|
FZGPUModules 2.0
GPU-accelerated modular compression pipelines
|
rANS entropy coding stage (GPU, via vendored dietGPU kernel templates). More...
#include "stage/stage.h"#include "fzm_format.h"#include <cuda_runtime.h>#include <cstdint>#include <cstring>#include <stdexcept>#include <string>#include <unordered_map>#include <vector>Go to the source code of this file.
Namespaces | |
| namespace | fz |
rANS entropy coding stage (GPU, via vendored dietGPU kernel templates).
Forward: uint8_t[] → ANS compressed bitstream (ANSCoalescedHeader + rANS data). Inverse: ANS bitstream → uint8_t[].
D2H copies occur in both directions, so isGraphCompatible() returns false.
Only prob_bits=10 is supported in this build (baked into GpuANS.cu template instantiations). setProbBits() is reserved for future use; calling execute() after setProbBits(n != 10) throws std::runtime_error.
Serialized FZM header layout (12 bytes): [0] prob_bits (uint8_t) [1..3] reserved (3 bytes, zero) [4..11] original_bytes_ (uint64_t LE, uncompressed byte count)