HIP API

HIP API#

HIP Runtime API Reference: HIP API

Collaboration diagram for HIP API:

Modules
	Initialization and Version

	Device Management

	Execution Control

	Error Handling

	Stream Management

	Event Management

	Memory Management

	PeerToPeer Device Memory Access

	Context Management

	Module Management

	Occupancy

	Profiler Control[Deprecated]

	Launch API to support the triple-chevron syntax

	Texture Management

Variables
unsigned	hipDeviceArch_t::hasGlobalInt32Atomics: 1
	32-bit integer atomics for global memory. More...

unsigned	hipDeviceArch_t::hasGlobalFloatAtomicExch: 1
	32-bit float atomic exch for global memory. More...

unsigned	hipDeviceArch_t::hasSharedInt32Atomics: 1
	32-bit integer atomics for shared memory. More...

unsigned	hipDeviceArch_t::hasSharedFloatAtomicExch: 1
	32-bit float atomic exch for shared memory. More...

unsigned	hipDeviceArch_t::hasFloatAtomicAdd: 1
	32-bit float atomic add in global and shared memory. More...

unsigned	hipDeviceArch_t::hasGlobalInt64Atomics: 1
	64-bit integer atomics for global memory. More...

unsigned	hipDeviceArch_t::hasSharedInt64Atomics: 1
	64-bit integer atomics for shared memory. More...

unsigned	hipDeviceArch_t::hasDoubles: 1
	Double-precision floating point. More...

unsigned	hipDeviceArch_t::hasWarpVote: 1
	Warp vote instructions (__any, __all). More...

unsigned	hipDeviceArch_t::hasWarpBallot: 1
	Warp ballot instructions (__ballot). More...

unsigned	hipDeviceArch_t::hasWarpShuffle: 1
	Warp shuffle operations. (__shfl_*). More...

unsigned	hipDeviceArch_t::hasFunnelShift: 1
	Funnel two words into one with shift&mask caps. More...

unsigned	hipDeviceArch_t::hasThreadFenceSystem: 1
	__threadfence_system. More...

unsigned	hipDeviceArch_t::hasSyncThreadsExt: 1
	__syncthreads_count, syncthreads_and, syncthreads_or. More...

unsigned	hipDeviceArch_t::hasSurfaceFuncs: 1
	Surface functions. More...

unsigned	hipDeviceArch_t::has3dGrid: 1
	Grid and group dims are 3D (rather than 2D). More...

unsigned	hipDeviceArch_t::hasDynamicParallelism: 1
	Dynamic parallelism. More...

char	hipDeviceProp_t::name [256]
	Device name. More...

size_t	hipDeviceProp_t::totalGlobalMem
	Size of global memory region (in bytes). More...

size_t	hipDeviceProp_t::sharedMemPerBlock
	Size of shared memory region (in bytes). More...

int	hipDeviceProp_t::regsPerBlock
	Registers per block. More...

int	hipDeviceProp_t::warpSize
	Warp size. More...

int	hipDeviceProp_t::maxThreadsPerBlock
	Max work items per work group or workgroup max size. More...

int	hipDeviceProp_t::maxThreadsDim [3]
	Max number of threads in each dimension (XYZ) of a block. More...

int	hipDeviceProp_t::maxGridSize [3]
	Max grid dimensions (XYZ). More...

int	hipDeviceProp_t::clockRate
	Max clock frequency of the multiProcessors in khz. More...

int	hipDeviceProp_t::memoryClockRate
	Max global memory clock frequency in khz. More...

int	hipDeviceProp_t::memoryBusWidth
	Global memory bus width in bits. More...

size_t	hipDeviceProp_t::totalConstMem
	Size of shared memory region (in bytes). More...

int	hipDeviceProp_t::major

int	hipDeviceProp_t::minor

int	hipDeviceProp_t::multiProcessorCount
	Number of multi-processors (compute units). More...

int	hipDeviceProp_t::l2CacheSize
	L2 cache size. More...

int	hipDeviceProp_t::maxThreadsPerMultiProcessor
	Maximum resident threads per multi-processor. More...

int	hipDeviceProp_t::computeMode
	Compute mode. More...

int	hipDeviceProp_t::clockInstructionRate

hipDeviceArch_t	hipDeviceProp_t::arch
	Architectural feature flags. New for HIP. More...

int	hipDeviceProp_t::concurrentKernels
	Device can possibly execute multiple kernels concurrently. More...

int	hipDeviceProp_t::pciDomainID
	PCI Domain ID. More...

int	hipDeviceProp_t::pciBusID
	PCI Bus ID. More...

int	hipDeviceProp_t::pciDeviceID
	PCI Device ID. More...

size_t	hipDeviceProp_t::maxSharedMemoryPerMultiProcessor
	Maximum Shared Memory Per Multiprocessor. More...

int	hipDeviceProp_t::isMultiGpuBoard
	1 if device is on a multi-GPU board, 0 if not. More...

int	hipDeviceProp_t::canMapHostMemory
	Check whether HIP can map host memory. More...

int	hipDeviceProp_t::gcnArch
	DEPRECATED: use gcnArchName instead. More...

char	hipDeviceProp_t::gcnArchName [256]
	AMD GCN Arch Name. More...

int	hipDeviceProp_t::integrated
	APU vs dGPU. More...

int	hipDeviceProp_t::cooperativeLaunch
	HIP device supports cooperative launch. More...

int	hipDeviceProp_t::cooperativeMultiDeviceLaunch
	HIP device supports cooperative launch on multiple devices. More...

int	hipDeviceProp_t::maxTexture1DLinear
	Maximum size for 1D textures bound to linear memory. More...

int	hipDeviceProp_t::maxTexture1D
	Maximum number of elements in 1D images. More...

int	hipDeviceProp_t::maxTexture2D [2]
	Maximum dimensions (width, height) of 2D images, in image elements. More...

int	hipDeviceProp_t::maxTexture3D [3]
	Maximum dimensions (width, height, depth) of 3D images, in image elements. More...

unsigned int *	hipDeviceProp_t::hdpMemFlushCntl
	Addres of HDP_MEM_COHERENCY_FLUSH_CNTL register. More...

unsigned int *	hipDeviceProp_t::hdpRegFlushCntl
	Addres of HDP_REG_COHERENCY_FLUSH_CNTL register. More...

size_t	hipDeviceProp_t::memPitch
	Maximum pitch in bytes allowed by memory copies. More...

size_t	hipDeviceProp_t::textureAlignment
	Alignment requirement for textures. More...

size_t	hipDeviceProp_t::texturePitchAlignment
	Pitch alignment requirement for texture references bound to pitched memory. More...

int	hipDeviceProp_t::kernelExecTimeoutEnabled
	Run time limit for kernels executed on the device. More...

int	hipDeviceProp_t::ECCEnabled
	Device has ECC support enabled. More...

int	hipDeviceProp_t::tccDriver
	1:If device is Tesla device using TCC driver, else 0 More...

int	hipDeviceProp_t::cooperativeMultiDeviceUnmatchedFunc

int	hipDeviceProp_t::cooperativeMultiDeviceUnmatchedGridDim

int	hipDeviceProp_t::cooperativeMultiDeviceUnmatchedBlockDim

int	hipDeviceProp_t::cooperativeMultiDeviceUnmatchedSharedMem

int	hipDeviceProp_t::isLargeBar
	1: if it is a large PCI bar device, else 0 More...

int	hipDeviceProp_t::asicRevision
	Revision of the GPU in this device. More...

int	hipDeviceProp_t::managedMemory
	Device supports allocating managed memory on this system. More...

int	hipDeviceProp_t::directManagedMemAccessFromHost
	Host can directly access managed memory on the device without migration. More...

int	hipDeviceProp_t::concurrentManagedAccess
	Device can coherently access managed memory concurrently with the CPU. More...

int	hipDeviceProp_t::pageableMemoryAccess

int	hipDeviceProp_t::pageableMemoryAccessUsesHostPageTables
	Device accesses pageable memory via the host's page tables. More...

enum hipMemoryType	hipPointerAttribute_t::memoryType

int	hipPointerAttribute_t::device

void *	hipPointerAttribute_t::devicePointer

void *	hipPointerAttribute_t::hostPointer

int	hipPointerAttribute_t::isManaged

unsigned	hipPointerAttribute_t::allocationFlags

	hipSuccess = 0
	Successful completion. More...

	hipErrorInvalidValue = 1

	hipErrorOutOfMemory = 2

	hipErrorMemoryAllocation = 2
	Memory allocation error. More...

	hipErrorNotInitialized = 3

	hipErrorInitializationError = 3

	hipErrorDeinitialized = 4

	hipErrorProfilerDisabled = 5

	hipErrorProfilerNotInitialized = 6

	hipErrorProfilerAlreadyStarted = 7

	hipErrorProfilerAlreadyStopped = 8

	hipErrorInvalidConfiguration = 9

	hipErrorInvalidPitchValue = 12

	hipErrorInvalidSymbol = 13

	hipErrorInvalidDevicePointer = 17
	Invalid Device Pointer. More...

	hipErrorInvalidMemcpyDirection = 21
	Invalid memory copy direction. More...

	hipErrorInsufficientDriver = 35

	hipErrorMissingConfiguration = 52

	hipErrorPriorLaunchFailure = 53

	hipErrorInvalidDeviceFunction = 98

	hipErrorNoDevice = 100
	Call to hipGetDeviceCount returned 0 devices. More...

	hipErrorInvalidDevice = 101
	DeviceID must be in range 0...#compute-devices. More...

	hipErrorInvalidImage = 200

	hipErrorInvalidContext = 201
	Produced when input context is invalid. More...

	hipErrorContextAlreadyCurrent = 202

	hipErrorMapFailed = 205

	hipErrorMapBufferObjectFailed = 205
	Produced when the IPC memory attach failed from ROCr. More...

	hipErrorUnmapFailed = 206

	hipErrorArrayIsMapped = 207

	hipErrorAlreadyMapped = 208

	hipErrorNoBinaryForGpu = 209

	hipErrorAlreadyAcquired = 210

	hipErrorNotMapped = 211

	hipErrorNotMappedAsArray = 212

	hipErrorNotMappedAsPointer = 213

	hipErrorECCNotCorrectable = 214

	hipErrorUnsupportedLimit = 215

	hipErrorContextAlreadyInUse = 216

	hipErrorPeerAccessUnsupported = 217

	hipErrorInvalidKernelFile = 218
	In CUDA DRV, it is CUDA_ERROR_INVALID_PTX. More...

	hipErrorInvalidGraphicsContext = 219

	hipErrorInvalidSource = 300

	hipErrorFileNotFound = 301

	hipErrorSharedObjectSymbolNotFound = 302

	hipErrorSharedObjectInitFailed = 303

	hipErrorOperatingSystem = 304

	hipErrorInvalidHandle = 400

	hipErrorInvalidResourceHandle = 400
	Resource handle (hipEvent_t or hipStream_t) invalid. More...

	hipErrorIllegalState = 401
	Resource required is not in a valid state to perform operation. More...

	hipErrorNotFound = 500

	hipErrorNotReady = 600

	hipErrorIllegalAddress = 700

	hipErrorLaunchOutOfResources = 701
	Out of resources error. More...

	hipErrorLaunchTimeOut = 702

	hipErrorPeerAccessAlreadyEnabled
	Peer access was already enabled from the current device. More...

	hipErrorPeerAccessNotEnabled
	Peer access was never enabled from the current device. More...

	hipErrorSetOnActiveProcess = 708

	hipErrorContextIsDestroyed = 709

	hipErrorAssert = 710
	Produced when the kernel calls assert. More...

	hipErrorHostMemoryAlreadyRegistered
	Produced when trying to lock a page-locked memory. More...

	hipErrorHostMemoryNotRegistered
	Produced when trying to unlock a non-page-locked memory. More...

	hipErrorLaunchFailure
	An exception occurred on the device while executing a kernel. More...

	hipErrorCooperativeLaunchTooLarge

	hipErrorNotSupported = 801
	Produced when the hip API is not supported/implemented. More...

	hipErrorStreamCaptureUnsupported = 900

	hipErrorStreamCaptureInvalidated = 901

	hipErrorStreamCaptureMerge = 902

	hipErrorStreamCaptureUnmatched = 903
	The capture was not initiated in this stream. More...

	hipErrorStreamCaptureUnjoined = 904

	hipErrorStreamCaptureIsolation = 905

	hipErrorStreamCaptureImplicit = 906

	hipErrorCapturedEvent = 907

	hipErrorStreamCaptureWrongThread = 908

	hipErrorGraphExecUpdateFailure = 910

	hipErrorUnknown = 999

	hipErrorRuntimeMemory = 1052

	hipErrorRuntimeOther = 1053

char	hipIpcMemHandle_t::reserved [HIP_IPC_HANDLE_SIZE]

char	hipIpcEventHandle_t::reserved [HIP_IPC_HANDLE_SIZE]

int	hipFuncAttributes::binaryVersion

int	hipFuncAttributes::cacheModeCA

size_t	hipFuncAttributes::constSizeBytes

size_t	hipFuncAttributes::localSizeBytes

int	hipFuncAttributes::maxDynamicSharedSizeBytes

int	hipFuncAttributes::maxThreadsPerBlock

int	hipFuncAttributes::numRegs

int	hipFuncAttributes::preferredShmemCarveout

int	hipFuncAttributes::ptxVersion

size_t	hipFuncAttributes::sharedSizeBytes

uint32_t	dim3::x
	x More...

uint32_t	dim3::y
	y More...

uint32_t	dim3::z
	z More...

void *	hipLaunchParams::func
	Device function symbol. More...

dim3	hipLaunchParams::gridDim
	Grid dimentions. More...

dim3	hipLaunchParams::blockDim
	Block dimentions. More...

void **	hipLaunchParams::args
	Arguments. More...

size_t	hipLaunchParams::sharedMem
	Shared memory. More...

hipStream_t	hipLaunchParams::stream
	Stream identifier. More...

hipExternalMemoryHandleType	hipExternalMemoryHandleDesc::type

int hipExternalMemoryHandleDesc::fd

void * hipExternalMemoryHandleDesc::handle

const void * hipExternalMemoryHandleDesc::name

struct {
void * hipExternalMemoryHandleDesc::handle

const void * hipExternalMemoryHandleDesc::name

} hipExternalMemoryHandleDesc::win32

union {
int hipExternalMemoryHandleDesc::fd

struct {
void * hipExternalMemoryHandleDesc::handle

const void * hipExternalMemoryHandleDesc::name

} hipExternalMemoryHandleDesc::win32

}	hipExternalMemoryHandleDesc::handle

unsigned long long	hipExternalMemoryHandleDesc::size

unsigned int	hipExternalMemoryHandleDesc::flags

unsigned long long	hipExternalMemoryBufferDesc::offset

unsigned long long	hipExternalMemoryBufferDesc::size

unsigned int	hipExternalMemoryBufferDesc::flags

hipExternalSemaphoreHandleType	hipExternalSemaphoreHandleDesc::type

int hipExternalSemaphoreHandleDesc::fd

void * hipExternalSemaphoreHandleDesc::handle

const void * hipExternalSemaphoreHandleDesc::name

struct {
void * hipExternalSemaphoreHandleDesc::handle

const void * hipExternalSemaphoreHandleDesc::name

} hipExternalSemaphoreHandleDesc::win32

union {
int hipExternalSemaphoreHandleDesc::fd

struct {
void * hipExternalSemaphoreHandleDesc::handle

const void * hipExternalSemaphoreHandleDesc::name

} hipExternalSemaphoreHandleDesc::win32

}	hipExternalSemaphoreHandleDesc::handle

unsigned int	hipExternalSemaphoreHandleDesc::flags

unsigned long long hipExternalSemaphoreSignalParams::value

struct {
unsigned long long hipExternalSemaphoreSignalParams::value

} hipExternalSemaphoreSignalParams::fence

unsigned long long hipExternalSemaphoreSignalParams::key

struct {
unsigned long long hipExternalSemaphoreSignalParams::key

} hipExternalSemaphoreSignalParams::keyedMutex

unsigned int hipExternalSemaphoreSignalParams::reserved [12]

struct {
struct {
unsigned long long hipExternalSemaphoreSignalParams::value

} hipExternalSemaphoreSignalParams::fence

struct {
unsigned long long hipExternalSemaphoreSignalParams::key

} hipExternalSemaphoreSignalParams::keyedMutex

unsigned int hipExternalSemaphoreSignalParams::reserved [12]

}	hipExternalSemaphoreSignalParams::params

unsigned int	hipExternalSemaphoreSignalParams::flags

unsigned long long hipExternalSemaphoreWaitParams::value

struct {
unsigned long long hipExternalSemaphoreWaitParams::value

} hipExternalSemaphoreWaitParams::fence

unsigned long long hipExternalSemaphoreWaitParams::key

unsigned int hipExternalSemaphoreWaitParams::timeoutMs

struct {
unsigned long long hipExternalSemaphoreWaitParams::key

unsigned int hipExternalSemaphoreWaitParams::timeoutMs

} hipExternalSemaphoreWaitParams::keyedMutex

unsigned int hipExternalSemaphoreWaitParams::reserved [10]

struct {
struct {
unsigned long long hipExternalSemaphoreWaitParams::value

} hipExternalSemaphoreWaitParams::fence

struct {
unsigned long long hipExternalSemaphoreWaitParams::key

unsigned int hipExternalSemaphoreWaitParams::timeoutMs

} hipExternalSemaphoreWaitParams::keyedMutex

unsigned int hipExternalSemaphoreWaitParams::reserved [10]

}	hipExternalSemaphoreWaitParams::params

unsigned int	hipExternalSemaphoreWaitParams::flags

hipHostFn_t	hipHostNodeParams::fn

void *	hipHostNodeParams::userData

dim3	hipKernelNodeParams::blockDim

void **	hipKernelNodeParams::extra

void *	hipKernelNodeParams::func

dim3	hipKernelNodeParams::gridDim

void **	hipKernelNodeParams::kernelParams

unsigned int	hipKernelNodeParams::sharedMemBytes

void *	hipMemsetParams::dst

unsigned int	hipMemsetParams::elementSize

size_t	hipMemsetParams::height

size_t	hipMemsetParams::pitch

unsigned int	hipMemsetParams::value

size_t	hipMemsetParams::width

Detailed Description

Defines the HIP API. See the individual sections for more information.

Variable Documentation

◆ allocationFlags

unsigned hipPointerAttribute_t::allocationFlags

◆ arch

hipDeviceArch_t hipDeviceProp_t::arch

Architectural feature flags. New for HIP.

◆ args

void** hipLaunchParams::args

Arguments.

◆ asicRevision

int hipDeviceProp_t::asicRevision

Revision of the GPU in this device.

◆ binaryVersion

int hipFuncAttributes::binaryVersion

◆ blockDim [1/2]

dim3 hipLaunchParams::blockDim

Block dimentions.

◆ blockDim [2/2]

dim3 hipKernelNodeParams::blockDim

◆ cacheModeCA

int hipFuncAttributes::cacheModeCA

◆ canMapHostMemory

int hipDeviceProp_t::canMapHostMemory

Check whether HIP can map host memory.

◆ clockInstructionRate

int hipDeviceProp_t::clockInstructionRate

Frequency in khz of the timer used by the device-side "clock*" instructions. New for HIP.

◆ clockRate

int hipDeviceProp_t::clockRate

Max clock frequency of the multiProcessors in khz.

◆ computeMode

int hipDeviceProp_t::computeMode

Compute mode.

◆ concurrentKernels

int hipDeviceProp_t::concurrentKernels

Device can possibly execute multiple kernels concurrently.

◆ concurrentManagedAccess

int hipDeviceProp_t::concurrentManagedAccess

Device can coherently access managed memory concurrently with the CPU.

◆ constSizeBytes

size_t hipFuncAttributes::constSizeBytes

◆ cooperativeLaunch

int hipDeviceProp_t::cooperativeLaunch

HIP device supports cooperative launch.

◆ cooperativeMultiDeviceLaunch

int hipDeviceProp_t::cooperativeMultiDeviceLaunch

HIP device supports cooperative launch on multiple devices.

◆ cooperativeMultiDeviceUnmatchedBlockDim

int hipDeviceProp_t::cooperativeMultiDeviceUnmatchedBlockDim

HIP device supports cooperative launch on multiple devices with unmatched block dimensions

◆ cooperativeMultiDeviceUnmatchedFunc

int hipDeviceProp_t::cooperativeMultiDeviceUnmatchedFunc

HIP device supports cooperative launch on multiple devices with unmatched functions

◆ cooperativeMultiDeviceUnmatchedGridDim

int hipDeviceProp_t::cooperativeMultiDeviceUnmatchedGridDim

HIP device supports cooperative launch on multiple devices with unmatched grid dimensions

◆ cooperativeMultiDeviceUnmatchedSharedMem

int hipDeviceProp_t::cooperativeMultiDeviceUnmatchedSharedMem

HIP device supports cooperative launch on multiple devices with unmatched shared memories

◆ device

int hipPointerAttribute_t::device

◆ devicePointer

void* hipPointerAttribute_t::devicePointer

◆ directManagedMemAccessFromHost

int hipDeviceProp_t::directManagedMemAccessFromHost

Host can directly access managed memory on the device without migration.

◆ dst

void* hipMemsetParams::dst

◆ ECCEnabled

int hipDeviceProp_t::ECCEnabled

Device has ECC support enabled.

◆ elementSize

unsigned int hipMemsetParams::elementSize

◆ extra

void** hipKernelNodeParams::extra

◆ fd [1/4]

int hipExternalMemoryHandleDesc::fd

◆ [2/4]

int { ... } ::fd

◆ fd [3/4]

int hipExternalSemaphoreHandleDesc::fd

◆ [4/4]

int { ... } ::fd

◆ [1/4]

struct { ... } hipExternalSemaphoreSignalParams::fence

◆ [2/4]

struct { ... } ::fence

◆ [3/4]

struct { ... } hipExternalSemaphoreWaitParams::fence

◆ [4/4]

struct { ... } ::fence

◆ flags [1/5]

unsigned int hipExternalMemoryHandleDesc::flags

◆ flags [2/5]

unsigned int hipExternalMemoryBufferDesc::flags

◆ flags [3/5]

unsigned int hipExternalSemaphoreHandleDesc::flags

◆ flags [4/5]

unsigned int hipExternalSemaphoreSignalParams::flags

◆ flags [5/5]

unsigned int hipExternalSemaphoreWaitParams::flags

◆ fn

hipHostFn_t hipHostNodeParams::fn

◆ func [1/2]

void* hipLaunchParams::func

Device function symbol.

◆ func [2/2]

void* hipKernelNodeParams::func

◆ gcnArch

int hipDeviceProp_t::gcnArch

DEPRECATED: use gcnArchName instead.

◆ gcnArchName

char hipDeviceProp_t::gcnArchName[256]

AMD GCN Arch Name.

◆ gridDim [1/2]

dim3 hipLaunchParams::gridDim

Grid dimentions.

◆ gridDim [2/2]

dim3 hipKernelNodeParams::gridDim

◆ handle [1/6]

void* hipExternalMemoryHandleDesc::handle

◆ [2/6]

void* { ... } ::handle

◆ [3/6]

union { ... } hipExternalMemoryHandleDesc::handle

◆ handle [4/6]

void* hipExternalSemaphoreHandleDesc::handle

◆ [5/6]

void* { ... } ::handle

◆ [6/6]

union { ... } hipExternalSemaphoreHandleDesc::handle

◆ has3dGrid

unsigned hipDeviceArch_t::has3dGrid

Grid and group dims are 3D (rather than 2D).

◆ hasDoubles

unsigned hipDeviceArch_t::hasDoubles

Double-precision floating point.

◆ hasDynamicParallelism

unsigned hipDeviceArch_t::hasDynamicParallelism

Dynamic parallelism.

◆ hasFloatAtomicAdd

unsigned hipDeviceArch_t::hasFloatAtomicAdd

32-bit float atomic add in global and shared memory.

◆ hasFunnelShift

unsigned hipDeviceArch_t::hasFunnelShift

Funnel two words into one with shift&mask caps.

◆ hasGlobalFloatAtomicExch

unsigned hipDeviceArch_t::hasGlobalFloatAtomicExch

32-bit float atomic exch for global memory.

◆ hasGlobalInt32Atomics

unsigned hipDeviceArch_t::hasGlobalInt32Atomics

32-bit integer atomics for global memory.

◆ hasGlobalInt64Atomics

unsigned hipDeviceArch_t::hasGlobalInt64Atomics

64-bit integer atomics for global memory.

◆ hasSharedFloatAtomicExch

unsigned hipDeviceArch_t::hasSharedFloatAtomicExch

32-bit float atomic exch for shared memory.

◆ hasSharedInt32Atomics

unsigned hipDeviceArch_t::hasSharedInt32Atomics

32-bit integer atomics for shared memory.

◆ hasSharedInt64Atomics

unsigned hipDeviceArch_t::hasSharedInt64Atomics

64-bit integer atomics for shared memory.

◆ hasSurfaceFuncs

unsigned hipDeviceArch_t::hasSurfaceFuncs

Surface functions.

◆ hasSyncThreadsExt

unsigned hipDeviceArch_t::hasSyncThreadsExt

__syncthreads_count, syncthreads_and, syncthreads_or.

◆ hasThreadFenceSystem

unsigned hipDeviceArch_t::hasThreadFenceSystem

__threadfence_system.

◆ hasWarpBallot

unsigned hipDeviceArch_t::hasWarpBallot

Warp ballot instructions (__ballot).

◆ hasWarpShuffle

unsigned hipDeviceArch_t::hasWarpShuffle

Warp shuffle operations. (__shfl_*).

◆ hasWarpVote

unsigned hipDeviceArch_t::hasWarpVote

Warp vote instructions (__any, __all).

◆ hdpMemFlushCntl

unsigned int* hipDeviceProp_t::hdpMemFlushCntl

Addres of HDP_MEM_COHERENCY_FLUSH_CNTL register.

◆ hdpRegFlushCntl

unsigned int* hipDeviceProp_t::hdpRegFlushCntl

Addres of HDP_REG_COHERENCY_FLUSH_CNTL register.

◆ height

size_t hipMemsetParams::height

◆ hipErrorAlreadyAcquired

hipErrorAlreadyAcquired = 210

◆ hipErrorAlreadyMapped

hipErrorAlreadyMapped = 208

◆ hipErrorArrayIsMapped

hipErrorArrayIsMapped = 207

◆ hipErrorAssert

hipErrorAssert = 710

Produced when the kernel calls assert.

◆ hipErrorCapturedEvent

hipErrorCapturedEvent = 907

The operation is not permitted on an event which was last recorded in a capturing stream.

◆ hipErrorContextAlreadyCurrent

hipErrorContextAlreadyCurrent = 202

◆ hipErrorContextAlreadyInUse

hipErrorContextAlreadyInUse = 216

◆ hipErrorContextIsDestroyed

hipErrorContextIsDestroyed = 709

◆ hipErrorCooperativeLaunchTooLarge

hipErrorCooperativeLaunchTooLarge

Initial value:

=

720

This error indicates that the number of blocks launched per grid for a kernel that was launched via cooperative launch APIs exceeds the maximum number of allowed blocks for the current device

◆ hipErrorDeinitialized

hipErrorDeinitialized = 4

◆ hipErrorECCNotCorrectable

hipErrorECCNotCorrectable = 214

◆ hipErrorFileNotFound

hipErrorFileNotFound = 301

◆ hipErrorGraphExecUpdateFailure

hipErrorGraphExecUpdateFailure = 910

This error indicates that the graph update not performed because it included changes which violated constraintsspecific to instantiated graph update.

◆ hipErrorHostMemoryAlreadyRegistered

hipErrorHostMemoryAlreadyRegistered

Initial value:

=

712

Produced when trying to lock a page-locked memory.

◆ hipErrorHostMemoryNotRegistered

hipErrorHostMemoryNotRegistered

Initial value:

=

713

Produced when trying to unlock a non-page-locked memory.

◆ hipErrorIllegalAddress

hipErrorIllegalAddress = 700

◆ hipErrorIllegalState

hipErrorIllegalState = 401

Resource required is not in a valid state to perform operation.

◆ hipErrorInitializationError

hipErrorInitializationError = 3

◆ hipErrorInsufficientDriver

hipErrorInsufficientDriver = 35

◆ hipErrorInvalidConfiguration

hipErrorInvalidConfiguration = 9

◆ hipErrorInvalidContext

hipErrorInvalidContext = 201

Produced when input context is invalid.

◆ hipErrorInvalidDevice

hipErrorInvalidDevice = 101

DeviceID must be in range 0...#compute-devices.

◆ hipErrorInvalidDeviceFunction

hipErrorInvalidDeviceFunction = 98

◆ hipErrorInvalidDevicePointer

hipErrorInvalidDevicePointer = 17

Invalid Device Pointer.

◆ hipErrorInvalidGraphicsContext

hipErrorInvalidGraphicsContext = 219

◆ hipErrorInvalidHandle

hipErrorInvalidHandle = 400

◆ hipErrorInvalidImage

hipErrorInvalidImage = 200

◆ hipErrorInvalidKernelFile

hipErrorInvalidKernelFile = 218

In CUDA DRV, it is CUDA_ERROR_INVALID_PTX.

◆ hipErrorInvalidMemcpyDirection

hipErrorInvalidMemcpyDirection = 21

Invalid memory copy direction.

◆ hipErrorInvalidPitchValue

hipErrorInvalidPitchValue = 12

◆ hipErrorInvalidResourceHandle

hipErrorInvalidResourceHandle = 400

Resource handle (hipEvent_t or hipStream_t) invalid.

◆ hipErrorInvalidSource

hipErrorInvalidSource = 300

◆ hipErrorInvalidSymbol

hipErrorInvalidSymbol = 13

◆ hipErrorInvalidValue

hipErrorInvalidValue = 1

One or more of the parameters passed to the API call is NULL or not in an acceptable range.

◆ hipErrorLaunchFailure

hipErrorLaunchFailure

Initial value:

=

719

An exception occurred on the device while executing a kernel.

◆ hipErrorLaunchOutOfResources

hipErrorLaunchOutOfResources = 701

Out of resources error.

◆ hipErrorLaunchTimeOut

hipErrorLaunchTimeOut = 702

◆ hipErrorMapBufferObjectFailed

hipErrorMapBufferObjectFailed = 205

Produced when the IPC memory attach failed from ROCr.

◆ hipErrorMapFailed

hipErrorMapFailed = 205

◆ hipErrorMemoryAllocation

hipErrorMemoryAllocation = 2

Memory allocation error.

◆ hipErrorMissingConfiguration

hipErrorMissingConfiguration = 52

◆ hipErrorNoBinaryForGpu

hipErrorNoBinaryForGpu = 209

◆ hipErrorNoDevice

hipErrorNoDevice = 100

Call to hipGetDeviceCount returned 0 devices.

◆ hipErrorNotFound

hipErrorNotFound = 500

◆ hipErrorNotInitialized

hipErrorNotInitialized = 3

◆ hipErrorNotMapped

hipErrorNotMapped = 211

◆ hipErrorNotMappedAsArray

hipErrorNotMappedAsArray = 212

◆ hipErrorNotMappedAsPointer

hipErrorNotMappedAsPointer = 213

◆ hipErrorNotReady

hipErrorNotReady = 600

Indicates that asynchronous operations enqueued earlier are not ready. This is not actually an error, but is used to distinguish from hipSuccess (which indicates completion). APIs that return this error include hipEventQuery and hipStreamQuery.

◆ hipErrorNotSupported

hipErrorNotSupported = 801

Produced when the hip API is not supported/implemented.

◆ hipErrorOperatingSystem

hipErrorOperatingSystem = 304

◆ hipErrorOutOfMemory

hipErrorOutOfMemory = 2

◆ hipErrorPeerAccessAlreadyEnabled

hipErrorPeerAccessAlreadyEnabled

Initial value:

=

704

Peer access was already enabled from the current device.

◆ hipErrorPeerAccessNotEnabled

hipErrorPeerAccessNotEnabled

Initial value:

=

705

Peer access was never enabled from the current device.

◆ hipErrorPeerAccessUnsupported

hipErrorPeerAccessUnsupported = 217

◆ hipErrorPriorLaunchFailure

hipErrorPriorLaunchFailure = 53

◆ hipErrorProfilerAlreadyStarted

hipErrorProfilerAlreadyStarted = 7

◆ hipErrorProfilerAlreadyStopped

hipErrorProfilerAlreadyStopped = 8

◆ hipErrorProfilerDisabled

hipErrorProfilerDisabled = 5

◆ hipErrorProfilerNotInitialized

hipErrorProfilerNotInitialized = 6

◆ hipErrorRuntimeMemory

hipErrorRuntimeMemory = 1052

HSA runtime memory call returned error. Typically not seen in production systems.

◆ hipErrorRuntimeOther

hipErrorRuntimeOther = 1053

HSA runtime call other than memory returned error. Typically not seen in production systems.

◆ hipErrorSetOnActiveProcess

hipErrorSetOnActiveProcess = 708

◆ hipErrorSharedObjectInitFailed

hipErrorSharedObjectInitFailed = 303

◆ hipErrorSharedObjectSymbolNotFound

hipErrorSharedObjectSymbolNotFound = 302

◆ hipErrorStreamCaptureImplicit

hipErrorStreamCaptureImplicit = 906

The operation would have resulted in a disallowed implicit dependency on a current capture sequence from hipStreamLegacy.

◆ hipErrorStreamCaptureInvalidated

hipErrorStreamCaptureInvalidated = 901

The current capture sequence on the stream has been invalidated due to a previous error.

◆ hipErrorStreamCaptureIsolation

hipErrorStreamCaptureIsolation = 905

A dependency would have been created which crosses the capture sequence boundary. Only implicit in-stream ordering dependencies are allowed to cross the boundary

◆ hipErrorStreamCaptureMerge

hipErrorStreamCaptureMerge = 902

The operation would have resulted in a merge of two independent capture sequences.

◆ hipErrorStreamCaptureUnjoined

hipErrorStreamCaptureUnjoined = 904

The capture sequence contains a fork that was not joined to the primary stream.

◆ hipErrorStreamCaptureUnmatched

hipErrorStreamCaptureUnmatched = 903

The capture was not initiated in this stream.

◆ hipErrorStreamCaptureUnsupported

hipErrorStreamCaptureUnsupported = 900

The operation is not permitted when the stream is capturing.

◆ hipErrorStreamCaptureWrongThread

hipErrorStreamCaptureWrongThread = 908

A stream capture sequence not initiated with the hipStreamCaptureModeRelaxed argument to hipStreamBeginCapture was passed to hipStreamEndCapture in a different thread.

◆ hipErrorUnknown

hipErrorUnknown = 999

◆ hipErrorUnmapFailed

hipErrorUnmapFailed = 206

◆ hipErrorUnsupportedLimit

hipErrorUnsupportedLimit = 215

◆ hipSuccess

hipSuccess = 0

Successful completion.

◆ hostPointer

void* hipPointerAttribute_t::hostPointer

◆ integrated

int hipDeviceProp_t::integrated

APU vs dGPU.

◆ isLargeBar

int hipDeviceProp_t::isLargeBar

1: if it is a large PCI bar device, else 0

◆ isManaged

int hipPointerAttribute_t::isManaged

◆ isMultiGpuBoard

int hipDeviceProp_t::isMultiGpuBoard

1 if device is on a multi-GPU board, 0 if not.

◆ kernelExecTimeoutEnabled

int hipDeviceProp_t::kernelExecTimeoutEnabled

Run time limit for kernels executed on the device.

◆ kernelParams

void** hipKernelNodeParams::kernelParams

◆ key [1/4]

unsigned long long hipExternalSemaphoreSignalParams::key

◆ [2/4]

unsigned { ... } ::key

◆ [3/4]

unsigned { ... } ::key

◆ key [4/4]

unsigned long long hipExternalSemaphoreWaitParams::key

◆ [1/4]

struct { ... } hipExternalSemaphoreSignalParams::keyedMutex

◆ [2/4]

struct { ... } ::keyedMutex

◆ [3/4]

struct { ... } hipExternalSemaphoreWaitParams::keyedMutex

◆ [4/4]

struct { ... } ::keyedMutex

◆ l2CacheSize

int hipDeviceProp_t::l2CacheSize

L2 cache size.

◆ localSizeBytes

size_t hipFuncAttributes::localSizeBytes

◆ major

int hipDeviceProp_t::major

Major compute capability. On HCC, this is an approximation and features may differ from CUDA CC. See the arch feature flags for portable ways to query feature caps.

◆ managedMemory

int hipDeviceProp_t::managedMemory

Device supports allocating managed memory on this system.

◆ maxDynamicSharedSizeBytes

int hipFuncAttributes::maxDynamicSharedSizeBytes

◆ maxGridSize

int hipDeviceProp_t::maxGridSize[3]

Max grid dimensions (XYZ).

◆ maxSharedMemoryPerMultiProcessor

size_t hipDeviceProp_t::maxSharedMemoryPerMultiProcessor

Maximum Shared Memory Per Multiprocessor.

◆ maxTexture1D

int hipDeviceProp_t::maxTexture1D

Maximum number of elements in 1D images.

◆ maxTexture1DLinear

int hipDeviceProp_t::maxTexture1DLinear

Maximum size for 1D textures bound to linear memory.

◆ maxTexture2D

int hipDeviceProp_t::maxTexture2D[2]

Maximum dimensions (width, height) of 2D images, in image elements.

◆ maxTexture3D

int hipDeviceProp_t::maxTexture3D[3]

Maximum dimensions (width, height, depth) of 3D images, in image elements.

◆ maxThreadsDim

int hipDeviceProp_t::maxThreadsDim[3]

Max number of threads in each dimension (XYZ) of a block.

◆ maxThreadsPerBlock [1/2]

int hipDeviceProp_t::maxThreadsPerBlock

Max work items per work group or workgroup max size.

◆ maxThreadsPerBlock [2/2]

int hipFuncAttributes::maxThreadsPerBlock

◆ maxThreadsPerMultiProcessor

int hipDeviceProp_t::maxThreadsPerMultiProcessor

Maximum resident threads per multi-processor.

◆ memoryBusWidth

int hipDeviceProp_t::memoryBusWidth

Global memory bus width in bits.

◆ memoryClockRate

int hipDeviceProp_t::memoryClockRate

Max global memory clock frequency in khz.

◆ memoryType

enum hipMemoryType hipPointerAttribute_t::memoryType

◆ memPitch

size_t hipDeviceProp_t::memPitch

Maximum pitch in bytes allowed by memory copies.

◆ minor

int hipDeviceProp_t::minor

Minor compute capability. On HCC, this is an approximation and features may differ from CUDA CC. See the arch feature flags for portable ways to query feature caps.

◆ multiProcessorCount

int hipDeviceProp_t::multiProcessorCount

Number of multi-processors (compute units).

◆ name [1/5]

char hipDeviceProp_t::name[256]

Device name.

◆ name [2/5]

const void* hipExternalMemoryHandleDesc::name

◆ [3/5]

const { ... } ::name

◆ name [4/5]

const void* hipExternalSemaphoreHandleDesc::name

◆ [5/5]

const { ... } ::name

◆ numRegs

int hipFuncAttributes::numRegs

◆ offset

unsigned long long hipExternalMemoryBufferDesc::offset

◆ pageableMemoryAccess

int hipDeviceProp_t::pageableMemoryAccess

Device supports coherently accessing pageable memory without calling hipHostRegister on it

◆ pageableMemoryAccessUsesHostPageTables

int hipDeviceProp_t::pageableMemoryAccessUsesHostPageTables

Device accesses pageable memory via the host's page tables.

◆ [1/2]

struct { ... } hipExternalSemaphoreSignalParams::params

◆ [2/2]

struct { ... } hipExternalSemaphoreWaitParams::params

◆ pciBusID

int hipDeviceProp_t::pciBusID

PCI Bus ID.

◆ pciDeviceID

int hipDeviceProp_t::pciDeviceID

PCI Device ID.

◆ pciDomainID

int hipDeviceProp_t::pciDomainID

PCI Domain ID.

◆ pitch

size_t hipMemsetParams::pitch

◆ preferredShmemCarveout

int hipFuncAttributes::preferredShmemCarveout

◆ ptxVersion

int hipFuncAttributes::ptxVersion

◆ regsPerBlock

int hipDeviceProp_t::regsPerBlock

Registers per block.

◆ [1/6]

unsigned { ... } ::reserved[10]

◆ reserved [2/6]

unsigned int hipExternalSemaphoreWaitParams::reserved

◆ [3/6]

unsigned { ... } ::reserved[12]

◆ reserved [4/6]

unsigned int hipExternalSemaphoreSignalParams::reserved

◆ reserved [5/6]

char hipIpcMemHandle_t::reserved[HIP_IPC_HANDLE_SIZE]

◆ reserved [6/6]

char hipIpcEventHandle_t::reserved[HIP_IPC_HANDLE_SIZE]

◆ sharedMem

size_t hipLaunchParams::sharedMem

Shared memory.

◆ sharedMemBytes

unsigned int hipKernelNodeParams::sharedMemBytes

◆ sharedMemPerBlock

size_t hipDeviceProp_t::sharedMemPerBlock

Size of shared memory region (in bytes).

◆ sharedSizeBytes

size_t hipFuncAttributes::sharedSizeBytes

◆ size [1/2]

unsigned long long hipExternalMemoryHandleDesc::size

◆ size [2/2]

unsigned long long hipExternalMemoryBufferDesc::size

◆ stream

hipStream_t hipLaunchParams::stream

Stream identifier.

◆ tccDriver

int hipDeviceProp_t::tccDriver

1:If device is Tesla device using TCC driver, else 0

◆ textureAlignment

size_t hipDeviceProp_t::textureAlignment

Alignment requirement for textures.

◆ texturePitchAlignment

size_t hipDeviceProp_t::texturePitchAlignment

Pitch alignment requirement for texture references bound to pitched memory.

◆ [1/2]

unsigned { ... } ::timeoutMs

◆ timeoutMs [2/2]

unsigned int hipExternalSemaphoreWaitParams::timeoutMs

◆ totalConstMem

size_t hipDeviceProp_t::totalConstMem

Size of shared memory region (in bytes).

◆ totalGlobalMem

size_t hipDeviceProp_t::totalGlobalMem

Size of global memory region (in bytes).

◆ type [1/2]

hipExternalMemoryHandleType hipExternalMemoryHandleDesc::type

◆ type [2/2]

hipExternalSemaphoreHandleType hipExternalSemaphoreHandleDesc::type

◆ userData

void* hipHostNodeParams::userData

◆ value [1/5]

unsigned long long hipExternalSemaphoreSignalParams::value

◆ [2/5]

unsigned { ... } ::value

◆ [3/5]

unsigned { ... } ::value

◆ value [4/5]

unsigned long long hipExternalSemaphoreWaitParams::value

◆ value [5/5]

unsigned int hipMemsetParams::value

◆ warpSize

int hipDeviceProp_t::warpSize

Warp size.

◆ width

size_t hipMemsetParams::width

◆ [1/4]

struct { ... } hipExternalMemoryHandleDesc::win32

◆ [2/4]

struct { ... } ::win32

◆ [3/4]

struct { ... } ::win32

◆ [4/4]

struct { ... } hipExternalSemaphoreHandleDesc::win32

◆ x

uint32_t dim3::x

x

◆ y

uint32_t dim3::y

y

◆ z

uint32_t dim3::z

z

HIP API

HIP API#

Modules

Variables

Detailed Description

Variable Documentation

◆ allocationFlags

◆ arch

◆ args

◆ asicRevision

◆ binaryVersion

◆ blockDim [1/2]

◆ blockDim [2/2]

◆ cacheModeCA

◆ canMapHostMemory

◆ clockInstructionRate

◆ clockRate

◆ computeMode

◆ concurrentKernels

◆ concurrentManagedAccess

◆ constSizeBytes

◆ cooperativeLaunch

◆ cooperativeMultiDeviceLaunch

◆ cooperativeMultiDeviceUnmatchedBlockDim

◆ cooperativeMultiDeviceUnmatchedFunc

◆ cooperativeMultiDeviceUnmatchedGridDim

◆ cooperativeMultiDeviceUnmatchedSharedMem

◆ device

◆ devicePointer

◆ directManagedMemAccessFromHost

◆ dst

◆ ECCEnabled

◆ elementSize

◆ extra

◆ fd [1/4]

◆ [2/4]

◆ fd [3/4]

◆ [4/4]

◆ [1/4]

◆ [2/4]

◆ [3/4]

◆ [4/4]

◆ flags [1/5]

◆ flags [2/5]

◆ flags [3/5]

◆ flags [4/5]

◆ flags [5/5]

◆ fn

◆ func [1/2]

◆ func [2/2]

◆ gcnArch

◆ gcnArchName

◆ gridDim [1/2]

◆ gridDim [2/2]

◆ handle [1/6]

◆ [2/6]

◆ [3/6]

◆ handle [4/6]

◆ [5/6]

◆ [6/6]

◆ has3dGrid

◆ hasDoubles

◆ hasDynamicParallelism

◆ hasFloatAtomicAdd

◆ hasFunnelShift

◆ hasGlobalFloatAtomicExch

◆ hasGlobalInt32Atomics

◆ hasGlobalInt64Atomics

◆ hasSharedFloatAtomicExch

◆ hasSharedInt32Atomics

◆ hasSharedInt64Atomics

◆ hasSurfaceFuncs

◆ hasSyncThreadsExt

◆ hasThreadFenceSystem

◆ hasWarpBallot

◆ hasWarpShuffle

◆ hasWarpVote

◆ hdpMemFlushCntl

◆ hdpRegFlushCntl

◆ height