#include <szx_stage.h>

Inheritance diagram for fz::SZxStage< T >:

Public Member Functions
void	setInverse (bool inv) override

bool	isGraphCompatible () const override

double	getConstantBlockFraction () const

void	execute (fz::stream_t stream, MemoryPool pool, const std::vector< void > &inputs, const std::vector< void * > &outputs, const std::vector< size_t > &sizes) override

void	postStreamSync (fz::stream_t stream) override

std::string	getName () const override

std::vector< size_t >	estimateOutputSizes (const std::vector< size_t > &input_sizes) const override

size_t	estimateScratchBytes (const std::vector< size_t > &input_sizes) const override

std::unordered_map< std::string, size_t >	getActualOutputSizesByName () const override

size_t	getActualOutputSize (int index) const override

std::vector< std::string >	getRunNotes () const override

uint16_t	getStageTypeId () const override

uint8_t	getOutputDataType (size_t) const override

uint8_t	getInputDataType (size_t) const override

size_t	serializeHeader (size_t, uint8_t *buf, size_t max_size) const override

void	deserializeHeader (const uint8_t *buf, size_t size) override

size_t	getMaxHeaderSize (size_t) const override

void	saveState () override

Public Member Functions inherited from fz::Stage
virtual size_t	getRequiredInputAlignment () const

virtual std::vector< std::string >	getOutputNames () const

int	getOutputIndex (const std::string &name) const

virtual void	setDims (const std::array< size_t, 3 > &dims)

virtual void	onFinalize (size_t, MemoryPool *)

virtual size_t	estimateDeviceFootprintBytes (size_t) const

virtual size_t	estimatePinnedFootprintBytes (size_t) const

virtual void	setTerminalOutput (bool terminal)

virtual FusionSpec	getFusionSpec () const

virtual FusedOpDecl	getFusedOp () const

virtual EncodingOracleDecl	getEncodingOracle () const

virtual bool	bindDownstreamEncodingOracle (const EncodingOracleDecl &)

virtual std::vector< FusedAuxOutputDecl >	getFusedAuxOutputs () const

virtual void	primeFusedForwardState (const FusedPrimeContext &)

virtual void	setFusedArchiveResult (size_t, size_t)

virtual void	setFusedSideOutput (int, size_t)

Additional Inherited Members
Static Public Member Functions inherited from fz::Stage
static constexpr bool	isSupportedOnBackend ()

Detailed Description

template<typename T>
class fz::SZxStage< T >

SZx ultrafast error-bounded compressor.

Per block of block_size elements the forward pass:

scans block min/max;
classifies the block as constant when max - min <= 2*eb (the whole block is representable by one value within the bound) or non-constant otherwise — a 2-bit type code per block;
constant blocks emit only a block reference value (the mean);
non-constant blocks subtract the block reference, quantize the residuals to fixed-length integers within [-2^b, 2^b) where b is the block's required bit width, and bit-pack them (no Huffman).

The archive is: [meta region: 2-bit type codes + per-block bit widths] followed by [payload region: reference values + packed residuals]. Output size is data dependent (constant blocks are ~one value), so estimateOutputSizes() returns a safe upper bound and postStreamSync() reads the true size back.

Template Parameters

T	float or double.

Member Function Documentation

◆ setInverse()

template<typename T >

void fz::SZxStage< T >::setInverse ( bool inverse )

inlineoverridevirtual

Switch between forward (compression) and inverse (decompression) mode. Affects getNumInputs()/getNumOutputs() for stages with asymmetric port counts.

Reimplemented from fz::Stage.

◆ isGraphCompatible()

template<typename T >

bool fz::SZxStage< T >::isGraphCompatible ( ) const

inlineoverridevirtual

Whether this stage is safe inside a CUDA Graph capture.

A stage is graph-compatible if execute() enqueues only device-side work (kernel launches, cudaMemcpyAsync D2D/H2D) and makes no host-synchronous calls. Override and return false if execute() contains D2H copies or dynamic decisions based on device data — the DAG will throw at setCaptureMode(true) time rather than producing a broken graph.

Default: true. Inverse-mode stages that do D2H reads (e.g. RZE inverse) must return false.

Reimplemented from fz::Stage.

◆ getConstantBlockFraction()

template<typename T >

double fz::SZxStage< T >::getConstantBlockFraction ( ) const

inline

Fraction of blocks classified constant on the last forward encode — a cheap compressibility probe, reported through getRunNotes().

◆ execute()

template<typename T >

void fz::SZxStage< T >::execute	(	fz::stream_t	stream,
		MemoryPool *	pool,
		const std::vector< void * > &	inputs,
		const std::vector< void * > &	outputs,
		const std::vector< size_t > &	sizes
	)

overridevirtual

Execute the stage. Inputs, outputs, and sizes are device pointers/bytes.

Stages may call cudaStreamSynchronize(stream) or issue blocking D2H copies when the algorithm requires it (e.g. Huffman histogram readback for codebook construction, ANS renormalization tables). Such stages must return false from isGraphCompatible() and must document the sync points.

Note: the DAG dispatches sibling nodes (same topological level) via a sequential CPU loop, each enqueuing to its own stream. A sync inside execute() blocks the CPU from dispatching subsequent siblings until the synced stream is idle — this delays parallel branches in wide DAGs. In a linear pipeline there are no siblings and no extra cost.

Implements fz::Stage.

◆ postStreamSync()

template<typename T >

void fz::SZxStage< T >::postStreamSync ( fz::stream_t stream )

overridevirtual

Called after dag->execute() and stream sync, before compress() returns. Use for D2H transfers that must not block mid-pipeline (e.g. Lorenzo's outlier count readback). The stream is already idle so a plain cudaMemcpy is safe here.

Reimplemented from fz::Stage.

◆ getName()

template<typename T >

std::string fz::SZxStage< T >::getName ( ) const

inlineoverridevirtual

Human-readable name used in error messages and debug output.

Implements fz::Stage.

◆ estimateOutputSizes()

template<typename T >

std::vector< size_t > fz::SZxStage< T >::estimateOutputSizes ( const std::vector< size_t > & input_sizes ) const

overridevirtual

Estimate output buffer sizes given input sizes. Used for buffer allocation planning in PREALLOCATE mode — must be a safe upper bound; under-estimation causes buffer overruns.

Implements fz::Stage.

◆ estimateScratchBytes()

template<typename T >

size_t fz::SZxStage< T >::estimateScratchBytes ( const std::vector< size_t > & input_sizes ) const

overridevirtual

Peak persistent scratch bytes this stage holds in the MemoryPool.

Only count allocations that are drawn from the pool and kept alive across execute() calls. Transient scratch freed within execute() is already captured by the pool's high-water mark and must not be included. Used by CompressionDAG::computeTopoPoolSize() to size the release threshold.

Reimplemented from fz::Stage.

◆ getActualOutputSizesByName()

template<typename T >

std::unordered_map< std::string, size_t > fz::SZxStage< T >::getActualOutputSizesByName ( ) const

inlineoverridevirtual

Actual output sizes after execute(), keyed by output port name.

Implements fz::Stage.

◆ getActualOutputSize()

template<typename T >

size_t fz::SZxStage< T >::getActualOutputSize ( int index ) const

inlineoverridevirtual

Actual size of a single output by index after execute(). Avoids constructing the map for the common single-output case. Default delegates to getActualOutputSizesByName(); override to return directly from an internal field.

Reimplemented from fz::Stage.

◆ getRunNotes()

template<typename T >

std::vector< std::string > fz::SZxStage< T >::getRunNotes ( ) const

overridevirtual

Notes about what this stage actually did on the last run, when that differs from what was configured in a way that affects comparability.

Motivating case: HuffmanStage silently falls back to an Adaptive book when a PerBlock/Fixed build drives a symbol past the 27-bit code field. The fallback is correct — it does not relax the error bound — but a field encoded with a different codebook is not compression-ratio comparable to one that was not, and getBookSource() deliberately keeps reporting what was asked for. Without a channel like this, a benchmark row records the two cases identically and the difference is unrecoverable after the fact.

Returns short stable machine-readable tokens (e.g. "adaptive_fallback"), not prose — these are meant to land in a benchmark row and be grouped on. Empty by default; a stage that never surprises its caller need not implement it.

Reimplemented from fz::Stage.

◆ getStageTypeId()

template<typename T >

uint16_t fz::SZxStage< T >::getStageTypeId ( ) const

inlineoverridevirtual

Stage type identifier written into the FZM file header.

Implements fz::Stage.

◆ getOutputDataType()

template<typename T >

uint8_t fz::SZxStage< T >::getOutputDataType ( size_t output_index ) const

inlineoverridevirtual

DataType enum of the given output port.

Implements fz::Stage.

◆ getInputDataType()

template<typename T >

uint8_t fz::SZxStage< T >::getInputDataType ( size_t ) const

inlineoverridevirtual

Expected DataType of the given input port.

Used by Pipeline::finalize() to detect type mismatches between connected stages before any execution. Return DataType::UNKNOWN to opt out of checking — byte-transparent stages (Bitshuffle, RZE, RRE) and mock stages must return UNKNOWN; finalize() skips any connection where either side is UNKNOWN.

Reimplemented from fz::Stage.

◆ serializeHeader()

template<typename T >

size_t fz::SZxStage< T >::serializeHeader	(	size_t	output_index,
		uint8_t *	header_buffer,
		size_t	max_size
	)		const

inlineoverridevirtual

Serialize stage config into header_buffer (max 128 bytes) for the FZM file. Return the number of bytes written, or 0 if the stage has no config.

Reimplemented from fz::Stage.

◆ deserializeHeader()

template<typename T >

void fz::SZxStage< T >::deserializeHeader	(	const uint8_t *	header_buffer,
		size_t	size
	)

inlineoverridevirtual

Restore stage config from header_buffer during decompression.

Reimplemented from fz::Stage.

◆ getMaxHeaderSize()

template<typename T >

size_t fz::SZxStage< T >::getMaxHeaderSize ( size_t output_index ) const

inlineoverridevirtual

Maximum bytes this stage writes into its per-output FZM header slot.

Reimplemented from fz::Stage.

◆ saveState()

template<typename T >

void fz::SZxStage< T >::saveState ( )

inlineoverridevirtual

Save/restore config state around a decompression pass. deserializeHeader() overwrites the stage's forward-pass config; saveState() is called before and restoreState() after so the stage returns to its original configuration.

Reimplemented from fz::Stage.

Public Member Functions

Additional Inherited Members

Detailed Description

Member Function Documentation

◆ setInverse()

◆ isGraphCompatible()

◆ getConstantBlockFraction()

◆ execute()

◆ postStreamSync()

◆ getName()

◆ estimateOutputSizes()

◆ estimateScratchBytes()

◆ getActualOutputSizesByName()

◆ getActualOutputSize()

◆ getRunNotes()

◆ getStageTypeId()

◆ getOutputDataType()

◆ getInputDataType()

◆ serializeHeader()

◆ deserializeHeader()

◆ getMaxHeaderSize()

◆ saveState()