.. kernel-doc:: include/linux/wait.h
+ :internal:
+
+.. kernel-doc:: kernel/sched/wait.c
+ :export:
+
Internal Functions
------------------
diff --git a/Documentation/driver-api/infrastructure.rst b/Documentation/driver-api/infrastructure.rst
index 683bd460e222..3d52dfdfa9fd 100644
--- a/Documentation/driver-api/infrastructure.rst
+++ b/Documentation/driver-api/infrastructure.rst
@@ -8,12 +8,24 @@ The Basic Device Driver-Model Structures
:internal:
:no-identifiers: device_link_state
+.. kernel-doc:: include/linux/device/bus.h
+ :identifiers: bus_type bus_notifier_event
+
+.. kernel-doc:: include/linux/device/class.h
+ :identifiers: class
+
+.. kernel-doc:: include/linux/device/driver.h
+ :identifiers: probe_type device_driver
+
Device Drivers Base
-------------------
.. kernel-doc:: drivers/base/init.c
:internal:
+.. kernel-doc:: include/linux/device/driver.h
+ :no-identifiers: probe_type device_driver
+
.. kernel-doc:: drivers/base/driver.c
:export:
@@ -23,6 +35,9 @@ Device Drivers Base
.. kernel-doc:: drivers/base/syscore.c
:export:
+.. kernel-doc:: include/linux/device/class.h
+ :no-identifiers: class
+
.. kernel-doc:: drivers/base/class.c
:export:
@@ -41,6 +56,9 @@ Device Drivers Base
.. kernel-doc:: drivers/base/platform.c
:export:
+.. kernel-doc:: include/linux/device/bus.h
+ :no-identifiers: bus_type bus_notifier_event
+
.. kernel-doc:: drivers/base/bus.c
:export:
diff --git a/Documentation/driver-api/interconnect.rst b/Documentation/driver-api/interconnect.rst
index 5ed4f57a6bac..a92d0f277a1f 100644
--- a/Documentation/driver-api/interconnect.rst
+++ b/Documentation/driver-api/interconnect.rst
@@ -113,3 +113,28 @@ through dot to generate diagrams in many graphical formats::
$ cat /sys/kernel/debug/interconnect/interconnect_graph | \
dot -Tsvg > interconnect_graph.svg
+
+The ``test-client`` directory provides interfaces for issuing BW requests to
+any arbitrary path. Note that for safety reasons, this feature is disabled by
+default without a Kconfig to enable it. Enabling it requires code changes to
+``#define INTERCONNECT_ALLOW_WRITE_DEBUGFS``. Example usage::
+
+ cd /sys/kernel/debug/interconnect/test-client/
+
+ # Configure node endpoints for the path from CPU to DDR on
+ # qcom/sm8550.
+ echo chm_apps > src_node
+ echo ebi > dst_node
+
+ # Get path between src_node and dst_node. This is only
+ # necessary after updating the node endpoints.
+ echo 1 > get
+
+ # Set desired BW to 1GBps avg and 2GBps peak.
+ echo 1000000 > avg_bw
+ echo 2000000 > peak_bw
+
+ # Vote for avg_bw and peak_bw on the latest path from "get".
+ # Voting for multiple paths is possible by repeating this
+ # process for different nodes endpoints.
+ echo 1 > commit
diff --git a/Documentation/driver-api/media/cec-core.rst b/Documentation/driver-api/media/cec-core.rst
index ae0d20798edc..f1ffdec388f3 100644
--- a/Documentation/driver-api/media/cec-core.rst
+++ b/Documentation/driver-api/media/cec-core.rst
@@ -109,9 +109,11 @@ your driver:
int (*adap_monitor_all_enable)(struct cec_adapter *adap, bool enable);
int (*adap_monitor_pin_enable)(struct cec_adapter *adap, bool enable);
int (*adap_log_addr)(struct cec_adapter *adap, u8 logical_addr);
- void (*adap_configured)(struct cec_adapter *adap, bool configured);
+ void (*adap_unconfigured)(struct cec_adapter *adap);
int (*adap_transmit)(struct cec_adapter *adap, u8 attempts,
u32 signal_free_time, struct cec_msg *msg);
+ void (*adap_nb_transmit_canceled)(struct cec_adapter *adap,
+ const struct cec_msg *msg);
void (*adap_status)(struct cec_adapter *adap, struct seq_file *file);
void (*adap_free)(struct cec_adapter *adap);
@@ -122,8 +124,8 @@ your driver:
...
};
-The seven low-level ops deal with various aspects of controlling the CEC adapter
-hardware:
+These low-level ops deal with various aspects of controlling the CEC adapter
+hardware. They are all called with the mutex adap->lock held.
To enable/disable the hardware::
@@ -179,14 +181,12 @@ can receive directed messages to that address.
Note that adap_log_addr must return 0 if logical_addr is CEC_LOG_ADDR_INVALID.
-Called when the adapter is fully configured or unconfigured::
+Called when the adapter is unconfigured::
- void (*adap_configured)(struct cec_adapter *adap, bool configured);
+ void (*adap_unconfigured)(struct cec_adapter *adap);
-If configured == true, then the adapter is fully configured, i.e. all logical
-addresses have been successfully claimed. If configured == false, then the
-adapter is unconfigured. If the driver has to take specific actions after
-(un)configuration, then that can be done through this optional callback.
+The adapter is unconfigured. If the driver has to take specific actions after
+unconfiguration, then that can be done through this optional callback.
To transmit a new message::
@@ -207,6 +207,19 @@ The CEC_FREE_TIME_TO_USEC macro can be used to convert signal_free_time to
microseconds (one data bit period is 2.4 ms).
+To pass on the result of a canceled non-blocking transmit::
+
+ void (*adap_nb_transmit_canceled)(struct cec_adapter *adap,
+ const struct cec_msg *msg);
+
+This optional callback can be used to obtain the result of a canceled
+non-blocking transmit with sequence number msg->sequence. This is
+called if the transmit was aborted, the transmit timed out (i.e. the
+hardware never signaled that the transmit finished), or the transmit
+was successful, but the wait for the expected reply was either aborted
+or it timed out.
+
+
To log the current CEC hardware status::
void (*adap_status)(struct cec_adapter *adap, struct seq_file *file);
@@ -372,7 +385,8 @@ Implementing the High-Level CEC Adapter
---------------------------------------
The low-level operations drive the hardware, the high-level operations are
-CEC protocol driven. The following high-level callbacks are available:
+CEC protocol driven. The high-level callbacks are called without the adap->lock
+mutex being held. The following high-level callbacks are available:
.. code-block:: none
@@ -384,9 +398,19 @@ CEC protocol driven. The following high-level callbacks are available:
...
/* High-level CEC message callback */
+ void (*configured)(struct cec_adapter *adap);
int (*received)(struct cec_adapter *adap, struct cec_msg *msg);
};
+Called when the adapter is configured::
+
+ void (*configured)(struct cec_adapter *adap);
+
+The adapter is fully configured, i.e. all logical addresses have been
+successfully claimed. If the driver has to take specific actions after
+configuration, then that can be done through this optional callback.
+
+
The received() callback allows the driver to optionally handle a newly
received CEC message::
diff --git a/Documentation/driver-api/media/dtv-common.rst b/Documentation/driver-api/media/dtv-common.rst
index f8b2c4dc8170..207a22bcaf4a 100644
--- a/Documentation/driver-api/media/dtv-common.rst
+++ b/Documentation/driver-api/media/dtv-common.rst
@@ -3,15 +3,6 @@
Digital TV Common functions
---------------------------
-Math functions
-~~~~~~~~~~~~~~
-
-Provide some commonly-used math functions, usually required in order to
-estimate signal strength and signal to noise measurements in dB.
-
-.. kernel-doc:: include/media/dvb_math.h
-
-
DVB devices
~~~~~~~~~~~
diff --git a/Documentation/driver-api/media/v4l2-cci.rst b/Documentation/driver-api/media/v4l2-cci.rst
new file mode 100644
index 000000000000..dd297a40ed20
--- /dev/null
+++ b/Documentation/driver-api/media/v4l2-cci.rst
@@ -0,0 +1,5 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+V4L2 CCI kAPI
+^^^^^^^^^^^^^
+.. kernel-doc:: include/media/v4l2-cci.h
diff --git a/Documentation/driver-api/media/v4l2-core.rst b/Documentation/driver-api/media/v4l2-core.rst
index 1a8c4a5f256b..239045ecc8f4 100644
--- a/Documentation/driver-api/media/v4l2-core.rst
+++ b/Documentation/driver-api/media/v4l2-core.rst
@@ -22,6 +22,7 @@ Video4Linux devices
v4l2-mem2mem
v4l2-async
v4l2-fwnode
+ v4l2-cci
v4l2-rect
v4l2-tuner
v4l2-common
diff --git a/Documentation/driver-api/media/v4l2-subdev.rst b/Documentation/driver-api/media/v4l2-subdev.rst
index 602dadaa81d8..e56b50b3f203 100644
--- a/Documentation/driver-api/media/v4l2-subdev.rst
+++ b/Documentation/driver-api/media/v4l2-subdev.rst
@@ -157,6 +157,9 @@ below.
Using one or the other registration method only affects the probing process, the
run-time bridge-subdevice interaction is in both cases the same.
+Registering synchronous sub-devices
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
In the **synchronous** case a device (bridge) driver needs to register the
:c:type:`v4l2_subdev` with the v4l2_device:
@@ -175,10 +178,12 @@ You can unregister a sub-device using:
:c:func:`v4l2_device_unregister_subdev <v4l2_device_unregister_subdev>`
(:c:type:`sd <v4l2_subdev>`).
-
Afterwards the subdev module can be unloaded and
:c:type:`sd <v4l2_subdev>`->dev == ``NULL``.
+Registering asynchronous sub-devices
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
In the **asynchronous** case subdevice probing can be invoked independently of
the bridge driver availability. The subdevice driver then has to verify whether
all the requirements for a successful probing are satisfied. This can include a
@@ -190,64 +195,89 @@ performed using the :c:func:`v4l2_async_unregister_subdev` call. Subdevices
registered this way are stored in a global list of subdevices, ready to be
picked up by bridge drivers.
-Bridge drivers in turn have to register a notifier object. This is
-performed using the :c:func:`v4l2_async_nf_register` call. To
-unregister the notifier the driver has to call
-:c:func:`v4l2_async_nf_unregister`. The former of the two functions
-takes two arguments: a pointer to struct :c:type:`v4l2_device` and a
-pointer to struct :c:type:`v4l2_async_notifier`.
+Asynchronous sub-device notifiers
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Bridge drivers in turn have to register a notifier object. This is performed
+using the :c:func:`v4l2_async_nf_register` call. To unregister the notifier the
+driver has to call :c:func:`v4l2_async_nf_unregister`. Before releasing memory
+of an unregister notifier, it must be cleaned up by calling
+:c:func:`v4l2_async_nf_cleanup`.
Before registering the notifier, bridge drivers must do two things: first, the
-notifier must be initialized using the :c:func:`v4l2_async_nf_init`.
-Second, bridge drivers can then begin to form a list of subdevice descriptors
-that the bridge device needs for its operation. Several functions are available
-to add subdevice descriptors to a notifier, depending on the type of device and
-the needs of the driver.
-
-:c:func:`v4l2_async_nf_add_fwnode_remote` and
-:c:func:`v4l2_async_nf_add_i2c` are for bridge and ISP drivers for
-registering their async sub-devices with the notifier.
-
-:c:func:`v4l2_async_register_subdev_sensor` is a helper function for
-sensor drivers registering their own async sub-device, but it also registers a
-notifier and further registers async sub-devices for lens and flash devices
-found in firmware. The notifier for the sub-device is unregistered with the
-async sub-device.
-
-These functions allocate an async sub-device descriptor which is of type struct
-:c:type:`v4l2_async_subdev` embedded in a driver-specific struct. The &struct
-:c:type:`v4l2_async_subdev` shall be the first member of this struct:
+notifier must be initialized using the :c:func:`v4l2_async_nf_init`. Second,
+bridge drivers can then begin to form a list of async connection descriptors
+that the bridge device needs for its
+operation. :c:func:`v4l2_async_nf_add_fwnode`,
+:c:func:`v4l2_async_nf_add_fwnode_remote` and :c:func:`v4l2_async_nf_add_i2c`
+
+Async connection descriptors describe connections to external sub-devices the
+drivers for which are not yet probed. Based on an async connection, a media data
+or ancillary link may be created when the related sub-device becomes
+available. There may be one or more async connections to a given sub-device but
+this is not known at the time of adding the connections to the notifier. Async
+connections are bound as matching async sub-devices are found, one by one.
+
+Asynchronous sub-device notifier for sub-devices
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+A driver that registers an asynchronous sub-device may also register an
+asynchronous notifier. This is called an asynchronous sub-device notifier andthe
+process is similar to that of a bridge driver apart from that the notifier is
+initialised using :c:func:`v4l2_async_subdev_nf_init` instead. A sub-device
+notifier may complete only after the V4L2 device becomes available, i.e. there's
+a path via async sub-devices and notifiers to a notifier that is not an
+asynchronous sub-device notifier.
+
+Asynchronous sub-device registration helper for camera sensor drivers
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+:c:func:`v4l2_async_register_subdev_sensor` is a helper function for sensor
+drivers registering their own async connection, but it also registers a notifier
+and further registers async connections for lens and flash devices found in
+firmware. The notifier for the sub-device is unregistered and cleaned up with
+the async sub-device, using :c:func:`v4l2_async_unregister_subdev`.
+
+Asynchronous sub-device notifier example
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+These functions allocate an async connection descriptor which is of type struct
+:c:type:`v4l2_async_connection` embedded in a driver-specific struct. The &struct
+:c:type:`v4l2_async_connection` shall be the first member of this struct:
.. code-block:: c
- struct my_async_subdev {
- struct v4l2_async_subdev asd;
+ struct my_async_connection {
+ struct v4l2_async_connection asc;
...
};
- struct my_async_subdev *my_asd;
+ struct my_async_connection *my_asc;
struct fwnode_handle *ep;
...
- my_asd = v4l2_async_nf_add_fwnode_remote(&notifier, ep,
- struct my_async_subdev);
+ my_asc = v4l2_async_nf_add_fwnode_remote(&notifier, ep,
+ struct my_async_connection);
fwnode_handle_put(ep);
- if (IS_ERR(asd))
- return PTR_ERR(asd);
+ if (IS_ERR(my_asc))
+ return PTR_ERR(my_asc);
+
+Asynchronous sub-device notifier callbacks
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-The V4L2 core will then use these descriptors to match asynchronously
-registered subdevices to them. If a match is detected the ``.bound()``
-notifier callback is called. After all subdevices have been located the
-.complete() callback is called. When a subdevice is removed from the
-system the .unbind() method is called. All three callbacks are optional.
+The V4L2 core will then use these connection descriptors to match asynchronously
+registered subdevices to them. If a match is detected the ``.bound()`` notifier
+callback is called. After all connections have been bound the .complete()
+callback is called. When a connection is removed from the system the
+``.unbind()`` method is called. All three callbacks are optional.
Drivers can store any type of custom data in their driver-specific
-:c:type:`v4l2_async_subdev` wrapper. If any of that data requires special
+:c:type:`v4l2_async_connection` wrapper. If any of that data requires special
handling when the structure is freed, drivers must implement the ``.destroy()``
notifier callback. The framework will call it right before freeing the
-:c:type:`v4l2_async_subdev`.
+:c:type:`v4l2_async_connection`.
Calling subdev operations
~~~~~~~~~~~~~~~~~~~~~~~~~
diff --git a/Documentation/driver-api/s390-drivers.rst b/Documentation/driver-api/s390-drivers.rst
index 5158577bc29b..8c0845c4eee7 100644
--- a/Documentation/driver-api/s390-drivers.rst
+++ b/Documentation/driver-api/s390-drivers.rst
@@ -27,7 +27,7 @@ not strictly considered I/O devices. They are considered here as well,
although they are not the focus of this document.
Some additional information can also be found in the kernel source under
-Documentation/s390/driver-model.rst.
+Documentation/arch/s390/driver-model.rst.
The css bus
===========
@@ -38,7 +38,7 @@ into several categories:
* Standard I/O subchannels, for use by the system. They have a child
device on the ccw bus and are described below.
* I/O subchannels bound to the vfio-ccw driver. See
- Documentation/s390/vfio-ccw.rst.
+ Documentation/arch/s390/vfio-ccw.rst.
* Message subchannels. No Linux driver currently exists.
* CHSC subchannels (at most one). The chsc subchannel driver can be used
to send asynchronous chsc commands.
diff --git a/Documentation/driver-api/tty/tty_buffer.rst b/Documentation/driver-api/tty/tty_buffer.rst
index a39d4781e0d2..4b5ca1776d4f 100644
--- a/Documentation/driver-api/tty/tty_buffer.rst
+++ b/Documentation/driver-api/tty/tty_buffer.rst
@@ -15,10 +15,13 @@ Flip Buffer Management
======================
.. kernel-doc:: drivers/tty/tty_buffer.c
- :identifiers: tty_prepare_flip_string tty_insert_flip_string_fixed_flag
- tty_insert_flip_string_flags __tty_insert_flip_char
+ :identifiers: tty_prepare_flip_string
tty_flip_buffer_push tty_ldisc_receive_buf
+.. kernel-doc:: include/linux/tty_flip.h
+ :identifiers: tty_insert_flip_string_fixed_flag tty_insert_flip_string_flags
+ tty_insert_flip_char
+
----
Other Functions
diff --git a/Documentation/driver-api/usb/usb.rst b/Documentation/driver-api/usb/usb.rst
index 2c94ff2f4385..fb41768696ec 100644
--- a/Documentation/driver-api/usb/usb.rst
+++ b/Documentation/driver-api/usb/usb.rst
@@ -420,6 +420,12 @@ USBDEVFS_CONNECTINFO
know the devnum value already, it's the DDD value of the device file
name.
+USBDEVFS_GET_SPEED
+ Returns the speed of the device. The speed is returned as a
+ nummerical value in accordance with enum usb_device_speed
+
+ File modification time is not updated by this request.
+
USBDEVFS_GETDRIVER
Returns the name of the kernel driver bound to a given interface (a
string). Parameter is a pointer to this structure, which is
@@ -771,8 +777,7 @@ Speed may be:
======= ======================================================
1.5 Mbit/s for low speed USB
12 Mbit/s for full speed USB
- 480 Mbit/s for high speed USB (added for USB 2.0);
- also used for Wireless USB, which has no fixed speed
+ 480 Mbit/s for high speed USB (added for USB 2.0)
5000 Mbit/s for SuperSpeed USB (added for USB 3.0)
======= ======================================================
diff --git a/Documentation/driver-api/vfio.rst b/Documentation/driver-api/vfio.rst
index 68abc089d6dd..633d11c7fa71 100644
--- a/Documentation/driver-api/vfio.rst
+++ b/Documentation/driver-api/vfio.rst
@@ -239,6 +239,137 @@ group and can access them as follows::
/* Gratuitous device reset and go... */
ioctl(device, VFIO_DEVICE_RESET);
+IOMMUFD and vfio_iommu_type1
+----------------------------
+
+IOMMUFD is the new user API to manage I/O page tables from userspace.
+It intends to be the portal of delivering advanced userspace DMA
+features (nested translation [5]_, PASID [6]_, etc.) while also providing
+a backwards compatibility interface for existing VFIO_TYPE1v2_IOMMU use
+cases. Eventually the vfio_iommu_type1 driver, as well as the legacy
+vfio container and group model is intended to be deprecated.
+
+The IOMMUFD backwards compatibility interface can be enabled two ways.
+In the first method, the kernel can be configured with
+CONFIG_IOMMUFD_VFIO_CONTAINER, in which case the IOMMUFD subsystem
+transparently provides the entire infrastructure for the VFIO
+container and IOMMU backend interfaces. The compatibility mode can
+also be accessed if the VFIO container interface, ie. /dev/vfio/vfio is
+simply symlink'd to /dev/iommu. Note that at the time of writing, the
+compatibility mode is not entirely feature complete relative to
+VFIO_TYPE1v2_IOMMU (ex. DMA mapping MMIO) and does not attempt to
+provide compatibility to the VFIO_SPAPR_TCE_IOMMU interface. Therefore
+it is not generally advisable at this time to switch from native VFIO
+implementations to the IOMMUFD compatibility interfaces.
+
+Long term, VFIO users should migrate to device access through the cdev
+interface described below, and native access through the IOMMUFD
+provided interfaces.
+
+VFIO Device cdev
+----------------
+
+Traditionally user acquires a device fd via VFIO_GROUP_GET_DEVICE_FD
+in a VFIO group.
+
+With CONFIG_VFIO_DEVICE_CDEV=y the user can now acquire a device fd
+by directly opening a character device /dev/vfio/devices/vfioX where
+"X" is the number allocated uniquely by VFIO for registered devices.
+cdev interface does not support noiommu devices, so user should use
+the legacy group interface if noiommu is wanted.
+
+The cdev only works with IOMMUFD. Both VFIO drivers and applications
+must adapt to the new cdev security model which requires using
+VFIO_DEVICE_BIND_IOMMUFD to claim DMA ownership before starting to
+actually use the device. Once BIND succeeds then a VFIO device can
+be fully accessed by the user.
+
+VFIO device cdev doesn't rely on VFIO group/container/iommu drivers.
+Hence those modules can be fully compiled out in an environment
+where no legacy VFIO application exists.
+
+So far SPAPR does not support IOMMUFD yet. So it cannot support device
+cdev either.
+
+vfio device cdev access is still bound by IOMMU group semantics, ie. there
+can be only one DMA owner for the group. Devices belonging to the same
+group can not be bound to multiple iommufd_ctx or shared between native
+kernel and vfio bus driver or other driver supporting the driver_managed_dma
+flag. A violation of this ownership requirement will fail at the
+VFIO_DEVICE_BIND_IOMMUFD ioctl, which gates full device access.
+
+Device cdev Example
+-------------------
+
+Assume user wants to access PCI device 0000:6a:01.0::
+
+ $ ls /sys/bus/pci/devices/0000:6a:01.0/vfio-dev/
+ vfio0
+
+This device is therefore represented as vfio0. The user can verify
+its existence::
+
+ $ ls -l /dev/vfio/devices/vfio0
+ crw------- 1 root root 511, 0 Feb 16 01:22 /dev/vfio/devices/vfio0
+ $ cat /sys/bus/pci/devices/0000:6a:01.0/vfio-dev/vfio0/dev
+ 511:0
+ $ ls -l /dev/char/511\:0
+ lrwxrwxrwx 1 root root 21 Feb 16 01:22 /dev/char/511:0 -> ../vfio/devices/vfio0
+
+Then provide the user with access to the device if unprivileged
+operation is desired::
+
+ $ chown user:user /dev/vfio/devices/vfio0
+
+Finally the user could get cdev fd by::
+
+ cdev_fd = open("/dev/vfio/devices/vfio0", O_RDWR);
+
+An opened cdev_fd doesn't give the user any permission of accessing
+the device except binding the cdev_fd to an iommufd. After that point
+then the device is fully accessible including attaching it to an
+IOMMUFD IOAS/HWPT to enable userspace DMA::
+
+ struct vfio_device_bind_iommufd bind = {
+ .argsz = sizeof(bind),
+ .flags = 0,
+ };
+ struct iommu_ioas_alloc alloc_data = {
+ .size = sizeof(alloc_data),
+ .flags = 0,
+ };
+ struct vfio_device_attach_iommufd_pt attach_data = {
+ .argsz = sizeof(attach_data),
+ .flags = 0,
+ };
+ struct iommu_ioas_map map = {
+ .size = sizeof(map),
+ .flags = IOMMU_IOAS_MAP_READABLE |
+ IOMMU_IOAS_MAP_WRITEABLE |
+ IOMMU_IOAS_MAP_FIXED_IOVA,
+ .__reserved = 0,
+ };
+
+ iommufd = open("/dev/iommu", O_RDWR);
+
+ bind.iommufd = iommufd;
+ ioctl(cdev_fd, VFIO_DEVICE_BIND_IOMMUFD, &bind);
+
+ ioctl(iommufd, IOMMU_IOAS_ALLOC, &alloc_data);
+ attach_data.pt_id = alloc_data.out_ioas_id;
+ ioctl(cdev_fd, VFIO_DEVICE_ATTACH_IOMMUFD_PT, &attach_data);
+
+ /* Allocate some space and setup a DMA mapping */
+ map.user_va = (int64_t)mmap(0, 1024 * 1024, PROT_READ | PROT_WRITE,
+ MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
+ map.iova = 0; /* 1MB starting at 0x0 from device view */
+ map.length = 1024 * 1024;
+ map.ioas_id = alloc_data.out_ioas_id;;
+
+ ioctl(iommufd, IOMMU_IOAS_MAP, &map);
+
+ /* Other device operations as stated in "VFIO Usage Example" */
+
VFIO User API
-------------------------------------------------------------------------------
@@ -279,6 +410,7 @@ similar to a file operations structure::
struct iommufd_ctx *ictx, u32 *out_device_id);
void (*unbind_iommufd)(struct vfio_device *vdev);
int (*attach_ioas)(struct vfio_device *vdev, u32 *pt_id);
+ void (*detach_ioas)(struct vfio_device *vdev);
int (*open_device)(struct vfio_device *vdev);
void (*close_device)(struct vfio_device *vdev);
ssize_t (*read)(struct vfio_device *vdev, char __user *buf,
@@ -315,9 +447,10 @@ container_of().
- The [un]bind_iommufd callbacks are issued when the device is bound to
and unbound from iommufd.
- - The attach_ioas callback is issued when the device is attached to an
- IOAS managed by the bound iommufd. The attached IOAS is automatically
- detached when the device is unbound from iommufd.
+ - The [de]attach_ioas callback is issued when the device is attached to
+ and detached from an IOAS managed by the bound iommufd. However, the
+ attached IOAS can also be automatically detached when the device is
+ unbound from iommufd.
- The read/write/mmap callbacks implement the device region access defined
by the device's own VFIO_DEVICE_GET_REGION_INFO ioctl.
@@ -564,3 +697,11 @@ This implementation has some specifics:
\-0d.1
00:1e.0 PCI bridge: Intel Corporation 82801 PCI Bridge (rev 90)
+
+.. [5] Nested translation is an IOMMU feature which supports two stage
+ address translations. This improves the address translation efficiency
+ in IOMMU virtualization.
+
+.. [6] PASID stands for Process Address Space ID, introduced by PCI
+ Express. It is a prerequisite for Shared Virtual Addressing (SVA)
+ and Scalable I/O Virtualization (Scalable IOV).