XRT Native Library C++ API¶

Buffer APIs¶

class bo¶

Public Types

enum flags¶

buffer object flags

Values:

normal = 0¶: Create normal BO with host side and device side buffers

cacheable = XRT_BO_FLAGS_CACHEABLE¶: Create cacheable BO. Only effective on embedded platforms.

device_only = XRT_BO_FLAGS_DEV_ONLY¶: Create a BO with a device side buffer only

host_only = XRT_BO_FLAGS_HOST_ONLY¶: Create a BO with a host side buffer only

p2p = XRT_BO_FLAGS_P2P¶: Create a BO for peer-to-peer use

svm = XRT_BO_FLAGS_SVM¶

Create a BO for SVM (supported on specific platforms only)

The flags used by xrt::bo are compatible with XCL style flags as define in xrt_mem.h

Public Functions

bo()¶: bo() - Constructor for empty bo

bo(xclDeviceHandle dhdl, void *userptr, size_t sz, bo::flags flags, memory_group grp)¶

bo() - Constructor with user host buffer and flags

Parameters

dhdl: Device handle
userptr: Pointer to aligned user memory
sz: Size of buffer
flags: Specify special flags per xrt_mem.h
grp: Device memory group to allocate buffer in

bo(xclDeviceHandle dhdl, void *userptr, size_t sz, memory_group grp)¶

bo() - Constructor with user host buffer

The buffer type is default buffer object with host buffer and device buffer, where the host buffer is managed by user.

Parameters

dhdl: Device handle
userptr: Pointer to aligned user memory
sz: Size of buffer
grp: Device memory group to allocate buffer in

bo(xclDeviceHandle dhdl, size_t size, bo::flags flags, memory_group grp)¶

bo() - Constructor where XRT manages host buffer if any

If the flags require a host buffer, then the host buffer is allocated by XRT and can be accessed by using

map()

Parameters

dhdl: Device handle
size: Size of buffer
flags: Specify special flags per xrt_mem.h
grp: Device memory group to allocate buffer in

bo(xclDeviceHandle dhdl, size_t size, memory_group grp)¶

bo() - Constructor, default flags, where XRT manages host buffer if any

The buffer type is default buffer object with host buffer and device buffer. The host buffer is allocated and managed by XRT.

Parameters

dhdl: Device handle
size: Size of buffer
grp: Device memory group to allocate buffer in

bo(xclDeviceHandle dhdl, xclBufferExportHandle ehdl)¶

bo() - Constructor to import an exported buffer

If the exported buffer handle acquired by using the export() method is from another process, then it must be transferred through proper IPC mechanism translating the underlying file-descriptor asscociated with the buffer, see also constructor taking process id as argument.

Parameters

dhdl: Device that imports the exported buffer
ehdl: Exported buffer handle, implementation specific type

bo(xclDeviceHandle dhdl, pid_type pid, xclBufferExportHandle ehdl)¶

bo() - Constructor to import an exported buffer from another process

The exported buffer handle is obtained from exporting process by calling

export(). This contructor requires that XRT is built on and running on a system with pidfd support. Also the importing process must have permission to duplicate the exporting process’ file descriptor. This permission is controlled by ptrace access mode PTRACE_MODE_ATTACH_REALCREDS check (see ptrace(2)).

Parameters

dhdl: Device that imports the exported buffer
pid: Process id of exporting process
ehdl: Exported buffer handle, implementation specific type

bo(const bo &parent, size_t size, size_t offset)¶

bo() - Constructor for sub-buffer

Parameters

parent: Parent buffer
size: Size of sub-buffer
offset: Offset into parent buffer

bo(const bo &rhs)¶: bo() - Copy ctor

bo &operator=(const bo &rhs)¶: operator= () - Copy assignment

bo(bo &&rhs)¶: bo() - Move ctor

bo &operator=(bo &&rhs)¶: operator= () - Move assignment

bool operator==(const bo &rhs) const¶: operator== () - Compare operator

operator bool() const¶: operator bool() - Check if bo handle is valid

size_t size() const¶

size() - Get the size of this buffer

Return: Size of buffer in bytes

uint64_t address() const¶

address() - Get the device address of this buffer

Return: Device address of buffer

memory_group get_memory_group() const¶

get_memory_group() - Get the memory group in which this buffer is allocated

Return: Memory group index with which the buffer was constructed

flags get_flags() const¶

get_flags() - Get the flags with which this buffer was constructed

Return: The xrt::bo::Flgas used when the buffer was contructed

xclBufferExportHandle export_buffer()¶

buffer_export() - Export this buffer

An exported buffer can be imported on another device by this process or another process. For multiprocess transfer, the exported buffer must be transferred through a proper IPC facility to translate the underlying file-descriptor properly into another process.

Return: Exported buffer handle

The lifetime of the exported buffer handle is associated with the exporting buffer (this). The handle is disposed of when the exporting buffer is destructed.

It is undefined behavior to use the export handle after the exporting buffer object has gone out of scope.

async_handle async(xclBOSyncDirection dir, size_t sz, size_t offset)¶

async() - Start buffer content txfer with device side

Asynchronously transfer specified size bytes of buffer starting at specified offset.

Parameters

dir: To device or from device
sz: Size of data to synchronize
offset: Offset within the BO

async_handle async(xclBOSyncDirection dir)¶

async() - Start buffer content txfer with device side

Asynchronously transfer entire buffer content in specified direction

Parameters

dir: To device or from device
sz: Size of data to synchronize
offset: Offset within the BO

void sync(xclBOSyncDirection dir, size_t sz, size_t offset)¶

sync() - Synchronize buffer content with device side

Sync specified size bytes of buffer starting at specified offset.

Parameters

dir: To device or from device
sz: Size of data to synchronize
offset: Offset within the BO

void sync(xclBOSyncDirection dir)¶

sync() - Synchronize buffer content with device side

Sync entire buffer content in specified direction.

Parameters

dir: To device or from device

void *map()¶

map() - Map the host side buffer into application

Map the contents of the buffer object into host memory

Return: Memory mapped buffer

template<typename MapType> MapType map()¶

map() - Map the host side buffer info application

Return

Memory mapped buffer

Template Parameters

MapType: Type of mapped data

void write(const void *src, size_t size, size_t seek)¶

write() - Copy-in user data to host backing storage of BO

Copy source data to host buffer of this buffer object.

seek specifies how many bytes to skip at the beginning of the BO before copying-in size bytes to host buffer.

Parameters

src: Source data pointer
size: Size of data to copy
seek: Offset within the BO

If BO has no host backing storage, e.g. a device only buffer, then write is directly to the device buffer.

void write(const void *src)¶

write() - Copy-in user data to host backing storage of BO

Copy specified source data to host buffer of this buffer object.

Parameters

src: Source data pointer

If BO has no host backing storage, e.g. a device only buffer, then write is directly to the device buffer.

void read(void *dst, size_t size, size_t skip)¶

read() - Copy-out user data from host backing storage of BO

Copy content of host buffer of this buffer object to specified destination.

skip specifies how many bytes to skip from the beginning of the BO before copying-out size bytes of host buffer.

Parameters

dst: Destination data pointer
size: Size of data to copy
skip: Offset within the BO

If BO has no host backing storage, e.g. a device only buffer, then read is directly from the device buffer.

void read(void *dst)¶

read() - Copy-out user data from host backing storage of BO

Copy content of host buffer of this buffer object to specified destination.

Parameters

dst: Destination data pointer

If BO has no host backing storage, e.g. a device only buffer, then read is directly from the device buffer.

void copy(const bo &src, size_t sz, size_t src_offset = 0, size_t dst_offset = 0)¶

copy() - Deep copy BO content from another buffer

Throws if copy size is 0 or sz + src/dst_offset is out of bounds.

Parameters

src: Source BO to copy from
sz: Size of data to copy
src_offset: Offset into src buffer copy from
dst_offset: Offset into this buffer to copy to

void copy(const bo &src)¶

copy() - Deep copy BO content from another buffer

Copy full content of specified src buffer object to this buffer object

Parameters

src: Source BO to copy from

class async_handle : public detail::pimpl<async_handle_impl>¶

xrt::bo::async_handle represents an asynchronously operation

A handle object is returned from asynchronous buffer object operations. It can be used to wait for the operation to complete.

Configuration APIs¶

namespace ini¶

APIs for XRT configuration control.

XRT can be configured through a json xrt.ini file co-located with the host executable. If present, XRT uses configuration options from the ini file when a given option is first accessed. Without an ini file, the configuration options take on default values.

The APIs in this file allow host application to specify configuration options for XRT programatically. It is only possible for the host application to change configuration options before a given option is used by XRT the very first time.

Functions

void set(const std::string &key, const std::string &value)¶

set() - Change xrt.ini string value for specified key

Throws if key value cannot be changed.

Parameters

key: Key to change value for
value: New value for key

void set(const std::string &key, unsigned int value)¶

set() - Change xrt.ini string value for specified key

Throws if key value cannot be changed.

Parameters

key: Key to change value for
value: New value for key

Custom IP APIs¶

class ip : public detail::pimpl<ip_impl>¶

xrt::ip represent the custom IP

The ip can be controlled through read- and write register only. If the IP supports interrupt notification, then xrt::ip objects supports enabling and control of underlying IP interrupt.

In order to construct an ip object, the following requirements must be met:

The custom IP must appear in IP_LAYOUT section of xclbin.
The custom IP must have a base address such that it can be controlled through register access at offsets from base address.
The custom IP must have an address range so that write and read access to base address offset can be validated.
XRT supports exclusive access only for the custom IP, this is to other processes from accessing the same IP at the same time.

Public Functions

ip()¶: ip() - Construct empty ip object

ip(const xrt::device &device, const xrt::uuid &xclbin_id, const std::string &name)¶

ip() - Constructor from a device and xclbin

The IP is opened with exclusive access meaning that no other

xrt::ip objects can use the same IP, nor will another process be able to use the IP while one process has been granted access.

Parameters

device: Device programmed with the IP
xclbin_id: UUID of the xclbin with the IP
name: Name of IP to construct

Constructor throws on error.

void write_register(uint32_t offset, uint32_t data)¶

write_register() - Write to the address range of an ip

Throws std::out_or_range if offset is outside the ip address space

Parameters

offset: Offset in register space to write to
data: Data to write

uint32_t read_register(uint32_t offset) const¶

read_register() - Read data from ip address range

Throws std::out_or_range if offset is outside the ip address space

Return

Value read from offset

Parameters

offset: Offset in register space to read from

interrupt create_interrupt_notify()¶

create_interrupt_notify() - Create xrt::ip::interrupt object

This function creates an

xrt::ip::interrupt object that can be used to control and wait for IP interrupt. On successful return the IP has interrupt enabled.

Return: xrt::ip::interrupt object can be used to control IP interrupt.

Throws if the custom IP doesn’t support interrupts.

class interrupt : public detail::pimpl<interrupt_impl>¶

xrt::ip::interrupt represents an IP interrupt event.

This class represents an IP interrupt event. The interrupt object is contructed via xrt::ip::create_interrupt_notify(). The object can be used to enable and disable IP interrupts and to wait for an interrupt to occur.

Upon construction, the IP interrupt is automatically enabled.

Public Functions

void enable()¶

enable() - Enable interrupt

Enables the IP interrupt if not already enabled. The IP interrupt is automatically enabled when the interrupt object is created.

void disable()¶

disable() - Disable interrupt

Disables the IP interrupt notification from IP.

void wait()¶

wait() - Wait for interrupt

Wait for interrupt from IP. Upon return, interrupt is re-enabled.

std::cv_status wait(const std::chrono::milliseconds &timeout) const¶

wait() - Wait for interrupt or timeout to occur

Blocks the current thread until an interrupt is received from the IP, or until after the specified timeout duration

Return

std::cv_status::timeout if the timeout specified expired, std::cv_status::no_timeout otherwise.

Parameters

timeout: Timout in milliseconds.

Device APIs¶

class device¶

xrt::device represents used for acceleration.

Public Functions

device()¶: device() - Constructor for empty device

~device()¶: device() - Dtor

device(unsigned int didx)¶

device() - Constructor from device index

Throws if no device is found matching the specified index.

Parameters

didx: Device index

device(const std::string &bdf)¶

device() - Constructor from string

If the string is in BDF format it matched against devices installed on the system. Otherwise the string is assumed to be a device index.

Parameters

bdf: String identifying the device to open.

Throws if string format is invalid or no matching device is found.

device(xclDeviceHandle dhdl)¶

device() - Create a managed device object from a shim xclDeviceHandle

Return

xrt::device object epresenting the opened device, or exception on error

Parameters

dhdl: Shim xclDeviceHandle

device(const device &rhs)¶: device() - Copy ctor

device &operator=(const device &rhs)¶: operator= () - Move assignment

device(device &&rhs)¶: device() - Move ctor

device &operator=(device &&rhs)¶: operator= () - Move assignment

template<info::device param> info::param_traits<info::device, param>::return_type get_info() const¶

get_info() - Retrieve device parameter information

This function is templated on the enumeration value as defined in the enumeration xrt::info::device.

The return type of the parameter is based on the instantiated param_traits for the given param enumeration supplied as template argument, see namespace xrt::info

This function guarantees return values conforming to the format used by the time the application was built and for a two year period minimum. In other words, XRT can be updated to new versions without affecting the format of the return type.

uuid load_xclbin(const axlf *xclbin)¶

load_xclbin() - Load an xclbin

Return

UUID of argument xclbin

Parameters

xclbin: Pointer to xclbin in memory image

uuid load_xclbin(const std::string &xclbin_fnm)¶

load_xclbin() - Read and load an xclbin file

This function reads the file from disk and loads the xclbin. Using this function allows one time allocation of data that needs to be kept in memory.

Return

UUID of argument xclbin

Parameters

xclbin_fnm: Full path to xclbin file

uuid load_xclbin(const xrt::xclbin &xclbin)¶

load_xclbin() - load an xclin from an xclbin object

This function uses the specified

xrt::xclbin object created by caller. The xrt::xclbin object must contain the complete axlf structure.

Return

UUID of argument xclbin

Parameters

xclbin: xrt::xclbin object

uuid get_xclbin_uuid() const¶

get_xclbin_uuid() - Get UUID of xclbin image loaded on device

Note that current UUID can be different from the UUID of the xclbin loaded by this process using

load_xclbin()

Return: UUID of currently loaded xclbin

template<typename SectionType> SectionType get_xclbin_section(axlf_section_kind section, const uuid &uuid) const¶

get_xclbin_section() - Retrieve specified xclbin section

Get the xclbin section of the xclbin currently loaded on the device. The function throws on error

Return

The specified section if available.

Parameters

section: The section to retrieve
uuid: Xclbin uuid of the xclbin with the section to retrieve

Note, this API may be replaced with more generic query request access

Info APIs¶

namespace info¶

Enums

enum device¶

Device information parameters.

Use with xrt::device::get_info() to retrieve properties of the device. The type of the device properties is compile time defined with param traits.

Values:

bdf¶: BDF for device (std::string)

interface_uuid¶: Interface UUID when device is programmed with 2RP shell (xrt::uuid)

kdma¶: Number of KDMA engines (std::uint32_t)

max_clock_frequency_mhz¶: Max clock frequency (unsigned long)

m2m¶: True if device contains m2m (bool)

name¶: Name (VBNV) of device (std::string)

nodma¶: True if device is a NoDMA device (bool)

offline¶: True if device is offline and in process of being reset (bool)

electrical¶: Electrical and power sensors present on the device (std::string)

thermal¶: Thermal sensors present on the device (std::string)

mechanical¶: Mechanical sensors on and surrounding the device (std::string)

memory¶: Memory information present on the device (std::string)

platform¶: Platforms flashed on the device (std::string)

pcie_info¶: Pcie information of the device (std::string)

host¶: Host information (std::string)

aie¶: AIE core information of the device (std::string)

aie_shim¶: AIE shim information of the device (std::string)

dynamic_regions¶: Information about xclbin on the device (std::string)

vmr¶: Information about vmr on the device (std::string)

Kernel APIs¶

class kernel¶

A kernel object represents a set of instances matching a specified name. The kernel is created by finding matching kernel instances in the currently loaded xclbin.

Most interaction with kernel objects are through xrt::run objects created from the kernel object to represent an execution of the kernel

Public Functions

kernel()¶: kernel() - Construct for empty kernel

kernel(const xrt::device &device, const xrt::uuid &xclbin_id, const std::string &name, cu_access_mode mode = cu_access_mode::shared)¶

kernel() - Constructor from a device and xclbin

The kernel name must uniquely identify compatible kernel instances (compute units). Optionally specify which kernel instance(s) to open using “kernelname:{instancename1,instancename2,…}” syntax. The compute units are default opened with shared access, meaning that other kernels and other process will have shared access to same compute units.

Parameters

device: Device on which the kernel should execute
xclbin_id: UUID of the xclbin with the kernel
name: Name of kernel to construct
mode: Open the kernel instances with specified access (default shared)

template<typename ...Args> run operator()(Args&&... args)¶

operator() - Invoke the kernel function

Return

Run object representing this kernel function invocation

Parameters

args: Kernel arguments

int group_id(int argno) const¶

group_id() - Get the memory bank group id of an kernel argument

The function throws if the group id is ambigious.

Return

The memory group id to use when allocating buffers (see xrt::bo)

Parameters

argno: The argument index

uint32_t offset(int argno) const¶

offset() - Get the offset of kernel argument

Use with

read_register() and write_register() if manually reading or writing kernel registers for explicit arguments.

Return

The kernel register offset of the argument with specified index

Parameters

argno: The argument index

void write_register(uint32_t offset, uint32_t data)¶

write() - Write to the address range of a kernel

Throws std::out_or_range if offset is outside the kernel address space

Parameters

offset: Offset in register space to write to
data: Data to write

The kernel must be associated with exactly one kernel instance (compute unit), which must be opened for exclusive access.

uint32_t read_register(uint32_t offset) const¶

read() - Read data from kernel address range

Throws std::out_or_range if offset is outside the kernel address space

Return

Value read from offset

Parameters

offset: Offset in register space to read from

The kernel must be associated with exactly one kernel instance (compute unit), which must be opened for exclusive access.

std::string get_name() const¶: get_name() - Return the name of the kernel

xrt::xclbin get_xclbin() const¶: get_xclbin() - Return the xclbin containing the kernel

class run¶

xrt::run represents one execution of a kernel

The run handle can be explicitly constructed from a kernel object or implicitly constructed from starting a kernel execution.

A run handle can be re-used to execute the same kernel again.

Public Functions

run()¶

run() - Construct empty run object

Can be used as lvalue in assignment.

run(const kernel &krnl)¶

run() - Construct run object from a kernel object

Parameters

krnl: Kernel object representing the kernel to execute

void start()¶

start() - Start one execution of a run.

This function is asynchronous, run status must be expclicit checked or wait() must be used to wait for the run to complete.

void start(const autostart &iterations)¶

start() - Start auto-restart execution of a run

An iteration count of zero means that the kernel should run forever, or until explicitly stopped using

stop().

Parameters

iterations: Number of times to iterate the same run.

This function is asynchronous, run status must be expclicit checked or wait() must be used to wait for the run to complete.

The kernel run object is complete only after all iterations have completed, or until run object has been explicitly stopped.

Changing kernel arguments set_arg() while the kernel is running has undefined behavior. To synchronize change of arguments, please use xrt::mailbox.

Currently autostart is only supported for kernels with one compute unit which must be opened in exclusive mode.

void stop()¶

stop() - Stop kernel run object at next safe iteration

If the kernel run object has been started by specifying an iteration count or by specifying default iteration count, then this function can be used to stop the iteration early.

The function is synchronous and waits for the kernel run object to complete.

If the kernel is not iterating, then calling this funciton is the same as calling wait().

ert_cmd_state abort()¶

abort() - Abort a run object that has been started

If the run object has been sent to scheduler for execution, then this function can be used to abort the scheduled command.

Return: State of aborted command

The function is synchronous and will wait for abort to complete. The return value is the state of the aborted command.

ert_cmd_state wait(const std::chrono::milliseconds &timeout = std::chrono::milliseconds{0}) const¶

wait() - Wait for a run to complete execution

The default timeout of 0ms indicates blocking until run completes.

Return

Command state upon return of wait, or ERT_CMD_STATE_TIMEOUT if timeout exceeded.

Parameters

timeout: Timeout for wait (default block till run completes)

The current thread will block until the run completes or timeout expires. Completion does not guarantee success, the run status should be checked by using state.

If specified time out is exceeded, the function returns with ERT_CMD_STATE_TIMEOUT, it is the callers responsibility to abort the run if it continues to time out.

The current implementation of this API can mask out the timeout of this run so that the call either never returns or doesn’t return until the run completes by itself. This can happen if other runs are continuosly completing within the specified timeout for this run. If the device is otherwise idle, or if the time between run completion exceeds the specified timeout, then this function will identify the timeout.

ert_cmd_state wait(unsigned int timeout_ms) const¶

wait() - Wait for specified milliseconds for run to complete

The default timeout of 0ms indicates blocking until run completes.

Return

Command state upon return of wait, or ERT_CMD_STATE_TIMEOUT if timeout exceeded.

Parameters

timeout_ms: Timeout in milliseconds

The current thread will block until the run completes or timeout expires. Completion does not guarantee success, the run status should be checked by using state.

If specified time out is exceeded, the function returns with ERT_CMD_STATE_TIMEOUT, it is the callers responsibility to abort the run if it continues to time out.

The current implementation of this API can mask out the timeout of this run so that the call either never returns or doesn’t return until the run completes by itself. This can happen if other runs are continuosly completing within the specified timeout for this run. If the device is otherwise idle, or if the time between run completion exceeds the specified timeout, then this function will identify the timeout.

ert_cmd_state state() const¶

state() - Check the current state of a run object

The state values are defined in

include/ert.h

Return: Current state of this run object

void add_callback(ert_cmd_state state, std::function<void(const void *, ert_cmd_state, void *)> callback, void *data, )¶

add_callback() - Add a callback function for run state

The function is called when the run object changes state to argument state or any error state. Only

ERT_CMD_STATE_COMPLETED is supported currently.

Parameters

state: State to invoke callback on
callback: Callback function
data: User data to pass to callback function

The function object’s first parameter is a unique ‘key’ for this xrt::run object implmentation on which the callback was added. This ‘key’ can be used to identify an actual run object that refers to the implementaion that is maybe shared by multiple xrt::run objects.

Any number of callbacks are supported.

operator bool() const¶

operator bool() - Check if run handle is valid

Return: True if run is associated with kernel object, false otherwise

bool operator<(const xrt::run &rhs) const¶

operator < () - Weak ordering

Return

True if object is ordered less that compared with other

Parameters

rhs: Object to compare with

template<typename ArgType> void set_arg(int index, ArgType &&arg)¶

set_arg() - Set a specific kernel scalar argument for this run

Use this API to explicit set or change a kernel argument prior to starting kernel execution. After setting arguments, the kernel can be started using

start() on the run object.

Parameters

index: Index of kernel argument to update
arg: The scalar argument value to set.

See also operator() to set all arguments and start kernel.

void set_arg(int index, xrt::bo &boh)¶

set_arg() - Set a specific kernel global argument for a run

Use this API to explicit set or change a kernel argument prior to starting kernel execution. After setting arguments, the kernel can be started using

start() on the run object.

Parameters

index: Index of kernel argument to set
boh: The global buffer argument value to set (lvalue).

See also operator() to set all arguments and start kernel.

void set_arg(int index, const xrt::bo &boh)¶: set_arg - xrt::bo variant for const lvalue

void set_arg(int index, xrt::bo &&boh)¶: set_arg - xrt::bo variant for rvalue

template<typename ArgType> void set_arg(const std::string &argnm, ArgType &&argvalue)¶

set_arg - set named argument

Throws if specified argument name doesn’t match kernel specification. Throws if argument value is incompatible with specified argument

Parameters

argnm: Name of kernel argument
argvalue: Argument value

template<typename ArgType> void update_arg(int index, ArgType &&arg)¶

udpdate_arg() - Asynchronous update of scalar kernel global argument

Use this API to asynchronously update a specific scalar argument of the kernel associated with the run object.

Parameters

index: Index of kernel argument to update
arg: The scalar argument value to set.

This API is only supported on Edge.

void update_arg(int index, const xrt::bo &boh)¶

update_arg() - Asynchronous update of kernel global argument for a run

Use this API to asynchronously update a specific kernel argument of an existing run.

Parameters

index: Index of kernel argument to update
boh: The global buffer argument value to set.

This API is only supported on Edge.

template<typename ...Args> void operator()(Args&&... args)¶

operator() - Set all kernel arguments and start the run

Use this API to explicitly set all kernel arguments and start kernel execution.

Parameters

args: Kernel arguments

template<typename ...Args> void operator()(autostart &&count, Args&&... args)¶

operator() - Set all kernel arguments and start the run

Use this API to explicitly set all kernel arguments and start kernel execution for specified number of iterations.

Parameters

count: Iteration count specifying number of iterations of the run
args: Kernel arguments

An iteration count of ‘1’ invokes the kernel once and is the same as calling the operator without specifying autostart.

The run is complete only after all iterations have completed or when the kernel has been explicitly stopped using stop().

Currently autostart is only supported for kernels with one compute unit which must be opened in exclusive mode.

Message APIs¶

namespace message¶

APIs for XRT messaging.

XRT internally uses a message system that supports dispatching of messages to null, console, file, or syslog under different verbosity levels. The sink and verbosity level is controlled statically through xrt.ini or at run-time using xrt::ini.

The APIs in this file allow host application to use the same message dispatch mechanism as XRT is configured to use.

Enums

enum level¶

Verbosity level for messages.

Use logging APIs to control at what verbosity level the messages should be issued. The default verbosity can be changed in xrt.ini or programatically by using xrt::ini::set.

Values:

emergency = xrtLogMsgLevel::XRT_EMERGENCY¶

alert = xrtLogMsgLevel::XRT_ALERT¶

critical = xrtLogMsgLevel::XRT_CRITICAL¶

error = xrtLogMsgLevel::XRT_ERROR¶

warning = xrtLogMsgLevel::XRT_WARNING¶

notice = xrtLogMsgLevel::XRT_NOTICE¶

info = xrtLogMsgLevel::XRT_INFO¶

debug = xrtLogMsgLevel::XRT_DEBUG¶

Functions

void log(level lvl, const std::string &tag, const std::string &msg)¶

log() - Dispatch composed log message

Parameters

lvl: Severity level, the message is ignored if configured level is less than specified level.
tag: The message tag to use.
msg: A formatted composed message

template<typename ...Args> void logf(level lvl, const std::string &tag, const char *format, Args&&... args)¶

logf() - Compose and dispatch formatted log message

This log function uses boost::format to compose the message from specified format string and arguments.

Parameters

lvl: Severity level, the message is ignored if configured level is less than specified level.
tag: The message tag to use.
format: A format string similar to printf or boost::format
args: Message arguments for the placeholders used in the format string

System APIs¶

namespace system¶

APIs for system level queries and control.

Functions

unsigned int enumerate_devices()¶

enumerate_devices() - Enumerate devices found in the system

Return: Number of devices in the system recognized by XRT

UUID APIs¶

class uuid¶

Wrapper class to treat uuid_t as a value type supporting copying.

xrt::uuid is used by many XRT APIs to match an expected xclbin against current device xclbin, or to get the uuid of current loaded shell on the device.

Public Functions

uuid()¶: uuid() - Construct cleared uuid

uuid(const xuid_t val)¶

uuid() - Construct uuid from a basic bare uuid

A basic uuid is either a uuid_t on Linux, or a typedef of equivalent basic type of other platforms

Parameters

val: The basic uuid to construct this object from

uuid(const std::string &uuid_str)¶

uuid() - Construct uuid from a string representaition

A uuid string is 36 bytes with ‘-‘ at 8, 13, 18, and 23

Parameters

uuid_str: A string formatted as a uuid string

uuid(const uuid &rhs)¶

uuid() - copy constructor

Parameters

rhs: Value to be copied

uuid &operator=(const uuid &rhs)¶

operator=() - assignment

Return

Reference to this

Parameters

rhs: Value to be assigned from

const xuid_t &get() const¶

get() - Get the underlying basis uuid type value

A basic uuid is either a uuid_t on Linux, or a typedef of equivalent basic type of other platforms

Return: Basic uuid value

std::string to_string() const¶

to_string() - Convert to string

Return: Lower case string representation of this uuid

operator bool() const¶

bool() - Conversion operator

Return: True if this uuid is not null

bool operator==(const xuid_t &xuid) const¶

operator==() - Compare to basic uuid

A basic uuid is either a uuid_t on Linux, or a typedef of equivalent basic type of other platforms

Return

True if equal, false otherwise

Parameters

xuid: Basic uuid to compare against

bool operator!=(const xuid_t &xuid) const¶

operator!=() - Compare to basic uuid

A basic uuid is either a uuid_t on Linux, or a typedef of equivalent basic type of other platforms

Return

False if equal, true otherwise

Parameters

xuid: Basic uuid to compare against

bool operator==(const uuid &rhs) const¶

operator==() - Comparison

Return

True if equal, false otherwise

Parameters

rhs: uuid to compare against

bool operator!=(const uuid &rhs) const¶

operator!=() - Comparison

Return

False if equal, true otherwise

Parameters

rhs: uuid to compare against

bool operator<(const uuid &rhs) const¶

operator<() - Comparison

Return

True of this is less that argument uuid, false otherwise

Parameters

rhs: uuid to compare against

XCLBIN APIs¶

class xclbin : public detail::pimpl<xclbin_impl>¶

xrt::xclbin represents an xclbin and provides APIs to access meta data.

The xclbin object is constructed by the user from a file.

When the xclbin object is constructed from a complete xclbin, then it can be used by xrt::device to program the xclbin onto the device.

First-class objects and class navigation

All meta data is rooted at xrt::xclbin.

_images/xclbin_navigation.png

From the xclbin object xrt::xclbin::kernel or xrt::xclbin::ip objects can be constructed.

The xrt:xclbin::kernel is a concept modelled only in the xclbin XML metadata, it corresponds to a function that can be executed by one or more compute units modelled by xrt::xclbin::ip objects. An xrt::xclbin::ip object corresponds to an entry in the xclbin IP_LAYOUT section, so the xrt::xclbin::kernel object is just a grouping of one or more of these.

In some cases the kernel concept is not needed, thus xrt::xclbin::ip objects corresponding to entries in the xclbin IP_LAYOUT sections can be accessed directly.

An xrt::xclbin::arg object corresponds to one or more entries in the xclbin CONNECTIVITY section decorated with additional meta data (offset, size, type, etc) from the XML section if available. An argument object represents a specific kernel or ip argument. If the argument is a global buffer, then it may connect to one or more memory objects.

Finally the xrt::xclbin::mem object corresponds to an entry in the MEM_TOPOLOGY section of the xclbin.

Public Types

enum target_type¶

Type of xclbin.

See xclbin.h

Values:

hw¶

sw_emu¶

hw_emu¶

Public Functions

xclbin()¶: xclbin() - Construct empty xclbin object

xclbin(const std::string &filename)¶

xclbin() - Constructor from an xclbin filename

Throws if file not found.

Parameters

filename: Path to the xclbin file

xclbin(const std::vector<char> &data)¶

xclbin() - Constructor from raw data

The raw data of the xclbin can be deleted after calling the constructor.

Parameters

data: Raw data of xclbin

xclbin(const axlf *top)¶

xclbin() - Constructor from raw data

The argument axlf is copied by the constructor.

Parameters

top: Raw data of xclbin file as axlf*

std::vector<kernel> get_kernels() const¶

get_kernels() - Get list of kernels from xclbin.

Kernels are extracted from embedded XML metadata in the xclbin. A kernel groups one or more compute units. A kernel has arguments from which offset, type, etc can be retrived.

Return: A list of xrt::xclbin::kernel from xclbin.

kernel get_kernel(const std::string &name) const¶

get_kernel() - Get a kernel by name from xclbin

A matching kernel is extracted from embedded XML metadata in the xclbin. A kernel groups one or more compute units. A kernel has arguments from which offset, type, etc can be retrived.

Return

The matching kernel from the xclbin or error if no matching kernel is found.

Parameters

name: Name of kernel to get.

std::vector<ip> get_ips() const¶

get_ips() - Get a list of IPs from the xclbin

The returned

xrt::xclbin::ip objects are extracted from the IP_LAYOUT section of the xclbin.

Return: A list of xrt::xclbin::ip objects from xclbin.

std::vector<ip> get_ips(const std::string &name) const¶

get_ips() - Get list of ips that matches name

The kernel name can optionally specify which kernel instance(s) to match “kernel:{ip1,ip2,…} syntax.

Return

A list of xrt::xclbin::ip objects that are compute units of this kernel object and matches the specified name.

Parameters

name: Name to match against, prefixed with kernel name

ip get_ip(const std::string &name) const¶

get_ip() - Get a specific IP from the xclbin

The returned

xrt::xclbin::ip object is extracted from the IP_LAYOUT section of the xclbin.

Return: A list of xrt::xclbin::ip objects from xclbin.

std::vector<mem> get_mems() const¶

get_mems() - Get list of memory objects

The returned

xrt::xclbin::mem objects are extracted from the xclbin.

Return: A list of xrt::xclbin::mem objects from xclbin

std::string get_xsa_name() const¶

get_xsa_name() - Get Xilinx Support Archive (XSA) name of xclbin

An exception is thrown if the data is missing.

Return: Name of XSA (vbnv name).

std::string get_fpga_device_name() const¶

get_fpga_device_name() - Get FPGA device name

Return: Name of fpga device per XML metadata.

uuid get_uuid() const¶

get_uuid() - Get the uuid of the xclbin

An exception is thrown if the data is missing.

Return: UUID of xclbin

target_type get_target_type() const¶

get_target_type() - Get the type of this xclbin

Return: Target type, which can be hw, sw_emu, or hw_emu

class arg : public detail::pimpl<arg_impl>¶

class arg - xrt::xclbin::arg represents a compute unit argument

The argument object constructed from the xclbin connectivity section. An argument is connected to a memory bank or a memory group, which dictates where in device memory a global buffer used with this kernel argument must be allocated.

Public Functions

std::string get_name() const¶

get_name() - Get argument name

Return: Name of argument.

std::vector<mem> get_mems() const¶

get_mems() - Get list of device memories from xclbin.

Return: A list of xrt::xclbin::mem objects to which this argument is connected.

std::string get_port() const¶

get_port() - Get port name of this argument

Return: Port name

uint64_t get_size() const¶

get_size() - Argument size in bytes

Return: Argument size

uint64_t get_offset() const¶

get_offset() - Argument offset

Return: Argument offset

std::string get_host_type() const¶

get_host_type() - Get the argument host type

Return: Argument host type

size_t get_index() const¶

get_index() - Get the index of this argument

Return: Argument index

class ip : public detail::pimpl<ip_impl>¶

xrt::xclbin::ip represents a IP in an xclbin.

The ip corresponds to an entry in the IP_LAYOUT section of the xclbin.

Public Types

enum control_type¶

See xclbin.h

Values:

hs = 0¶

chain = 1¶

none = 2¶

fa = 5¶

enum ip_type¶

See xclbin.h

Values:

pl = IP_KERNEL¶

ps = IP_PS_KERNEL¶

Public Functions

std::string get_name() const¶

get_name() - Get name of IP

Return: IP name.

ip_type get_type() const¶

get_type() - Get the IP type

Return: IP type

control_type get_control_type() const¶

get_control_type() - Get the IP control protocol

Return: Control type

size_t get_num_args() const¶

get_num_args() - Number of arguments

Return: Number of arguments for this IP per CONNECTIVITY section

std::vector<arg> get_args() const¶

get_args() - Get list of IP arguments

An argument may have multiple memory connections

Return: A list sorted of xclbin::arg sorted by argument indices

arg get_arg(int32_t index) const¶

get_arg() - Get argument at index.

The argument may have multiple memory connections

Return

The argument a specified index

Parameters

index: Index of argument

uint64_t get_base_address() const¶

get_base_address() - Get the base address of the cu

Return: The base address of the IP

size_t get_size() const¶

get_size() - Get the address range size of this IP.

The address range is a property of the kernel and as such only valid for for kernel compute units.

Return: The size of this IP

For IPs that are not associated with a kernel, the size return is 0.

class kernel : public detail::pimpl<kernel_impl>¶

Public Functions

std::string get_name() const¶

get_name() - Get kernel name

Return: The name of the kernel

std::vector<ip> get_cus() const¶

get_cus() - Get list of cu from kernel.

Return: A list of xrt::xclbin::ip objects corresponding the compute units for this kernel object.

std::vector<ip> get_cus(const std::string &name) const¶

get_cus() - Get list of compute units that matches name

The kernel name can optionally specify which kernel instance(s) to match “kernel:{cu1,cu2,…} syntax.

Return

A list of xrt::xclbin::ip objects that are compute units of this kernel object and matches the specified name.

Parameters

name: Name to match against, prefixed with kernel name

ip get_cu(const std::string &name) const¶

get_cu() - Get compute unit by name

Return: The xct::xclbin::ip object matching the specified name, or error if not present.

size_t get_num_args() const¶

get_num_args() - Number of arguments

Return: Number of arguments for this kernel.

std::vector<arg> get_args() const¶

get_args() - Get list of kernel arguments

An argument may have multiple memory connections

Return: A list sorted of xclbin::arg sorted by argument indices

arg get_arg(int32_t index) const¶

get_arg() - Get kernel argument at index.

The memory connections of an argument is the union of the connections for each compute unit for that particular argument. In other words, for each connection of the argument returned by

get_arg() there is at least one compute unit that has that connection.

Return: The xrt::xclbin::arg object at specified argument index.

class mem : public detail::pimpl<mem_impl>¶

xrt::xclbin::mem represents a physical device memory bank

A memory object is constructed from an entry in the MEM_TOPOLOGY section of an xclbin.

Public Types

enum memory_type¶

type of memory

See xclbin.h

Values:

ddr3 = MEM_DDR3¶

ddr4 = MEM_DDR4¶

dram = MEM_DRAM¶

streaming = MEM_STREAMING¶

preallocated_global = MEM_PREALLOCATED_GLOB¶

are = MEM_ARE¶

hbm = MEM_HBM¶

bram = MEM_BRAM¶

uram = MEM_URAM¶

streaming_connection = MEM_STREAMING_CONNECTION¶

host = MEM_HOST¶

Public Functions

std::string get_tag() const¶

get_name() - Get tag name

Return: Memory tag name

uint64_t get_base_address() const¶

get_base_address() - Get the base address of the memory bank

Return: Base address of the memory bank, or -1 for invalid base address

uint64_t get_size_kb() const¶

get_size() - Get the size of the memory in KB

Return: Size of memory in KB, or -1 for invalid size

bool get_used() const¶

get_used() - Get used status of the memory

A value of false indicates that no buffer can be allocated in this memory bank.

Return: True of this memory bank is used by kernels in the xclbin or false otherwise.

memory_type get_type() const¶

get_type() - Get the type of the memory

Return: Memory type

int32_t get_index() const¶

get_index() - Get the index of the memory

The returned index can be used when allocating buffers using

xrt::bo provided the memory bank is connected / used.

Return: Index of the memory within the memory topology

XRT Native Library C API¶

Buffer APIs¶

typedef xrtDeviceHandle¶: opaque device handle

typedef xrtBufferHandle¶: opaque buffer handle

typedef xrtBufferFlags¶: flags for BO

Description

See xrt_mem.h for available flags

typedef xrtMemoryGroup¶: Memory bank group for buffer

XCL_DRIVER_DLLESPEC xrtBufferHandle xrtBOAllocUserPtr(xrtDeviceHandle dhdl, void * userptr, size_t size, xrtBufferFlags flags, xrtMemoryGroup grp)¶: Allocate a BO using userptr provided by the user

Parameters

xrtDeviceHandle dhdl: Device handle
void * userptr: Pointer to 4K aligned user memory
size_t size: Size of buffer
xrtBufferFlags flags: Specify type of buffer
xrtMemoryGroup grp: Specify bank information

Return

xrtBufferHandle on success or NULL

XCL_DRIVER_DLLESPEC xrtBufferHandle xrtBOAlloc(xrtDeviceHandle dhdl, size_t size, xrtBufferFlags flags, xrtMemoryGroup grp)¶: Allocate a BO of requested size with appropriate flags

Parameters

xrtDeviceHandle dhdl: Device handle
size_t size: Size of buffer
xrtBufferFlags flags: Specify type of buffer
xrtMemoryGroup grp: Specify bank information

Return

xrtBufferHandle on success or NULL

XCL_DRIVER_DLLESPEC xrtBufferHandle xrtBOImport(xrtDeviceHandle dhdl, xclBufferExportHandle ehdl)¶: Allocate a BO imported from another device

Parameters

xrtDeviceHandle dhdl: Device that imports the exported buffer
xclBufferExportHandle ehdl: Exported buffer handle, implementation specific type

Description

The exported buffer handle is acquired by using the export() method and can be passed to another process.

XCL_DRIVER_DLLESPEC xclBufferExportHandle xrtBOExport(xrtBufferHandle bhdl)¶: Export this buffer

Parameters

xrtBufferHandle bhdl: Buffer handle

Return

Exported buffer handle

An exported buffer can be imported on another device by this process or another process.

XCL_DRIVER_DLLESPEC xrtBufferHandle xrtBOSubAlloc(xrtBufferHandle parent, size_t size, size_t offset)¶: Allocate a sub buffer from a parent buffer

Parameters

xrtBufferHandle parent: Parent buffer handle
size_t size: Size of sub buffer
size_t offset: Offset into parent buffer

Return

xrtBufferHandle on success or NULL

XCL_DRIVER_DLLESPEC int xrtBOFree(xrtBufferHandle bhdl)¶: Free a previously allocated BO

Parameters

xrtBufferHandle bhdl: Buffer handle

Return

0 on success, or err code on error

XCL_DRIVER_DLLESPEC size_t xrtBOSize(xrtBufferHandle bhdl)¶: Get the size of this buffer

Parameters

xrtBufferHandle bhdl: Buffer handle

Return

Size of buffer in bytes

XCL_DRIVER_DLLESPEC uint64_t xrtBOAddress(xrtBufferHandle bhdl)¶: Get the physical address of this buffer

Parameters

xrtBufferHandle bhdl: Buffer handle

Return

Device address of this BO, or LLONG_MAX on error

XCL_DRIVER_DLLESPEC int xrtBOSync(xrtBufferHandle bhdl, enum xclBOSyncDirection dir, size_t size, size_t offset)¶: Synchronize buffer contents in requested direction

Parameters

xrtBufferHandle bhdl: Bufferhandle
enum xclBOSyncDirection dir: To device or from device
size_t size: Size of data to synchronize
size_t offset: Offset within the BO

Return

0 on success or error

Synchronize the buffer contents between host and device. Depending on the memory model this may require DMA to/from device or CPU cache flushing/invalidation

XCL_DRIVER_DLLESPEC void* xrtBOMap(xrtBufferHandle bhdl)¶: Memory map BO into user’s address space

Parameters

xrtBufferHandle bhdl: Buffer handle

Return

Memory mapped buffer, or NULL on error

Map the contents of the buffer object into host memory. The buffer object is unmapped when freed.

XCL_DRIVER_DLLESPEC int xrtBOWrite(xrtBufferHandle bhdl, const void * src, size_t size, size_t seek)¶: Copy-in user data to host backing storage of BO

Parameters

xrtBufferHandle bhdl: Buffer handle
const void * src: Source data pointer
size_t size: Size of data to copy
size_t seek: Offset within the BO

Return

0 on success or appropriate error number

Copy host buffer contents to previously allocated device memory. seek specifies how many bytes to skip at the beginning of the BO before copying-in size bytes of host buffer.

XCL_DRIVER_DLLESPEC int xrtBORead(xrtBufferHandle bhdl, void * dst, size_t size, size_t skip)¶: Copy-out user data from host backing storage of BO

Parameters

xrtBufferHandle bhdl: Buffer handle
void * dst: Destination data pointer
size_t size: Size of data to copy
size_t skip: Offset within the BO

Return

0 on success or appropriate error number

Copy contents of previously allocated device memory to host buffer. skip specifies how many bytes to skip from the beginning of the BO before copying-out size bytes of device buffer.

XCL_DRIVER_DLLESPEC int xrtBOCopy(xrtBufferHandle dst, xrtBufferHandle src, size_t sz, size_t dst_offset, size_t src_offset)¶: Deep copy BO content from another buffer

Parameters

xrtBufferHandle dst: Destination BO to copy to
xrtBufferHandle src: Source BO to copy from
size_t sz: Size of data to copy
size_t dst_offset: Offset into destination buffer to copy to
size_t src_offset: Offset into src buffer to copy from

Return

0 on success or appropriate error number

It is an error if sz is 0 bytes or sz + src/dst_offset is out of bounds.

Configuration APIs¶

XCL_DRIVER_DLLESPEC int xrtIniStringSet(const char * key, const char * value)¶: Change xrt.ini string value for specified key

Parameters

const char * key: Key to change value for
const char * value: New value for key

Return

0 on success, error if key value cannot be changed

XCL_DRIVER_DLLESPEC int xrtIniUintSet(const char * key, unsigned int value)¶: Change xrt.ini unsigned int value for specified key

Parameters

const char * key: Key to change value for
unsigned int value: New value for key

Return

0 on success, error if key value cannot be changed

Device and XCLBIN APIs¶

typedef xrtDeviceHandle: opaque device handle

XCL_DRIVER_DLLESPEC xrtDeviceHandle xrtDeviceOpen(unsigned int index)¶: Open a device and obtain its handle

Parameters

unsigned int index: Device index

Return

Handle representing the opened device, or nullptr on error

XCL_DRIVER_DLLESPEC xrtDeviceHandle xrtDeviceOpenByBDF(const char * bdf)¶: Open a device and obtain its handle

Parameters

const char * bdf: PCIe BDF identifying the device to open

Return

Handle representing the opened device, or nullptr on error

XCL_DRIVER_DLLESPEC xrtDeviceHandle xrtDeviceOpenFromXcl(xclDeviceHandle xhdl)¶: Open a device from a shim xclDeviceHandle

Parameters

xclDeviceHandle xhdl: Shim xclDeviceHandle

Return

Handle representing the opened device, or nullptr on error

The returned XRT device handle must be explicitly closed when nolonger needed.

XCL_DRIVER_DLLESPEC int xrtDeviceClose(xrtDeviceHandle dhdl)¶: Close an opened device

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen

Return

0 on success, error otherwise

XCL_DRIVER_DLLESPEC int xrtDeviceLoadXclbin(xrtDeviceHandle dhdl, const struct axlf * xclbin)¶: Load an xclbin image

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen
const struct axlf * xclbin: Pointer to complete axlf in memory image

Return

0 on success, error otherwise

The xclbin image can safely be deleted after calling this funciton.

XCL_DRIVER_DLLESPEC int xrtDeviceLoadXclbinFile(xrtDeviceHandle dhdl, const char * xclbin_fnm)¶: Read and load an xclbin file

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen
const char * xclbin_fnm: Full path to xclbin file

Return

0 on success, error otherwise

This function read the file from disk and loads the xclbin. Using this function allows one time allocation of data that needs to be kept in memory.

XCL_DRIVER_DLLESPEC int xrtDeviceLoadXclbinHandle(xrtDeviceHandle dhdl, xrtXclbinHandle xhdl)¶: load an xclbin from an xrt::xclbin handle

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen
xrtXclbinHandle xhdl: xrt::xclbin handle

Return

0 on success, error otherwise

This function uses the specified xrt::xclbin object created by caller. The xrt::xclbin object must contain the complete axlf structure.

XCL_DRIVER_DLLESPEC int xrtDeviceLoadXclbinUUID(xrtDeviceHandle dhdl, const xuid_t uuid)¶: load an xclbin from an xrt::xclbin handle

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen
const xuid_t uuid: uuid_t struct of xclbin id

Return

0 on success, error otherwise

This function reads the xclbin id already loaded in the system and comapres it with the input uuid. If they match, load the cached xclbin metadata into caller’s process. Otherwise returns error.

XCL_DRIVER_DLLESPEC int xrtDeviceGetXclbinUUID(xrtDeviceHandle dhdl, xuid_t out)¶: Get UUID of xclbin image loaded on device

Parameters

xrtDeviceHandle dhdl: Handle to device previously opened with xrtDeviceOpen
xuid_t out: Return xclbin id in this uuid_t struct

Return

0 on success or appropriate error number

Note that current UUID can be different from the UUID of the xclbin loaded by this process using load_xclbin()

typedef xrtXclbinHandle¶: opaque xclbin handle

XCL_DRIVER_DLLESPEC xrtXclbinHandle xrtXclbinAllocFilename(const char * filename)¶: Allocate a xclbin using xclbin filename

Parameters

const char * filename: path to the xclbin file

Return

xrtXclbinHandle on success or NULL with errno set

XCL_DRIVER_DLLESPEC xrtXclbinHandle xrtXclbinAllocAxlf(const struct axlf * top_axlf)¶: Allocate a xclbin using an axlf

Parameters

const struct axlf * top_axlf: an axlf

Return

xrtXclbinHandle on success or NULL with errno set

XCL_DRIVER_DLLESPEC xrtXclbinHandle xrtXclbinAllocRawData(const char * data, int size)¶: Allocate a xclbin using raw data

Parameters

const char * data: raw data buffer of xclbin
int size: size (in bytes) of raw data buffer of xclbin

Return

xrtXclbinHandle on success or NULL with errno set

XCL_DRIVER_DLLESPEC int xrtXclbinFreeHandle(xrtXclbinHandle xhdl)¶: Deallocate the xclbin handle

Parameters

xrtXclbinHandle xhdl: xclbin handle

Return

0 on success, -1 on error

XCL_DRIVER_DLLESPEC int xrtXclbinGetXSAName(xrtXclbinHandle xhdl, char * name, int size, int * ret_size)¶: Get Xilinx Support Archive (XSA) Name of xclbin handle

Parameters

xrtXclbinHandle xhdl: Xclbin handle
char * name: Return name of XSA. If the value is nullptr, the content of this value will not be populated. Otherwise, the the content of this value will be populated.
int size: size (in bytes) of name.
int * ret_size: Return size (in bytes) of XSA name. If the value is nullptr, the content of this value will not be populated. Otherwise, the the content of this value will be populated.

Return

0 on success or appropriate error number

XCL_DRIVER_DLLESPEC int xrtXclbinGetUUID(xrtXclbinHandle xhdl, xuid_t ret_uuid)¶: Get UUID of xclbin handle

Parameters

xrtXclbinHandle xhdl: Xclbin handle
xuid_t ret_uuid: Return xclbin id in this uuid_t struct

Return

0 on success or appropriate error number

XCL_DRIVER_DLLESPEC size_t xrtXclbinGetNumKernels(xrtXclbinHandle xhdl)¶: Get number of PL kernels in xclbin

Parameters

xrtXclbinHandle xhdl: Xclbin handle obtained from an xrtXclbinAlloc function

Return

The number of PL kernels in the xclbin

Kernels are extracted from embedded XML metadata in the xclbin. A kernel groups one or more compute units. A kernel has arguments from which offset, type, etc can be retrived.

XCL_DRIVER_DLLESPEC size_t xrtXclbinGetNumKernelComputeUnits(xrtXclbinHandle xhdl)¶: Get number of CUs in xclbin

Parameters

xrtXclbinHandle xhdl: Xclbin handle obtained from an xrtXclbinAlloc function

Return

The number of compute units

Compute units are associated with kernels. This function returns the total number of compute units as the sum of compute units over all kernels.

XCL_DRIVER_DLLESPEC int xrtXclbinGetData(xrtXclbinHandle xhdl, char * data, int size, int * ret_size)¶: Get the raw data of the xclbin handle

Parameters

xrtXclbinHandle xhdl: Xclbin handle
char * data: Return raw data. If the value is nullptr, the content of this value will not be populated. Otherwise, the the content of this value will be populated.
int size: Size (in bytes) of data
int * ret_size: Return size (in bytes) of XSA name. If the value is nullptr, the content of this value will not be populated. Otherwise, the the content of this value will be populated.

Return

0 on success or appropriate error number

Kernel APIs¶

typedef xrtKernelHandle¶: opaque kernel handle

Description

A kernel handle is obtained by opening a kernel. Clients pass this kernel handle to APIs that operate on a kernel.

typedef xrtRunHandle¶: opaque handle to a specific kernel run

Description

A run handle is obtained by running a kernel. Clients use a run handle to check or wait for kernel completion.

XCL_DRIVER_DLLESPEC xrtKernelHandle xrtPLKernelOpen(xrtDeviceHandle deviceHandle, const xuid_t xclbinId, const char * name)¶: Open a PL kernel and obtain its handle.

Parameters

xrtDeviceHandle deviceHandle: Handle to the device with the kernel
const xuid_t xclbinId: The uuid of the xclbin with the specified kernel.
const char * name: Name of kernel to open.

Return

Handle representing the opened kernel.

The kernel name must uniquely identify compatible kernel instances (compute units). Optionally specify which kernel instance(s) to open using “kernelname:{instancename1,instancename2,…}” syntax. The compute units are opened with shared access, meaning that other kernels and other process will have shared access to same compute units. If exclusive access is needed then open the kernel using xrtPLKernelOpenExclusve().

An xclbin with the specified kernel must have been loaded prior to calling this function. An XRT_NULL_HANDLE is returned on error and errno is set accordingly.

A kernel handle is thread safe and can be shared between threads.

XCL_DRIVER_DLLESPEC xrtKernelHandle xrtPLKernelOpenExclusive(xrtDeviceHandle deviceHandle, const xuid_t xclbinId, const char * name)¶: Open a PL kernel and obtain its handle.

Parameters

xrtDeviceHandle deviceHandle: Handle to the device with the kernel
const xuid_t xclbinId: The uuid of the xclbin with the specified kernel.
const char * name: Name of kernel to open.

Return

Handle representing the opened kernel.

Same as xrtPLKernelOpen(), but opens compute units with exclusive access. Fails if any compute unit is already opened with either exclusive or shared access.

XCL_DRIVER_DLLESPEC int xrtKernelClose(xrtKernelHandle kernelHandle)¶: Close an opened kernel

Parameters

xrtKernelHandle kernelHandle: Handle to kernel previously opened with xrtKernelOpen

Return

0 on success, -1 on error

XCL_DRIVER_DLLESPEC int xrtKernelArgGroupId(xrtKernelHandle kernelHandle, int argno)¶: Acquire bank group id for kernel argument

Parameters

xrtKernelHandle kernelHandle: Handle to kernel previously opened with xrtKernelOpen
int argno: Index of kernel argument

Return

Group id or negative error code on error

A valid group id is a non-negative integer. The group id is required when constructing a buffer object.

The kernel argument group id is ambigious if kernel has multiple kernel with different connectivity for specified argument. In this case the API returns error.

XCL_DRIVER_DLLESPEC uint32_t xrtKernelArgOffset(xrtKernelHandle khdl, int argno)¶: Get the offset of kernel argument

Parameters

xrtKernelHandle khdl: Handle to kernel previously opened with xrtKernelOpen
int argno: Index of kernel argument

Return

The kernel register offset of the argument with specified index

Use with :c:func:`xrtKernelReadRegister()` and :c:func:`xrtKernelWriteRegister()` if manually reading or writing kernel registers for explicit arguments.

XCL_DRIVER_DLLESPEC int xrtKernelReadRegister(xrtKernelHandle kernelHandle, uint32_t offset, uint32_t * datap)¶: Read data from kernel address range

Parameters

xrtKernelHandle kernelHandle: Handle to kernel previously opened with xrtKernelOpen
uint32_t offset: Offset in register space to read from
uint32_t * datap: Pointer to location where to write data

Return

0 on success, errcode otherwise

The kernel must be associated with exactly one kernel instance (compute unit), which must be opened for exclusive access.

XCL_DRIVER_DLLESPEC int xrtKernelWriteRegister(xrtKernelHandle kernelHandle, uint32_t offset, uint32_t data)¶: Write to the address range of a kernel

Parameters

xrtKernelHandle kernelHandle: Handle to kernel previously opened with xrtKernelOpen
uint32_t offset: Offset in register space to write to
uint32_t data: Data to write

Return

0 on success, errcode otherwise

The kernel must be associated with exactly one kernel instance (compute unit), which must be opened for exclusive access.

XCL_DRIVER_DLLESPEC xrtRunHandle xrtKernelRun(xrtKernelHandle kernelHandle, ...)¶: Start a kernel execution

Parameters

xrtKernelHandle kernelHandle: Handle to the kernel to run
...: Kernel arguments

Return

Run handle which must be closed with xrtRunClose()

A run handle is specific to one execution of a kernel. Once execution completes, the run handle can be re-used to execute the same kernel again. When no longer needed, then run handle must be closed with xrtRunClose().

XCL_DRIVER_DLLESPEC xrtRunHandle xrtRunOpen(xrtKernelHandle kernelHandle)¶: Open a new run handle for a kernel without starting kernel

Parameters

xrtKernelHandle kernelHandle: Handle to the kernel to associate the run handle with

Return

Run handle which must be closed with xrtRunClose()

The handle can be used repeatedly to start an execution of the associated kernel. This API allows application to manage run handles without maintaining corresponding kernel handle.

XCL_DRIVER_DLLESPEC int xrtRunSetArg(xrtRunHandle rhdl, int index, ...)¶: Set a specific kernel argument for this run

Parameters

xrtRunHandle rhdl: Handle to the run object to modify
int index: Index of kernel argument to set
...: The argument value to set.

Return

0 on success, -1 on error

Use this API to explicitly set specific kernel arguments prior to starting kernel execution. After setting all arguments, the kernel execution can be start with xrtRunStart()

XCL_DRIVER_DLLESPEC int xrtRunUpdateArg(xrtRunHandle rhdl, int index, ...)¶: Asynchronous update of kernel argument

Parameters

xrtRunHandle rhdl: Handle to the run object to modify
int index: Index of kernel argument to update
...: The argument value to update.

Return

0 on success, -1 on error

Use this API to asynchronously update a specific kernel argument of an existing run.

This API is only supported on Edge.

XCL_DRIVER_DLLESPEC int xrtRunStart(xrtRunHandle rhdl)¶: Start existing run handle

Parameters

xrtRunHandle rhdl: Handle to the run object to start

Return

0 on success, -1 on error

Use this API when re-using a run handle for more than one execution of the kernel associated with the run handle.

XCL_DRIVER_DLLESPEC enum ert_cmd_state xrtRunWait(xrtRunHandle rhdl)¶: Wait for a run to complete

Parameters

xrtRunHandle rhdl: Handle to the run object to start

Return

Run command state for completed run,: or ERT_CMD_STATE_ABORT on error

Blocks current thread until job has completed

XCL_DRIVER_DLLESPEC enum ert_cmd_state xrtRunWaitFor(xrtRunHandle rhdl, unsigned int timeout_ms)¶: Wait for a run to complete

Parameters

xrtRunHandle rhdl: Handle to the run object to start
unsigned int timeout_ms: Timeout in millisecond

Return

Run command state for completed run, or: current status if timeout.

Blocks current thread until job has completed

XCL_DRIVER_DLLESPEC enum ert_cmd_state xrtRunState(xrtRunHandle rhdl)¶: Check the current state of a run

Parameters

xrtRunHandle rhdl: Handle to check

Return

The underlying command execution state per ert.h

XCL_DRIVER_DLLESPEC int xrtRunSetCallback(xrtRunHandle rhdl, enum ert_cmd_state state, void (*callback) (xrtRunHandle, enum ert_cmd_state, void*, void * data)¶: Set a callback function

Parameters

xrtRunHandle rhdl: Handle to set callback on
enum ert_cmd_state state: State to invoke callback on
void (*)(xrtRunHandle, enum ert_cmd_state, void*) callback: Callback function
void * data: User data to pass to callback function

Description

Register a run callback function that is invoked when the run changes underlying execution state to specified state. Support states are: ERT_CMD_STATE_COMPLETED (to be extended)

XCL_DRIVER_DLLESPEC int xrtRunClose(xrtRunHandle rhdl)¶: Close a run handle

Parameters

xrtRunHandle rhdl: Handle to close

Return

0 on success, -1 on error