VdfSMBLData holds per-output data that is meant to be consumed by the executor. This data is an optional part of VdfExecutorData and it is specific to sparse mung buffer locking. More...

#include <smblData.h>

Public Member Functions
	VdfSMBLData (const VdfSMBLData &)=delete

VdfSMBLData &	operator= (const VdfSMBLData &)=delete

	VdfSMBLData ()

VDF_API	~VdfSMBLData ()

Memoized mask computations
VdfMask	InvalidateCacheMask (const VdfMask &cacheMask, const VdfMask &invalidationMask)

void	ExtendLockedCacheMask (VdfMask *lockedCacheMask, const VdfMask &cacheMask)

void	RemoveUncachedMask (VdfMask *lockedCacheMask, const VdfMask &cacheMask, const VdfMask &keepMask)

bool	ComputeAffectiveness (const VdfMask &lockedCacheMask, const VdfMask &affectsMask)

Local Cache
VdfVector *	Retain (const VdfOutputSpec &spec, VdfVector *cache, const VdfMask &cacheMask)

void	Release ()

VDF_API void	Clear ()

VdfVector *	GetCache () const

const VdfMask &	GetCacheMask () const

bool	HasCache () const

Detailed Description

VdfSMBLData holds per-output data that is meant to be consumed by the executor. This data is an optional part of VdfExecutorData and it is specific to sparse mung buffer locking.

Definition at line 30 of file smblData.h.

Constructor & Destructor Documentation

VdfSMBLData::VdfSMBLData ( const VdfSMBLData & )

delete

Noncopyable.

VdfSMBLData::VdfSMBLData ( )

inline

Constructs an SMBL data object.

Definition at line 40 of file smblData.h.

VDF_API VdfSMBLData::~VdfSMBLData ( )

Member Function Documentation

VDF_API void VdfSMBLData::Clear ( )

Clear any of the data this object is holding on to.

bool VdfSMBLData::ComputeAffectiveness	(	const VdfMask &	lockedCacheMask,
		const VdfMask &	affectsMask
	)

inline

Computes the affectiveness of the corresponding output given the accumulated lockedCacheMask and the scheduled affectsMask of the output. This is a memoized computation.

Definition at line 87 of file smblData.h.

void VdfSMBLData::ExtendLockedCacheMask	(	VdfMask *	lockedCacheMask,
		const VdfMask &	cacheMask
	)

inline

Extends the lockedCacheMask by appending the bits stored in the executor cacheMask. This is a memoized computation.

Definition at line 65 of file smblData.h.

VdfVector* VdfSMBLData::GetCache ( ) const

inline

Returns a pointer to the locally retained cache, if any.

Definition at line 130 of file smblData.h.

const VdfMask& VdfSMBLData::GetCacheMask ( ) const

inline

Returns a mask indicating data available in the locally retained cache.

Definition at line 136 of file smblData.h.

bool VdfSMBLData::HasCache ( ) const

inline

Returns true, if a cache has been retained locally, and false if there is no such cache.

Definition at line 143 of file smblData.h.

VdfMask VdfSMBLData::InvalidateCacheMask	(	const VdfMask &	cacheMask,
		const VdfMask &	invalidationMask
	)

inline

Invalidates the executor cacheMask given an invalidationMask. Returns the cacheMask with the bits in the invalidationMask removed. This is a memoized computation.

Definition at line 240 of file smblData.h.

VdfSMBLData& VdfSMBLData::operator= ( const VdfSMBLData & )

delete

void VdfSMBLData::Release ( )

inline

Releases the cache, which has been retained by this object, if any. Note, that this method does NOT release ownership of any of its heap allocated data! It merely demotes the retained cache to a free cache for future use.

Definition at line 117 of file smblData.h.

void VdfSMBLData::RemoveUncachedMask	(	VdfMask *	lockedCacheMask,
		const VdfMask &	cacheMask,
		const VdfMask &	keepMask
	)

inline

Make sure that all the bits in the keepMask are provided by the cacheMask. Any data bits that are not provided by the cacheMask should not be contained in the lockedCacheMask, so remove them from the lockedCacheMask. This makes it so that nodes, which must provide data to be kept at the output, do not become un-affective.

Definition at line 250 of file smblData.h.

VdfVector * VdfSMBLData::Retain	(	const VdfOutputSpec &	spec,
		VdfVector *	cache,
		const VdfMask &	cacheMask
	)

inline

Locally retains the passed in cache with the given cacheMask. This method returns a pointer to a (new) cache, which the client is now free to use. Consequently, this method passes the ownership of the cache pointer to this object, while giving up the ownership of the returned vector. This is to avoid a copy of the underlying data.

Definition at line 268 of file smblData.h.

The documentation for this class was generated from the following file:

pxr/exec/vdf/smblData.h

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation