Crate nccl_sys

Structs

CUevent_st

CUgraphDeviceUpdatableNode_st

CUgraphNode_st

CUgraph_st

CUstream_st

cudaAccessPolicyWindow

Specifies an access policy for a window, a contiguous extent of memory beginning at base_ptr and ending at base_ptr + num_bytes. Partition into many segments and assign segments such that. sum of “hit segments” / window == approx. ratio. sum of “miss segments” / window == approx 1-ratio. Segments and ratio specifications are fitted to the capabilities of the architecture. Accesses in a hit segment apply the hitProp access policy. Accesses in a miss segment apply the missProp access policy.

cudaAccessProperty

Specifies performance hint with ::cudaAccessPolicyWindow for hitProp and missProp members.

cudaClusterSchedulingPolicy

Cluster scheduling policies. These may be passed to ::cudaFuncSetAttribute

cudaError

CUDA error types

cudaError_t

CUDA Error types CUDA error types

cudaGraphEdgeData_st

Optional annotation for edges in a CUDA graph. Note, all edges implicitly have annotations and default to a zero-initialized value if not specified. A zero-initialized struct indicates a standard full serialization of two nodes with memory visibility.

cudaLaunchAttributeID

Launch attributes enum; used as id field of ::cudaLaunchAttribute

cudaLaunchAttributeValue__bindgen_ty_1

Value of launch attribute ::cudaLaunchAttributeClusterDimension that represents the desired cluster dimensions for the kernel. Opaque type with the following fields: - \p x - The X dimension of the cluster, in blocks. Must be a divisor of the grid X dimension. - \p y - The Y dimension of the cluster, in blocks. Must be a divisor of the grid Y dimension. - \p z - The Z dimension of the cluster, in blocks. Must be a divisor of the grid Z dimension.

cudaLaunchAttributeValue__bindgen_ty_2

Value of launch attribute ::cudaLaunchAttributeProgrammaticEvent with the following fields: - \p cudaEvent_t event - Event to fire when all blocks trigger it. - \p int flags; - Event record flags, see ::cudaEventRecordWithFlags. Does not accept ::cudaEventRecordExternal. - \p int triggerAtBlockStart - If this is set to non-0, each block launch will automatically trigger the event.

cudaLaunchAttributeValue__bindgen_ty_3

Value of launch attribute ::cudaLaunchAttributeLaunchCompletionEvent with the following fields: - \p cudaEvent_t event - Event to fire when the last block launches. - \p int flags - Event record flags, see ::cudaEventRecordWithFlags. Does not accept ::cudaEventRecordExternal.

cudaLaunchAttributeValue__bindgen_ty_4

Value of launch attribute ::cudaLaunchAttributeDeviceUpdatableKernelNode with the following fields:

cudaLaunchMemSyncDomain

Memory Synchronization Domain

cudaLaunchMemSyncDomainMap_st

Memory Synchronization Domain map

cudaStreamCaptureMode

Possible modes for stream capture thread interactions. For more details see ::cudaStreamBeginCapture and ::cudaThreadExchangeStreamCaptureMode

cudaStreamCaptureStatus

Possible stream capture statuses returned by ::cudaStreamIsCapturing

cudaSynchronizationPolicy

ncclComm

ncclConfig_v22700

ncclDataType_t

ncclRedOp_t

ncclResult_t

ncclScalarResidence_t

ncclSimInfo_v22200

ncclUniqueId

Constants

NCCL_MAJOR

NCCL_MINOR

NCCL_PATCH

NCCL_SPLIT_NOCOLOR

Functions

cudaSetDevice^⚠

\brief Set device to be used for GPU executions

cudaStreamAddCallback^⚠

\brief Add a callback to a compute stream

cudaStreamAttachMemAsync^⚠

cudaStreamBeginCapture^⚠

\brief Begins graph capture on a stream

cudaStreamBeginCaptureToGraph^⚠

\brief Begins graph capture on a stream to an existing graph

cudaStreamCopyAttributes^⚠

\brief Copies attributes from source stream to destination stream.

cudaStreamCreate^⚠

\brief Create an asynchronous stream

cudaStreamCreateWithFlags^⚠

\brief Create an asynchronous stream

cudaStreamCreateWithPriority^⚠

\brief Create an asynchronous stream with the specified priority

cudaStreamDestroy^⚠

\brief Destroys and cleans up an asynchronous stream

cudaStreamEndCapture^⚠

\brief Ends capture on a stream, returning the captured graph

cudaStreamGetAttribute^⚠

\brief Queries stream attribute.

cudaStreamGetCaptureInfo_v2^⚠

\brief Query a stream’s capture state

cudaStreamGetCaptureInfo_v3^⚠

\brief Query a stream’s capture state (12.3+)

cudaStreamGetFlags^⚠

\brief Query the flags of a stream

cudaStreamGetId^⚠

\brief Query the Id of a stream

cudaStreamGetPriority^⚠

\brief Query the priority of a stream

cudaStreamIsCapturing^⚠

\brief Returns a stream’s capture status

cudaStreamQuery^⚠

\brief Queries an asynchronous stream for completion status

cudaStreamSetAttribute^⚠

\brief Sets stream attribute.

cudaStreamSynchronize^⚠

\brief Waits for stream tasks to complete

cudaStreamUpdateCaptureDependencies^⚠

\brief Update the set of dependencies in a capturing stream (11.3+)

cudaStreamUpdateCaptureDependencies_v2^⚠

\brief Update the set of dependencies in a capturing stream (12.3+)

cudaStreamWaitEvent^⚠

\brief Make a compute stream wait on an event

ncclAllGather^⚠

ncclAllReduce^⚠

ncclBroadcast^⚠

ncclCommAbort^⚠

ncclCommCount^⚠

ncclCommCuDevice^⚠

ncclCommDeregister^⚠

ncclCommDestroy^⚠

ncclCommFinalize^⚠

ncclCommGetAsyncError^⚠

ncclCommInitAll^⚠

ncclCommInitRank^⚠

ncclCommInitRankConfig^⚠

ncclCommInitRankScalable^⚠

ncclCommRegister^⚠

ncclCommSplit^⚠

ncclCommUserRank^⚠

ncclGetErrorString^⚠

ncclGetLastError^⚠

ncclGetUniqueId^⚠

ncclGetVersion^⚠

ncclGroupEnd^⚠

ncclGroupSimulateEnd^⚠

ncclGroupStart^⚠

ncclMemAlloc^⚠

ncclMemFree^⚠

ncclRecv^⚠

ncclRedOpCreatePreMulSum^⚠

ncclRedOpDestroy^⚠

ncclReduce^⚠

ncclReduceScatter^⚠

ncclSend^⚠

Type Aliases

cudaEvent_t

CUDA event types

cudaGraphDeviceNode_t

CUDA device node handle for device-side node update

cudaGraphEdgeData

Optional annotation for edges in a CUDA graph. Note, all edges implicitly have annotations and default to a zero-initialized value if not specified. A zero-initialized struct indicates a standard full serialization of two nodes with memory visibility.

cudaGraphNode_t

CUDA graph node.

cudaGraph_t

CUDA graph

cudaLaunchMemSyncDomainMap

Memory Synchronization Domain map

cudaStreamCallback_t

Type of stream callback functions. \param stream The stream as passed to ::cudaStreamAddCallback, may be NULL. \param status ::cudaSuccess or any persistent error on the stream. \param userData User parameter provided at registration.

cudaStream_t

CUDA stream

ncclComm_t

ncclConfig_t

ncclSimInfo_t

Unions

cudaLaunchAttributeValue

Launch attributes union; used as value field of ::cudaLaunchAttribute