ne::PipelineManager::ComputePipelines Struct Reference

Classes
struct	Resources

Public Member Functions
	ComputePipelines (const ComputePipelines &)=delete
ComputePipelines &	operator= (const ComputePipelines &)=delete
std::variant< PipelineSharedPtr, Error >	getComputePipelineForShader (PipelineManager *pPipelineManager, ComputeShaderInterface *pComputeShaderInterface)
std::optional< Error >	onPipelineNoLongerUsedByComputeShaderInterface (const std::string &sComputeShaderName, ComputeShaderInterface *pComputeShaderInterface)
std::optional< Error >	queueShaderExecutionOnGraphicsQueue (ComputeShaderInterface *pComputeShaderInterface)
size_t	getComputePipelineCount ()
std::pair< std::recursive_mutex *, QueuedForExecutionComputeShaders * >	getComputeShadersForGraphicsQueueExecution ()

Static Private Member Functions
static std::optional< Error >	queueComputeShaderInterfaceForExecution (std::unordered_map< Pipeline *, std::unordered_set< ComputeShaderInterface * > > &pipelineShaders, ComputeShaderInterface *pComputeShaderInterface)

Private Attributes
std::pair< std::recursive_mutex, Resources >	mtxResources

Detailed Description

Groups information about compute pipelines.

Member Function Documentation

getComputePipelineCount()

size_t ne::PipelineManager::ComputePipelines::getComputePipelineCount

(

)

Returns the total number of existing compute pipelines.

Returns

Number of compute pipelines.

getComputePipelineForShader()

std::variant< PipelineSharedPtr, Error > ne::PipelineManager::ComputePipelines::getComputePipelineForShader	(	PipelineManager *	pPipelineManager,
		ComputeShaderInterface *	pComputeShaderInterface
	)

Look for already created pipeline that uses the specified shader and returns it, otherwise creates a new pipeline.

Remarks

If creating a new pipeline, loads the specified shader from disk into the memory, it will be released from the memory once the pipeline object is destroyed (not the shared pointer) and no other object is using it.

Parameters

pPipelineManager	Pipeline manager.
pComputeShaderInterface	Compute shader interface to add.

Returns

Error if something went wrong, otherwise compute pipeline.

getComputeShadersForGraphicsQueueExecution()

std::pair< std::recursive_mutex *, QueuedForExecutionComputeShaders * > ne::PipelineManager::ComputePipelines::getComputeShadersForGraphicsQueueExecution ( )

inline

Returns all compute shaders and their pipelines to be executed on the graphics queue.

Warning

Do not delete (free) returned pointers.

Returns

Shaders and pipelines.

onPipelineNoLongerUsedByComputeShaderInterface()

std::optional< Error > ne::PipelineManager::ComputePipelines::onPipelineNoLongerUsedByComputeShaderInterface	(	const std::string &	sComputeShaderName,
		ComputeShaderInterface *	pComputeShaderInterface
	)

Removes the specified compute shader interface and if no other interface references the compute pipeline (that the shader used) also destroys the pipeline.

Remarks

If you used getComputePipelineForShader to get a compute pipeline for your shader you don't need to call this function as it will be automatically called by PipelineSharedPtr's destructor.

Parameters

sComputeShaderName	Name of the compute shader that compute pipeline uses.
pComputeShaderInterface	Compute shader interface to remove.

Returns

Error if something went wrong.

queueComputeShaderInterfaceForExecution()

std::optional< Error > ne::PipelineManager::ComputePipelines::queueComputeShaderInterfaceForExecution ( std::unordered_map< Pipeline *, std::unordered_set< ComputeShaderInterface * > > &  pipelineShaders, ComputeShaderInterface *  pComputeShaderInterface  )

staticprivate

Adds a compute shader interface to the specified map to be executed.

Warning

Expects that mtxResources is locked during the function call.

Parameters

pipelineShaders	Map to add the new interface to.
pComputeShaderInterface	Interface to add.

Returns

Error if something went wrong.

queueShaderExecutionOnGraphicsQueue()

std::optional< Error > ne::PipelineManager::ComputePipelines::queueShaderExecutionOnGraphicsQueue

(

ComputeShaderInterface *

pComputeShaderInterface

)

Adds a compute shader interface to be executed on the graphics queue according to shader's execution stage and group.

Remarks

Added shader will be executed only once, if you want your shader to be executed again you would need to call this function again but later after a frame was submitted (if you call it right now nothing will happen as it's already queued).

Parameters

pComputeShaderInterface

Compute shader interface to queue for execution.

Returns

Error if something went wrong.

Member Data Documentation

mtxResources

std::pair<std::recursive_mutex, Resources> ne::PipelineManager::ComputePipelines::mtxResources

private

Pipeline data.

---

The documentation for this struct was generated from the following files:

• src/engine_lib/private/render/general/pipeline/PipelineManager.h
• src/engine_lib/private/render/general/pipeline/PipelineManager.cpp