ddidderr/zstd-rs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

feat(rust): migrate high-level runtime paths

Browse Source 
Move long-distance matching and high-level decompression from C shims into
Rust. The decoder now owns context, dictionary, parameter, one-shot, and
buffered streaming state while C retains allocation/configuration, legacy,
and trace leaves.

Move CLI parsing, safety policy, and dispatch into a separate Rust static
archive. Keeping it separate prevents library builds from retaining FIO
symbols, while C continues to own file opening, replacement, and I/O.
Program targets now select matching compression/decompression archives.

The remaining C boundary is intentional: high-level compression, optimal
parsing, dictionary building, legacy callbacks, and CLI file I/O still need
migration.

Test Plan:
- cargo test --all-targets (native and i686)
- cargo test --all-targets in rust/cli (native and i686)
- CLI crate compression-only and decompression-only feature tests
- native and i686 fuzzer/zstreamtest runs, plus legacy and dictionary tests
- ZSTD_C_PREDICT and ZSTD_HEAPMODE=0 fuzzer coverage
- library, dynamic-link, and program-target build/round-trip matrix

Refs: rust/README.md
This commit is contained in:
ddidderr
2026-07-11 09:03:41 +02:00
parent 031592da1e
commit fe7e24c770
14 changed files with 6886 additions and 4651 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/compress/zstd_ldm.c
+274 -676
View File
@@ -8,265 +8,225 @@
* You may select, at your option, one of the above-listed licenses.
*/
/* The LDM algorithms live in rust/src/zstd_ldm.rs. This file retains private
* match-state extraction and block-compressor dispatch only. */
#include "zstd_ldm.h"
#include "../common/debug.h"
#include "../common/xxhash.h"
#include "zstd_fast.h" /* ZSTD_fillHashTable() */
#include "zstd_double_fast.h" /* ZSTD_fillDoubleHashTable() */
#include "zstd_fast.h"
#include "zstd_double_fast.h"
#include "zstd_ldm_geartab.h"
#define LDM_BUCKET_SIZE_LOG 4
#define LDM_MIN_MATCH_LENGTH 64
#define LDM_HASH_RLOG 7
typedef struct {
U32 offset;
U32 checksum;
} ZSTD_rust_ldm_entry_layout;
typedef struct {
U64 rolling;
U64 stopMask;
} ldmRollingHashState_t;
U32 offset;
U32 litLength;
U32 matchLength;
} ZSTD_rust_raw_seq_layout;
/** ZSTD_ldm_gear_init():
*
* Initializes the rolling hash state such that it will honor the
* settings in params. */
static void ZSTD_ldm_gear_init(ldmRollingHashState_t* state, ldmParams_t const* params)
typedef struct {
rawSeq* seq;
size_t pos;
size_t posInSequence;
size_t size;
size_t capacity;
} ZSTD_rust_raw_seq_store_layout;
typedef struct {
ZSTD_ParamSwitch_e enableLdm;
U32 hashLog;
U32 bucketSizeLog;
U32 minMatchLength;
U32 hashRateLog;
U32 windowLog;
} ZSTD_rust_ldm_params_layout;
typedef struct {
BYTE const* nextSrc;
BYTE const* base;
BYTE const* dictBase;
U32 dictLimit;
U32 lowLimit;
U32 nbOverflowCorrections;
} ZSTD_rust_ldm_window_layout;
typedef char ZSTD_rust_ldm_entry_layout_check[
sizeof(ldmEntry_t) == sizeof(ZSTD_rust_ldm_entry_layout) ? 1 : -1];
typedef char ZSTD_rust_raw_seq_layout_check[
sizeof(rawSeq) == sizeof(ZSTD_rust_raw_seq_layout) ? 1 : -1];
typedef char ZSTD_rust_raw_seq_store_layout_check[
sizeof(RawSeqStore_t) == sizeof(ZSTD_rust_raw_seq_store_layout) ? 1 : -1];
typedef char ZSTD_rust_ldm_params_layout_check[
sizeof(ldmParams_t) == sizeof(ZSTD_rust_ldm_params_layout) ? 1 : -1];
typedef char ZSTD_rust_ldm_window_layout_check[
sizeof(ZSTD_window_t) == sizeof(ZSTD_rust_ldm_window_layout) ? 1 : -1];
typedef char ZSTD_rust_param_switch_layout_check[
sizeof(ZSTD_ParamSwitch_e) == sizeof(int) ? 1 : -1];
#define ZSTD_RUST_LDM_OFFSET_CHECK(name, c_type, c_field, rust_type, rust_field) \
typedef char name[offsetof(c_type, c_field) == offsetof(rust_type, rust_field) ? 1 : -1]
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_entry_offset_check,
ldmEntry_t, offset,
ZSTD_rust_ldm_entry_layout, offset);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_entry_checksum_check,
ldmEntry_t, checksum,
ZSTD_rust_ldm_entry_layout, checksum);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_offset_check,
rawSeq, offset,
ZSTD_rust_raw_seq_layout, offset);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_lit_length_check,
rawSeq, litLength,
ZSTD_rust_raw_seq_layout, litLength);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_match_length_check,
rawSeq, matchLength,
ZSTD_rust_raw_seq_layout, matchLength);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_store_seq_check,
RawSeqStore_t, seq,
ZSTD_rust_raw_seq_store_layout, seq);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_store_pos_check,
RawSeqStore_t, pos,
ZSTD_rust_raw_seq_store_layout, pos);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_store_pos_in_sequence_check,
RawSeqStore_t, posInSequence,
ZSTD_rust_raw_seq_store_layout, posInSequence);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_store_size_check,
RawSeqStore_t, size,
ZSTD_rust_raw_seq_store_layout, size);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_raw_seq_store_capacity_check,
RawSeqStore_t, capacity,
ZSTD_rust_raw_seq_store_layout, capacity);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_enable_check,
ldmParams_t, enableLdm,
ZSTD_rust_ldm_params_layout, enableLdm);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_hash_log_check,
ldmParams_t, hashLog,
ZSTD_rust_ldm_params_layout, hashLog);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_bucket_size_log_check,
ldmParams_t, bucketSizeLog,
ZSTD_rust_ldm_params_layout, bucketSizeLog);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_min_match_length_check,
ldmParams_t, minMatchLength,
ZSTD_rust_ldm_params_layout, minMatchLength);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_hash_rate_log_check,
ldmParams_t, hashRateLog,
ZSTD_rust_ldm_params_layout, hashRateLog);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_params_window_log_check,
ldmParams_t, windowLog,
ZSTD_rust_ldm_params_layout, windowLog);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_next_src_check,
ZSTD_window_t, nextSrc,
ZSTD_rust_ldm_window_layout, nextSrc);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_base_check,
ZSTD_window_t, base,
ZSTD_rust_ldm_window_layout, base);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_dict_base_check,
ZSTD_window_t, dictBase,
ZSTD_rust_ldm_window_layout, dictBase);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_dict_limit_check,
ZSTD_window_t, dictLimit,
ZSTD_rust_ldm_window_layout, dictLimit);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_low_limit_check,
ZSTD_window_t, lowLimit,
ZSTD_rust_ldm_window_layout, lowLimit);
ZSTD_RUST_LDM_OFFSET_CHECK(
ZSTD_rust_ldm_window_overflow_corrections_check,
ZSTD_window_t, nbOverflowCorrections,
ZSTD_rust_ldm_window_layout, nbOverflowCorrections);
#undef ZSTD_RUST_LDM_OFFSET_CHECK
typedef char ZSTD_rust_ldm_rep_count[(ZSTD_REP_NUM == 3) ? 1 : -1];
void ZSTD_rust_ldm_adjustParameters(
void* params, U32 windowLog, int strategy,
U32 hashLogMax, U32 bucketSizeLogMax, int btultra);
size_t ZSTD_rust_ldm_getTableSize(
const void* params, int enableLdm, size_t redzoneSize);
size_t ZSTD_rust_ldm_getMaxNbSeq(
const void* params, int enableLdm, size_t maxChunkSize);
void ZSTD_rust_ldm_fillHashTable(
void* hashTable, BYTE* bucketOffsets, const BYTE* base,
const BYTE* ip, const BYTE* iend, const void* params);
size_t ZSTD_rust_ldm_generateSequences(
void* hashTable, BYTE* bucketOffsets, void* window, U32* loadedDictEnd,
void* rawSeqStore, const void* params,
const void* src, size_t srcSize, int overflowCorrectFrequently);
void ZSTD_rust_ldm_skipSequences(void* rawSeqStore, size_t srcSize, U32 minMatch);
void ZSTD_rust_ldm_skipRawSeqStoreBytes(void* rawSeqStore, size_t nbBytes);
size_t ZSTD_rust_ldm_blockCompress(
void* rawSeqStore, void* blockContext, void* seqStore, U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize, U32 minMatch, int useOptimalParser);
const U64* ZSTD_ldm_rust_gearTable(void);
void ZSTD_ldm_rust_prepareBlock(void* blockContext, const void* anchor);
size_t ZSTD_ldm_rust_compressLiterals(
void* blockContext, void* seqStore, U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize);
void ZSTD_ldm_rust_storeSeq(
void* seqStore, size_t litLength, const void* literals, const void* litLimit,
U32 offBase, size_t matchLength);
void ZSTD_ldm_rust_setLdmSeqStore(void* blockContext, const void* rawSeqStore);
const U64* ZSTD_ldm_rust_gearTable(void)
{
unsigned maxBitsInMask = MIN(params->minMatchLength, 64);
unsigned hashRateLog = params->hashRateLog;
return ZSTD_ldm_gearTab;
}
state->rolling = ~(U32)0;
typedef struct {
ZSTD_MatchState_t* ms;
ZSTD_BlockCompressor_f blockCompressor;
} ZSTD_rust_ldm_block_context;
/* The choice of the splitting criterion is subject to two conditions:
* 1. it has to trigger on average every 2^(hashRateLog) bytes;
* 2. ideally, it has to depend on a window of minMatchLength bytes.
*
* In the gear hash algorithm, bit n depends on the last n bytes;
* so in order to obtain a good quality splitting criterion it is
* preferable to use bits with high weight.
*
* To match condition 1 we use a mask with hashRateLog bits set
* and, because of the previous remark, we make sure these bits
* have the highest possible weight while still respecting
* condition 2.
*/
if (hashRateLog > 0 && hashRateLog <= maxBitsInMask) {
state->stopMask = (((U64)1 << hashRateLog) - 1) << (maxBitsInMask - hashRateLog);
} else {
/* In this degenerate case we simply honor the hash rate. */
state->stopMask = ((U64)1 << hashRateLog) - 1;
static void ZSTD_rust_ldm_limitTableUpdate(ZSTD_MatchState_t* ms, const BYTE* anchor)
{
U32 const curr = (U32)(anchor - ms->window.base);
if (curr > ms->nextToUpdate + 1024) {
ms->nextToUpdate = curr - MIN(512, curr - ms->nextToUpdate - 1024);
}
}
/** ZSTD_ldm_gear_reset()
* Feeds [data, data + minMatchLength) into the hash without registering any
* splits. This effectively resets the hash state. This is used when skipping
* over data, either at the beginning of a block, or skipping sections.
*/
static void ZSTD_ldm_gear_reset(ldmRollingHashState_t* state,
BYTE const* data, size_t minMatchLength)
static void ZSTD_rust_ldm_fillFastTables(ZSTD_MatchState_t* ms, const BYTE* end)
{
U64 hash = state->rolling;
size_t n = 0;
#define GEAR_ITER_ONCE() do { \
hash = (hash << 1) + ZSTD_ldm_gearTab[data[n] & 0xff]; \
n += 1; \
} while (0)
while (n + 3 < minMatchLength) {
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
}
while (n < minMatchLength) {
GEAR_ITER_ONCE();
}
#undef GEAR_ITER_ONCE
}
/** ZSTD_ldm_gear_feed():
*
* Registers in the splits array all the split points found in the first
* size bytes following the data pointer. This function terminates when
* either all the data has been processed or LDM_BATCH_SIZE splits are
* present in the splits array.
*
* Precondition: The splits array must not be full.
* Returns: The number of bytes processed. */
static size_t ZSTD_ldm_gear_feed(ldmRollingHashState_t* state,
BYTE const* data, size_t size,
size_t* splits, unsigned* numSplits)
{
size_t n;
U64 hash, mask;
hash = state->rolling;
mask = state->stopMask;
n = 0;
#define GEAR_ITER_ONCE() do { \
hash = (hash << 1) + ZSTD_ldm_gearTab[data[n] & 0xff]; \
n += 1; \
if (UNLIKELY((hash & mask) == 0)) { \
splits[*numSplits] = n; \
*numSplits += 1; \
if (*numSplits == LDM_BATCH_SIZE) \
goto done; \
} \
} while (0)
while (n + 3 < size) {
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
GEAR_ITER_ONCE();
}
while (n < size) {
GEAR_ITER_ONCE();
}
#undef GEAR_ITER_ONCE
done:
state->rolling = hash;
return n;
}
void ZSTD_ldm_adjustParameters(ldmParams_t* params,
const ZSTD_compressionParameters* cParams)
{
params->windowLog = cParams->windowLog;
ZSTD_STATIC_ASSERT(LDM_BUCKET_SIZE_LOG <= ZSTD_LDM_BUCKETSIZELOG_MAX);
DEBUGLOG(4, "ZSTD_ldm_adjustParameters");
if (params->hashRateLog == 0) {
if (params->hashLog > 0) {
/* if params->hashLog is set, derive hashRateLog from it */
assert(params->hashLog <= ZSTD_HASHLOG_MAX);
if (params->windowLog > params->hashLog) {
params->hashRateLog = params->windowLog - params->hashLog;
}
} else {
assert(1 <= (int)cParams->strategy && (int)cParams->strategy <= 9);
/* mapping from [fast, rate7] to [btultra2, rate4] */
params->hashRateLog = 7 - (cParams->strategy/3);
}
}
if (params->hashLog == 0) {
params->hashLog = BOUNDED(ZSTD_HASHLOG_MIN, params->windowLog - params->hashRateLog, ZSTD_HASHLOG_MAX);
}
if (params->minMatchLength == 0) {
params->minMatchLength = LDM_MIN_MATCH_LENGTH;
if (cParams->strategy >= ZSTD_btultra)
params->minMatchLength /= 2;
}
if (params->bucketSizeLog==0) {
assert(1 <= (int)cParams->strategy && (int)cParams->strategy <= 9);
params->bucketSizeLog = BOUNDED(LDM_BUCKET_SIZE_LOG, (U32)cParams->strategy, ZSTD_LDM_BUCKETSIZELOG_MAX);
}
params->bucketSizeLog = MIN(params->bucketSizeLog, params->hashLog);
}
size_t ZSTD_ldm_getTableSize(ldmParams_t params)
{
size_t const ldmHSize = ((size_t)1) << params.hashLog;
size_t const ldmBucketSizeLog = MIN(params.bucketSizeLog, params.hashLog);
size_t const ldmBucketSize = ((size_t)1) << (params.hashLog - ldmBucketSizeLog);
size_t const totalSize = ZSTD_cwksp_alloc_size(ldmBucketSize)
+ ZSTD_cwksp_alloc_size(ldmHSize * sizeof(ldmEntry_t));
return params.enableLdm == ZSTD_ps_enable ? totalSize : 0;
}
size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize)
{
return params.enableLdm == ZSTD_ps_enable ? (maxChunkSize / params.minMatchLength) : 0;
}
/** ZSTD_ldm_getBucket() :
* Returns a pointer to the start of the bucket associated with hash. */
static ldmEntry_t* ZSTD_ldm_getBucket(
const ldmState_t* ldmState, size_t hash, U32 const bucketSizeLog)
{
return ldmState->hashTable + (hash << bucketSizeLog);
}
/** ZSTD_ldm_insertEntry() :
* Insert the entry with corresponding hash into the hash table */
static void ZSTD_ldm_insertEntry(ldmState_t* ldmState,
size_t const hash, const ldmEntry_t entry,
U32 const bucketSizeLog)
{
BYTE* const pOffset = ldmState->bucketOffsets + hash;
unsigned const offset = *pOffset;
*(ZSTD_ldm_getBucket(ldmState, hash, bucketSizeLog) + offset) = entry;
*pOffset = (BYTE)((offset + 1) & ((1u << bucketSizeLog) - 1));
}
/** ZSTD_ldm_countBackwardsMatch() :
* Returns the number of bytes that match backwards before pIn and pMatch.
*
* We count only bytes where pMatch >= pBase and pIn >= pAnchor. */
static size_t ZSTD_ldm_countBackwardsMatch(
const BYTE* pIn, const BYTE* pAnchor,
const BYTE* pMatch, const BYTE* pMatchBase)
{
size_t matchLength = 0;
while (pIn > pAnchor && pMatch > pMatchBase && pIn[-1] == pMatch[-1]) {
pIn--;
pMatch--;
matchLength++;
}
return matchLength;
}
/** ZSTD_ldm_countBackwardsMatch_2segments() :
* Returns the number of bytes that match backwards from pMatch,
* even with the backwards match spanning 2 different segments.
*
* On reaching `pMatchBase`, start counting from mEnd */
static size_t ZSTD_ldm_countBackwardsMatch_2segments(
const BYTE* pIn, const BYTE* pAnchor,
const BYTE* pMatch, const BYTE* pMatchBase,
const BYTE* pExtDictStart, const BYTE* pExtDictEnd)
{
size_t matchLength = ZSTD_ldm_countBackwardsMatch(pIn, pAnchor, pMatch, pMatchBase);
if (pMatch - matchLength != pMatchBase || pMatchBase == pExtDictStart) {
/* If backwards match is entirely in the extDict or prefix, immediately return */
return matchLength;
}
DEBUGLOG(7, "ZSTD_ldm_countBackwardsMatch_2segments: found 2-parts backwards match (length in prefix==%zu)", matchLength);
matchLength += ZSTD_ldm_countBackwardsMatch(pIn - matchLength, pAnchor, pExtDictEnd, pExtDictStart);
DEBUGLOG(7, "final backwards match length = %zu", matchLength);
return matchLength;
}
/** ZSTD_ldm_fillFastTables() :
*
* Fills the relevant tables for the ZSTD_fast and ZSTD_dfast strategies.
* This is similar to ZSTD_loadDictionaryContent.
*
* The tables for the other strategies are filled within their
* block compressors. */
static size_t ZSTD_ldm_fillFastTables(ZSTD_MatchState_t* ms,
void const* end)
{
const BYTE* const iend = (const BYTE*)end;
switch(ms->cParams.strategy)
{
switch (ms->cParams.strategy) {
case ZSTD_fast:
ZSTD_fillHashTable(ms, iend, ZSTD_dtlm_fast, ZSTD_tfp_forCCtx);
ZSTD_fillHashTable(ms, end, ZSTD_dtlm_fast, ZSTD_tfp_forCCtx);
break;
case ZSTD_dfast:
#ifndef ZSTD_EXCLUDE_DFAST_BLOCK_COMPRESSOR
ZSTD_fillDoubleHashTable(ms, iend, ZSTD_dtlm_fast, ZSTD_tfp_forCCtx);
ZSTD_fillDoubleHashTable(ms, end, ZSTD_dtlm_fast, ZSTD_tfp_forCCtx);
#else
assert(0); /* shouldn't be called: cparams should've been adjusted. */
assert(0);
#endif
break;
case ZSTD_greedy:
case ZSTD_lazy:
case ZSTD_lazy2:
@@ -276,406 +236,95 @@ static size_t ZSTD_ldm_fillFastTables(ZSTD_MatchState_t* ms,
case ZSTD_btultra2:
break;
default:
assert(0); /* not possible : not a valid strategy id */
assert(0);
}
}
return 0;
void ZSTD_ldm_rust_prepareBlock(void* blockContext, const void* anchor)
{
ZSTD_rust_ldm_block_context* const context = (ZSTD_rust_ldm_block_context*)blockContext;
ZSTD_rust_ldm_limitTableUpdate(context->ms, (const BYTE*)anchor);
ZSTD_rust_ldm_fillFastTables(context->ms, (const BYTE*)anchor);
}
size_t ZSTD_ldm_rust_compressLiterals(
void* blockContext, void* seqStore, U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize)
{
ZSTD_rust_ldm_block_context* const context = (ZSTD_rust_ldm_block_context*)blockContext;
return context->blockCompressor(context->ms, (SeqStore_t*)seqStore, rep, src, srcSize);
}
void ZSTD_ldm_rust_storeSeq(
void* seqStore, size_t litLength, const void* literals, const void* litLimit,
U32 offBase, size_t matchLength)
{
ZSTD_storeSeq((SeqStore_t*)seqStore, litLength, (const BYTE*)literals,
(const BYTE*)litLimit, offBase, matchLength);
}
void ZSTD_ldm_rust_setLdmSeqStore(void* blockContext, const void* rawSeqStore)
{
ZSTD_rust_ldm_block_context* const context = (ZSTD_rust_ldm_block_context*)blockContext;
context->ms->ldmSeqStore = (const RawSeqStore_t*)rawSeqStore;
}
void ZSTD_ldm_adjustParameters(ldmParams_t* params,
const ZSTD_compressionParameters* cParams)
{
ZSTD_rust_ldm_adjustParameters(
params, cParams->windowLog, (int)cParams->strategy,
ZSTD_HASHLOG_MAX, ZSTD_LDM_BUCKETSIZELOG_MAX, (int)ZSTD_btultra);
}
size_t ZSTD_ldm_getTableSize(ldmParams_t params)
{
#if ZSTD_ADDRESS_SANITIZER && !defined(ZSTD_ASAN_DONT_POISON_WORKSPACE)
size_t const redzoneSize = ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#else
size_t const redzoneSize = 0;
#endif
return ZSTD_rust_ldm_getTableSize(
&params, params.enableLdm == ZSTD_ps_enable, redzoneSize);
}
size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize)
{
return ZSTD_rust_ldm_getMaxNbSeq(
&params, params.enableLdm == ZSTD_ps_enable, maxChunkSize);
}
void ZSTD_ldm_fillHashTable(
ldmState_t* ldmState, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params)
ldmState_t* ldmState, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params)
{
U32 const minMatchLength = params->minMatchLength;
U32 const bucketSizeLog = params->bucketSizeLog;
U32 const hBits = params->hashLog - bucketSizeLog;
BYTE const* const base = ldmState->window.base;
BYTE const* const istart = ip;
ldmRollingHashState_t hashState;
size_t* const splits = ldmState->splitIndices;
unsigned numSplits;
DEBUGLOG(5, "ZSTD_ldm_fillHashTable");
ZSTD_ldm_gear_init(&hashState, params);
while (ip < iend) {
size_t hashed;
unsigned n;
numSplits = 0;
hashed = ZSTD_ldm_gear_feed(&hashState, ip, (size_t)(iend - ip), splits, &numSplits);
for (n = 0; n < numSplits; n++) {
if (ip + splits[n] >= istart + minMatchLength) {
BYTE const* const split = ip + splits[n] - minMatchLength;
U64 const xxhash = XXH64(split, minMatchLength, 0);
U32 const hash = (U32)(xxhash & (((U32)1 << hBits) - 1));
ldmEntry_t entry;
entry.offset = (U32)(split - base);
entry.checksum = (U32)(xxhash >> 32);
ZSTD_ldm_insertEntry(ldmState, hash, entry, params->bucketSizeLog);
}
}
ip += hashed;
}
}
/** ZSTD_ldm_limitTableUpdate() :
*
* Sets cctx->nextToUpdate to a position corresponding closer to anchor
* if it is far way
* (after a long match, only update tables a limited amount). */
static void ZSTD_ldm_limitTableUpdate(ZSTD_MatchState_t* ms, const BYTE* anchor)
{
U32 const curr = (U32)(anchor - ms->window.base);
if (curr > ms->nextToUpdate + 1024) {
ms->nextToUpdate =
curr - MIN(512, curr - ms->nextToUpdate - 1024);
}
}
static
ZSTD_ALLOW_POINTER_OVERFLOW_ATTR
size_t ZSTD_ldm_generateSequences_internal(
ldmState_t* ldmState, RawSeqStore_t* rawSeqStore,
ldmParams_t const* params, void const* src, size_t srcSize)
{
/* LDM parameters */
int const extDict = ZSTD_window_hasExtDict(ldmState->window);
U32 const minMatchLength = params->minMatchLength;
U32 const entsPerBucket = 1U << params->bucketSizeLog;
U32 const hBits = params->hashLog - params->bucketSizeLog;
/* Prefix and extDict parameters */
U32 const dictLimit = ldmState->window.dictLimit;
U32 const lowestIndex = extDict ? ldmState->window.lowLimit : dictLimit;
BYTE const* const base = ldmState->window.base;
BYTE const* const dictBase = extDict ? ldmState->window.dictBase : NULL;
BYTE const* const dictStart = extDict ? dictBase + lowestIndex : NULL;
BYTE const* const dictEnd = extDict ? dictBase + dictLimit : NULL;
BYTE const* const lowPrefixPtr = base + dictLimit;
/* Input bounds */
BYTE const* const istart = (BYTE const*)src;
BYTE const* const iend = istart + srcSize;
BYTE const* const ilimit = iend - HASH_READ_SIZE;
/* Input positions */
BYTE const* anchor = istart;
BYTE const* ip = istart;
/* Rolling hash state */
ldmRollingHashState_t hashState;
/* Arrays for staged-processing */
size_t* const splits = ldmState->splitIndices;
ldmMatchCandidate_t* const candidates = ldmState->matchCandidates;
unsigned numSplits;
if (srcSize < minMatchLength)
return iend - anchor;
/* Initialize the rolling hash state with the first minMatchLength bytes */
ZSTD_ldm_gear_init(&hashState, params);
ZSTD_ldm_gear_reset(&hashState, ip, minMatchLength);
ip += minMatchLength;
while (ip < ilimit) {
size_t hashed;
unsigned n;
numSplits = 0;
hashed = ZSTD_ldm_gear_feed(&hashState, ip, ilimit - ip,
splits, &numSplits);
for (n = 0; n < numSplits; n++) {
BYTE const* const split = ip + splits[n] - minMatchLength;
U64 const xxhash = XXH64(split, minMatchLength, 0);
U32 const hash = (U32)(xxhash & (((U32)1 << hBits) - 1));
candidates[n].split = split;
candidates[n].hash = hash;
candidates[n].checksum = (U32)(xxhash >> 32);
candidates[n].bucket = ZSTD_ldm_getBucket(ldmState, hash, params->bucketSizeLog);
PREFETCH_L1(candidates[n].bucket);
}
for (n = 0; n < numSplits; n++) {
size_t forwardMatchLength = 0, backwardMatchLength = 0,
bestMatchLength = 0, mLength;
U32 offset;
BYTE const* const split = candidates[n].split;
U32 const checksum = candidates[n].checksum;
U32 const hash = candidates[n].hash;
ldmEntry_t* const bucket = candidates[n].bucket;
ldmEntry_t const* cur;
ldmEntry_t const* bestEntry = NULL;
ldmEntry_t newEntry;
newEntry.offset = (U32)(split - base);
newEntry.checksum = checksum;
/* If a split point would generate a sequence overlapping with
* the previous one, we merely register it in the hash table and
* move on */
if (split < anchor) {
ZSTD_ldm_insertEntry(ldmState, hash, newEntry, params->bucketSizeLog);
continue;
}
for (cur = bucket; cur < bucket + entsPerBucket; cur++) {
size_t curForwardMatchLength, curBackwardMatchLength,
curTotalMatchLength;
if (cur->checksum != checksum || cur->offset <= lowestIndex) {
continue;
}
if (extDict) {
BYTE const* const curMatchBase =
cur->offset < dictLimit ? dictBase : base;
BYTE const* const pMatch = curMatchBase + cur->offset;
BYTE const* const matchEnd =
cur->offset < dictLimit ? dictEnd : iend;
BYTE const* const lowMatchPtr =
cur->offset < dictLimit ? dictStart : lowPrefixPtr;
curForwardMatchLength =
ZSTD_count_2segments(split, pMatch, iend, matchEnd, lowPrefixPtr);
if (curForwardMatchLength < minMatchLength) {
continue;
}
curBackwardMatchLength = ZSTD_ldm_countBackwardsMatch_2segments(
split, anchor, pMatch, lowMatchPtr, dictStart, dictEnd);
} else { /* !extDict */
BYTE const* const pMatch = base + cur->offset;
curForwardMatchLength = ZSTD_count(split, pMatch, iend);
if (curForwardMatchLength < minMatchLength) {
continue;
}
curBackwardMatchLength =
ZSTD_ldm_countBackwardsMatch(split, anchor, pMatch, lowPrefixPtr);
}
curTotalMatchLength = curForwardMatchLength + curBackwardMatchLength;
if (curTotalMatchLength > bestMatchLength) {
bestMatchLength = curTotalMatchLength;
forwardMatchLength = curForwardMatchLength;
backwardMatchLength = curBackwardMatchLength;
bestEntry = cur;
}
}
/* No match found -- insert an entry into the hash table
* and process the next candidate match */
if (bestEntry == NULL) {
ZSTD_ldm_insertEntry(ldmState, hash, newEntry, params->bucketSizeLog);
continue;
}
/* Match found */
offset = (U32)(split - base) - bestEntry->offset;
mLength = forwardMatchLength + backwardMatchLength;
{
rawSeq* const seq = rawSeqStore->seq + rawSeqStore->size;
/* Out of sequence storage */
if (rawSeqStore->size == rawSeqStore->capacity)
return ERROR(dstSize_tooSmall);
seq->litLength = (U32)(split - backwardMatchLength - anchor);
seq->matchLength = (U32)mLength;
seq->offset = offset;
rawSeqStore->size++;
}
/* Insert the current entry into the hash table --- it must be
* done after the previous block to avoid clobbering bestEntry */
ZSTD_ldm_insertEntry(ldmState, hash, newEntry, params->bucketSizeLog);
anchor = split + forwardMatchLength;
/* If we find a match that ends after the data that we've hashed
* then we have a repeating, overlapping, pattern. E.g. all zeros.
* If one repetition of the pattern matches our `stopMask` then all
* repetitions will. We don't need to insert them all into out table,
* only the first one. So skip over overlapping matches.
* This is a major speed boost (20x) for compressing a single byte
* repeated, when that byte ends up in the table.
*/
if (anchor > ip + hashed) {
ZSTD_ldm_gear_reset(&hashState, anchor - minMatchLength, minMatchLength);
/* Continue the outer loop at anchor (ip + hashed == anchor). */
ip = anchor - hashed;
break;
}
}
ip += hashed;
}
return iend - anchor;
}
/*! ZSTD_ldm_reduceTable() :
* reduce table indexes by `reducerValue` */
static void ZSTD_ldm_reduceTable(ldmEntry_t* const table, U32 const size,
U32 const reducerValue)
{
U32 u;
for (u = 0; u < size; u++) {
if (table[u].offset < reducerValue) table[u].offset = 0;
else table[u].offset -= reducerValue;
}
ZSTD_rust_ldm_fillHashTable(
ldmState->hashTable, ldmState->bucketOffsets, ldmState->window.base,
ip, iend, params);
}
size_t ZSTD_ldm_generateSequences(
ldmState_t* ldmState, RawSeqStore_t* sequences,
ldmParams_t const* params, void const* src, size_t srcSize)
{
U32 const maxDist = 1U << params->windowLog;
BYTE const* const istart = (BYTE const*)src;
BYTE const* const iend = istart + srcSize;
size_t const kMaxChunkSize = 1 << 20;
size_t const nbChunks = (srcSize / kMaxChunkSize) + ((srcSize % kMaxChunkSize) != 0);
size_t chunk;
size_t leftoverSize = 0;
assert(ZSTD_CHUNKSIZE_MAX >= kMaxChunkSize);
/* Check that ZSTD_window_update() has been called for this chunk prior
* to passing it to this function.
*/
assert(ldmState->window.nextSrc >= (BYTE const*)src + srcSize);
/* The input could be very large (in zstdmt), so it must be broken up into
* chunks to enforce the maximum distance and handle overflow correction.
*/
assert(ldmState->window.nextSrc >= (const BYTE*)src + srcSize);
assert(sequences->pos <= sequences->size);
assert(sequences->size <= sequences->capacity);
for (chunk = 0; chunk < nbChunks && sequences->size < sequences->capacity; ++chunk) {
BYTE const* const chunkStart = istart + chunk * kMaxChunkSize;
size_t const remaining = (size_t)(iend - chunkStart);
BYTE const *const chunkEnd =
(remaining < kMaxChunkSize) ? iend : chunkStart + kMaxChunkSize;
size_t const chunkSize = chunkEnd - chunkStart;
size_t newLeftoverSize;
size_t const prevSize = sequences->size;
assert(chunkStart < iend);
/* 1. Perform overflow correction if necessary. */
if (ZSTD_window_needOverflowCorrection(ldmState->window, 0, maxDist, ldmState->loadedDictEnd, chunkStart, chunkEnd)) {
U32 const ldmHSize = 1U << params->hashLog;
U32 const correction = ZSTD_window_correctOverflow(
&ldmState->window, /* cycleLog */ 0, maxDist, chunkStart);
ZSTD_ldm_reduceTable(ldmState->hashTable, ldmHSize, correction);
/* invalidate dictionaries on overflow correction */
ldmState->loadedDictEnd = 0;
}
/* 2. We enforce the maximum offset allowed.
*
* kMaxChunkSize should be small enough that we don't lose too much of
* the window through early invalidation.
* TODO: * Test the chunk size.
* * Try invalidation after the sequence generation and test the
* offset against maxDist directly.
*
* NOTE: Because of dictionaries + sequence splitting we MUST make sure
* that any offset used is valid at the END of the sequence, since it may
* be split into two sequences. This condition holds when using
* ZSTD_window_enforceMaxDist(), but if we move to checking offsets
* against maxDist directly, we'll have to carefully handle that case.
*/
ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, &ldmState->loadedDictEnd, NULL);
/* 3. Generate the sequences for the chunk, and get newLeftoverSize. */
newLeftoverSize = ZSTD_ldm_generateSequences_internal(
ldmState, sequences, params, chunkStart, chunkSize);
if (ZSTD_isError(newLeftoverSize))
return newLeftoverSize;
/* 4. We add the leftover literals from previous iterations to the first
* newly generated sequence, or add the `newLeftoverSize` if none are
* generated.
*/
/* Prepend the leftover literals from the last call */
if (prevSize < sequences->size) {
sequences->seq[prevSize].litLength += (U32)leftoverSize;
leftoverSize = newLeftoverSize;
} else {
assert(newLeftoverSize == chunkSize);
leftoverSize += chunkSize;
}
}
return 0;
return ZSTD_rust_ldm_generateSequences(
ldmState->hashTable, ldmState->bucketOffsets, &ldmState->window,
&ldmState->loadedDictEnd, sequences, params, src, srcSize,
ZSTD_WINDOW_OVERFLOW_CORRECT_FREQUENTLY);
}
void
ZSTD_ldm_skipSequences(RawSeqStore_t* rawSeqStore, size_t srcSize, U32 const minMatch)
void ZSTD_ldm_skipSequences(RawSeqStore_t* rawSeqStore, size_t srcSize,
U32 const minMatch)
{
while (srcSize > 0 && rawSeqStore->pos < rawSeqStore->size) {
rawSeq* seq = rawSeqStore->seq + rawSeqStore->pos;
if (srcSize <= seq->litLength) {
/* Skip past srcSize literals */
seq->litLength -= (U32)srcSize;
return;
}
srcSize -= seq->litLength;
seq->litLength = 0;
if (srcSize < seq->matchLength) {
/* Skip past the first srcSize of the match */
seq->matchLength -= (U32)srcSize;
if (seq->matchLength < minMatch) {
/* The match is too short, omit it */
if (rawSeqStore->pos + 1 < rawSeqStore->size) {
seq[1].litLength += seq[0].matchLength;
}
rawSeqStore->pos++;
}
return;
}
srcSize -= seq->matchLength;
seq->matchLength = 0;
rawSeqStore->pos++;
}
ZSTD_rust_ldm_skipSequences(rawSeqStore, srcSize, minMatch);
}
/**
* If the sequence length is longer than remaining then the sequence is split
* between this block and the next.
*
* Returns the current sequence to handle, or if the rest of the block should
* be literals, it returns a sequence with offset == 0.
*/
static rawSeq maybeSplitSequence(RawSeqStore_t* rawSeqStore,
U32 const remaining, U32 const minMatch)
void ZSTD_ldm_skipRawSeqStoreBytes(RawSeqStore_t* rawSeqStore, size_t nbBytes)
{
rawSeq sequence = rawSeqStore->seq[rawSeqStore->pos];
assert(sequence.offset > 0);
/* Likely: No partial sequence */
if (remaining >= sequence.litLength + sequence.matchLength) {
rawSeqStore->pos++;
return sequence;
}
/* Cut the sequence short (offset == 0 ==> rest is literals). */
if (remaining <= sequence.litLength) {
sequence.offset = 0;
} else if (remaining < sequence.litLength + sequence.matchLength) {
sequence.matchLength = remaining - sequence.litLength;
if (sequence.matchLength < minMatch) {
sequence.offset = 0;
}
}
/* Skip past `remaining` bytes for the future sequences. */
ZSTD_ldm_skipSequences(rawSeqStore, remaining, minMatch);
return sequence;
}
void ZSTD_ldm_skipRawSeqStoreBytes(RawSeqStore_t* rawSeqStore, size_t nbBytes) {
U32 currPos = (U32)(rawSeqStore->posInSequence + nbBytes);
while (currPos && rawSeqStore->pos < rawSeqStore->size) {
rawSeq currSeq = rawSeqStore->seq[rawSeqStore->pos];
if (currPos >= currSeq.litLength + currSeq.matchLength) {
currPos -= currSeq.litLength + currSeq.matchLength;
rawSeqStore->pos++;
} else {
rawSeqStore->posInSequence = currPos;
break;
}
}
if (currPos == 0 || rawSeqStore->pos == rawSeqStore->size) {
rawSeqStore->posInSequence = 0;
}
ZSTD_rust_ldm_skipRawSeqStoreBytes(rawSeqStore, nbBytes);
}
size_t ZSTD_ldm_blockCompress(RawSeqStore_t* rawSeqStore,
@@ -683,63 +332,12 @@ size_t ZSTD_ldm_blockCompress(RawSeqStore_t* rawSeqStore,
ZSTD_ParamSwitch_e useRowMatchFinder,
void const* src, size_t srcSize)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
unsigned const minMatch = cParams->minMatch;
ZSTD_BlockCompressor_f const blockCompressor =
ZSTD_selectBlockCompressor(cParams->strategy, useRowMatchFinder, ZSTD_matchState_dictMode(ms));
/* Input bounds */
BYTE const* const istart = (BYTE const*)src;
BYTE const* const iend = istart + srcSize;
/* Input positions */
BYTE const* ip = istart;
DEBUGLOG(5, "ZSTD_ldm_blockCompress: srcSize=%zu", srcSize);
/* If using opt parser, use LDMs only as candidates rather than always accepting them */
if (cParams->strategy >= ZSTD_btopt) {
size_t lastLLSize;
ms->ldmSeqStore = rawSeqStore;
lastLLSize = blockCompressor(ms, seqStore, rep, src, srcSize);
ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore, srcSize);
return lastLLSize;
}
assert(rawSeqStore->pos <= rawSeqStore->size);
assert(rawSeqStore->size <= rawSeqStore->capacity);
/* Loop through each sequence and apply the block compressor to the literals */
while (rawSeqStore->pos < rawSeqStore->size && ip < iend) {
/* maybeSplitSequence updates rawSeqStore->pos */
rawSeq const sequence = maybeSplitSequence(rawSeqStore,
(U32)(iend - ip), minMatch);
/* End signal */
if (sequence.offset == 0)
break;
assert(ip + sequence.litLength + sequence.matchLength <= iend);
/* Fill tables for block compressor */
ZSTD_ldm_limitTableUpdate(ms, ip);
ZSTD_ldm_fillFastTables(ms, ip);
/* Run the block compressor */
DEBUGLOG(5, "pos %u : calling block compressor on segment of size %u", (unsigned)(ip-istart), sequence.litLength);
{
int i;
size_t const newLitLength =
blockCompressor(ms, seqStore, rep, ip, sequence.litLength);
ip += sequence.litLength;
/* Update the repcodes */
for (i = ZSTD_REP_NUM - 1; i > 0; i--)
rep[i] = rep[i-1];
rep[0] = sequence.offset;
/* Store the sequence */
ZSTD_storeSeq(seqStore, newLitLength, ip - newLitLength, iend,
OFFSET_TO_OFFBASE(sequence.offset),
sequence.matchLength);
ip += sequence.matchLength;
}
}
/* Fill the tables for the block compressor */
ZSTD_ldm_limitTableUpdate(ms, ip);
ZSTD_ldm_fillFastTables(ms, ip);
/* Compress the last literals */
return blockCompressor(ms, seqStore, rep, ip, iend - ip);
ZSTD_rust_ldm_block_context context;
context.ms = ms;
context.blockCompressor = ZSTD_selectBlockCompressor(
ms->cParams.strategy, useRowMatchFinder, ZSTD_matchState_dictMode(ms));
assert(context.blockCompressor != NULL);
return ZSTD_rust_ldm_blockCompress(
rawSeqStore, &context, seqStore, rep, src, srcSize,
ms->cParams.minMatch, ms->cParams.strategy >= ZSTD_btopt);
}
lib/decompress/zstd_decompress.c
+395 -2323
View File
@@ -1,925 +1,336 @@
/*
* Copyright (c) Meta Platforms, Inc. and affiliates.
* All rights reserved.
* Decoder orchestration is implemented in rust/src/zstd_decompress.rs.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
* `ZSTD_DCtx_s` is intentionally still allocated and laid out by C: its
* optional members vary with the build configuration and the block decoder
* shares that object. This translation unit is therefore a deliberately
* narrow ABI adapter. It projects field addresses to Rust, keeps allocation
* ownership in the C allocator domain, and retains the configuration-bound
* legacy and trace leaves.
*/
#define ZSTD_STATIC_LINKING_ONLY
#include "../common/zstd_deps.h"
#include "../common/allocations.h"
#include "../common/error_private.h"
#include "../common/mem.h"
#include "../common/zstd_internal.h"
#include "zstd_decompress_internal.h"
#include "zstd_ddict.h"
/* ***************************************************************
* Tuning parameters
*****************************************************************/
/*!
* HEAPMODE :
* Select how default decompression function ZSTD_decompress() allocates its context,
* on stack (0), or into heap (1, default; requires malloc()).
* Note that functions with explicit context such as ZSTD_decompressDCtx() are unaffected.
*/
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
# include "../legacy/zstd_legacy.h"
#endif
/* Keep the three public build-time tuning knobs in the C configuration domain.
* Rust queries them through the small leaves below rather than baking a second
* set of defaults into its source. */
#ifndef ZSTD_HEAPMODE
# define ZSTD_HEAPMODE 1
#endif
/*!
* LEGACY_SUPPORT :
* if set to 1+, ZSTD_decompress() can decode older formats (v0.1+)
*/
#ifndef ZSTD_LEGACY_SUPPORT
# define ZSTD_LEGACY_SUPPORT 0
#endif
/*!
* MAXWINDOWSIZE_DEFAULT :
* maximum window size accepted by DStream __by default__.
* Frames requiring more memory will be rejected.
* It's possible to set a different limit using ZSTD_DCtx_setMaxWindowSize().
*/
#ifndef ZSTD_MAXWINDOWSIZE_DEFAULT
# define ZSTD_MAXWINDOWSIZE_DEFAULT (((U32)1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT) + 1)
#endif
/*!
* NO_FORWARD_PROGRESS_MAX :
* maximum allowed nb of calls to ZSTD_decompressStream()
* without any forward progress
* (defined as: no byte read from input, and no byte flushed to output)
* before triggering an error.
*/
#ifndef ZSTD_NO_FORWARD_PROGRESS_MAX
# define ZSTD_NO_FORWARD_PROGRESS_MAX 16
#endif
/* Keep this in lock-step with ZSTD_rustDctxView in zstd_decompress.rs.
* Every pointer that names a scalar is the address of that scalar; array and
* embedded-object entries name their first byte. */
typedef struct {
void* dctx;
void* llt_ptr;
void* mlt_ptr;
void* oft_ptr;
void* huf_ptr;
void* entropy;
void* workspace;
size_t workspace_size;
void* previous_dst_end;
void* prefix_start;
void* virtual_start;
void* dict_end;
void* expected;
void* f_params;
void* processed_c_size;
void* decoded_size;
void* b_type;
void* stage;
void* lit_entropy;
void* fse_entropy;
void* xxh_state;
void* header_size;
void* format;
void* force_ignore_checksum;
void* validate_checksum;
void* lit_ptr;
void* custom_mem;
void* lit_size;
void* rle_size;
void* static_size;
void* is_frame_decompression;
void* ddict_local;
void* ddict;
void* dict_id;
void* ddict_is_cold;
void* dict_uses;
void* ddict_set;
void* ref_multiple_ddicts;
void* disable_huf_asm;
void* max_block_size_param;
void* stream_stage;
void* in_buff;
void* in_buff_size;
void* in_pos;
void* max_window_size;
void* out_buff;
void* out_buff_size;
void* out_start;
void* out_end;
void* lh_size;
void* legacy_context;
void* previous_legacy_version;
void* legacy_version;
void* hostage_byte;
void* no_forward_progress;
void* out_buffer_mode;
void* expected_out_buffer;
void* lit_buffer;
void* lit_buffer_end;
void* lit_buffer_location;
void* lit_extra_buffer;
size_t lit_extra_buffer_size;
void* header_buffer;
size_t header_buffer_size;
void* oversized_duration;
void* fuzz_begin;
void* fuzz_end;
size_t dctx_size;
} ZSTD_rustDctxView;
/*-*******************************************************
* Dependencies
*********************************************************/
#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */
#include "../common/allocations.h" /* ZSTD_customMalloc, ZSTD_customCalloc, ZSTD_customFree */
#include "../common/error_private.h"
#include "../common/zstd_internal.h" /* blockProperties_t */
#include "../common/mem.h" /* low level memory routines */
#include "../common/bits.h" /* ZSTD_highbit32 */
#define FSE_STATIC_LINKING_ONLY
#include "../common/fse.h"
#include "../common/huf.h"
#include "../common/xxhash.h" /* XXH64_reset, XXH64_update, XXH64_digest, XXH64 */
#include "zstd_decompress_internal.h" /* ZSTD_DCtx */
#include "zstd_ddict.h" /* ZSTD_DDictDictContent */
#include "zstd_decompress_block.h" /* ZSTD_decompressBlock_internal */
void ZSTD_rust_dctx_view(ZSTD_DCtx* dctx, ZSTD_rustDctxView* out);
size_t ZSTD_rust_dctx_sizeof(void);
ZSTD_DCtx* ZSTD_rust_dctx_alloc(ZSTD_customMem customMem);
void ZSTD_rust_dctx_free_storage(ZSTD_DCtx* dctx, ZSTD_customMem customMem);
void ZSTD_rust_dctx_init_platform(ZSTD_DCtx* dctx);
size_t ZSTD_rust_dctx_default_max_window_size(void);
int ZSTD_rust_no_forward_progress_max(void);
int ZSTD_rust_heapmode(void);
size_t ZSTD_rust_decompress_stack(void* dst, size_t dstCapacity,
const void* src, size_t srcSize);
void* ZSTD_rust_custom_malloc(size_t size, ZSTD_customMem customMem);
void* ZSTD_rust_custom_calloc(size_t size, ZSTD_customMem customMem);
void ZSTD_rust_custom_free(void* allocation, ZSTD_customMem customMem);
ZSTD_DDict* ZSTD_rust_create_ddict(const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType,
ZSTD_customMem customMem);
void ZSTD_rust_dctx_copy_prefix(ZSTD_DCtx* dstDCtx, const ZSTD_DCtx* srcDCtx);
void ZSTD_rust_dctx_trace_begin(ZSTD_DCtx* dctx);
void ZSTD_rust_dctx_trace_end(ZSTD_DCtx* dctx, U64 uncompressedSize,
U64 compressedSize, int streaming);
unsigned ZSTD_rust_legacy_is(const void* src, size_t srcSize);
unsigned long long ZSTD_rust_legacy_get_decompressed_size(const void* src,
size_t srcSize);
size_t ZSTD_rust_legacy_find_compressed_size(const void* src, size_t srcSize);
size_t ZSTD_rust_legacy_frame_size_info(const void* src, size_t srcSize,
size_t* compressedSize,
unsigned long long* decompressedBound,
size_t* nbBlocks);
size_t ZSTD_rust_legacy_decompress(void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const void* dict, size_t dictSize);
size_t ZSTD_rust_legacy_decompress_stream(ZSTD_DCtx* dctx,
ZSTD_outBuffer* output,
ZSTD_inBuffer* input,
const void* dict, size_t dictSize);
void ZSTD_rust_legacy_free_stream(ZSTD_DCtx* dctx);
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
# include "../legacy/zstd_legacy.h"
void ZSTD_rust_dctx_view(ZSTD_DCtx* dctx, ZSTD_rustDctxView* out)
{
ZSTD_memset(out, 0, sizeof(*out));
if (dctx == NULL) return;
out->dctx = dctx;
out->llt_ptr = &dctx->LLTptr;
out->mlt_ptr = &dctx->MLTptr;
out->oft_ptr = &dctx->OFTptr;
out->huf_ptr = &dctx->HUFptr;
out->entropy = &dctx->entropy;
out->workspace = dctx->workspace;
out->workspace_size = sizeof(dctx->workspace);
out->previous_dst_end = &dctx->previousDstEnd;
out->prefix_start = &dctx->prefixStart;
out->virtual_start = &dctx->virtualStart;
out->dict_end = &dctx->dictEnd;
out->expected = &dctx->expected;
out->f_params = &dctx->fParams;
out->processed_c_size = &dctx->processedCSize;
out->decoded_size = &dctx->decodedSize;
out->b_type = &dctx->bType;
out->stage = &dctx->stage;
out->lit_entropy = &dctx->litEntropy;
out->fse_entropy = &dctx->fseEntropy;
out->xxh_state = &dctx->xxhState;
out->header_size = &dctx->headerSize;
out->format = &dctx->format;
out->force_ignore_checksum = &dctx->forceIgnoreChecksum;
out->validate_checksum = &dctx->validateChecksum;
out->lit_ptr = &dctx->litPtr;
out->custom_mem = &dctx->customMem;
out->lit_size = &dctx->litSize;
out->rle_size = &dctx->rleSize;
out->static_size = &dctx->staticSize;
out->is_frame_decompression = &dctx->isFrameDecompression;
out->ddict_local = &dctx->ddictLocal;
out->ddict = &dctx->ddict;
out->dict_id = &dctx->dictID;
out->ddict_is_cold = &dctx->ddictIsCold;
out->dict_uses = &dctx->dictUses;
out->ddict_set = &dctx->ddictSet;
out->ref_multiple_ddicts = &dctx->refMultipleDDicts;
out->disable_huf_asm = &dctx->disableHufAsm;
out->max_block_size_param = &dctx->maxBlockSizeParam;
out->stream_stage = &dctx->streamStage;
out->in_buff = &dctx->inBuff;
out->in_buff_size = &dctx->inBuffSize;
out->in_pos = &dctx->inPos;
out->max_window_size = &dctx->maxWindowSize;
out->out_buff = &dctx->outBuff;
out->out_buff_size = &dctx->outBuffSize;
out->out_start = &dctx->outStart;
out->out_end = &dctx->outEnd;
out->lh_size = &dctx->lhSize;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
out->legacy_context = &dctx->legacyContext;
out->previous_legacy_version = &dctx->previousLegacyVersion;
out->legacy_version = &dctx->legacyVersion;
#endif
/*************************************
* Multiple DDicts Hashset internals *
*************************************/
#define DDICT_HASHSET_MAX_LOAD_FACTOR_COUNT_MULT 4
#define DDICT_HASHSET_MAX_LOAD_FACTOR_SIZE_MULT 3 /* These two constants represent SIZE_MULT/COUNT_MULT load factor without using a float.
* Currently, that means a 0.75 load factor.
* So, if count * COUNT_MULT / size * SIZE_MULT != 0, then we've exceeded
* the load factor of the ddict hash set.
*/
#define DDICT_HASHSET_TABLE_BASE_SIZE 64
#define DDICT_HASHSET_RESIZE_FACTOR 2
/* Hash function to determine starting position of dict insertion within the table
* Returns an index between [0, hashSet->ddictPtrTableSize]
*/
static size_t ZSTD_DDictHashSet_getIndex(const ZSTD_DDictHashSet* hashSet, U32 dictID) {
const U64 hash = XXH64(&dictID, sizeof(U32), 0);
/* DDict ptr table size is a multiple of 2, use size - 1 as mask to get index within [0, hashSet->ddictPtrTableSize) */
return hash & (hashSet->ddictPtrTableSize - 1);
}
/* Adds DDict to a hashset without resizing it.
* If inserting a DDict with a dictID that already exists in the set, replaces the one in the set.
* Returns 0 if successful, or a zstd error code if something went wrong.
*/
static size_t ZSTD_DDictHashSet_emplaceDDict(ZSTD_DDictHashSet* hashSet, const ZSTD_DDict* ddict) {
const U32 dictID = ZSTD_getDictID_fromDDict(ddict);
size_t idx = ZSTD_DDictHashSet_getIndex(hashSet, dictID);
const size_t idxRangeMask = hashSet->ddictPtrTableSize - 1;
RETURN_ERROR_IF(hashSet->ddictPtrCount == hashSet->ddictPtrTableSize, GENERIC, "Hash set is full!");
DEBUGLOG(4, "Hashed index: for dictID: %u is %zu", dictID, idx);
while (hashSet->ddictPtrTable[idx] != NULL) {
/* Replace existing ddict if inserting ddict with same dictID */
if (ZSTD_getDictID_fromDDict(hashSet->ddictPtrTable[idx]) == dictID) {
DEBUGLOG(4, "DictID already exists, replacing rather than adding");
hashSet->ddictPtrTable[idx] = ddict;
return 0;
}
idx &= idxRangeMask;
idx++;
}
DEBUGLOG(4, "Final idx after probing for dictID %u is: %zu", dictID, idx);
hashSet->ddictPtrTable[idx] = ddict;
hashSet->ddictPtrCount++;
return 0;
}
/* Expands hash table by factor of DDICT_HASHSET_RESIZE_FACTOR and
* rehashes all values, allocates new table, frees old table.
* Returns 0 on success, otherwise a zstd error code.
*/
static size_t ZSTD_DDictHashSet_expand(ZSTD_DDictHashSet* hashSet, ZSTD_customMem customMem) {
size_t newTableSize = hashSet->ddictPtrTableSize * DDICT_HASHSET_RESIZE_FACTOR;
const ZSTD_DDict** newTable = (const ZSTD_DDict**)ZSTD_customCalloc(sizeof(ZSTD_DDict*) * newTableSize, customMem);
const ZSTD_DDict** oldTable = hashSet->ddictPtrTable;
size_t oldTableSize = hashSet->ddictPtrTableSize;
size_t i;
DEBUGLOG(4, "Expanding DDict hash table! Old size: %zu new size: %zu", oldTableSize, newTableSize);
RETURN_ERROR_IF(!newTable, memory_allocation, "Expanded hashset allocation failed!");
hashSet->ddictPtrTable = newTable;
hashSet->ddictPtrTableSize = newTableSize;
hashSet->ddictPtrCount = 0;
for (i = 0; i < oldTableSize; ++i) {
if (oldTable[i] != NULL) {
FORWARD_IF_ERROR(ZSTD_DDictHashSet_emplaceDDict(hashSet, oldTable[i]), "");
}
}
ZSTD_customFree((void*)oldTable, customMem);
DEBUGLOG(4, "Finished re-hash");
return 0;
}
/* Fetches a DDict with the given dictID
* Returns the ZSTD_DDict* with the requested dictID. If it doesn't exist, then returns NULL.
*/
static const ZSTD_DDict* ZSTD_DDictHashSet_getDDict(ZSTD_DDictHashSet* hashSet, U32 dictID) {
size_t idx = ZSTD_DDictHashSet_getIndex(hashSet, dictID);
const size_t idxRangeMask = hashSet->ddictPtrTableSize - 1;
DEBUGLOG(4, "Hashed index: for dictID: %u is %zu", dictID, idx);
for (;;) {
size_t currDictID = ZSTD_getDictID_fromDDict(hashSet->ddictPtrTable[idx]);
if (currDictID == dictID || currDictID == 0) {
/* currDictID == 0 implies a NULL ddict entry */
break;
} else {
idx &= idxRangeMask; /* Goes to start of table when we reach the end */
idx++;
}
}
DEBUGLOG(4, "Final idx after probing for dictID %u is: %zu", dictID, idx);
return hashSet->ddictPtrTable[idx];
}
/* Allocates space for and returns a ddict hash set
* The hash set's ZSTD_DDict* table has all values automatically set to NULL to begin with.
* Returns NULL if allocation failed.
*/
static ZSTD_DDictHashSet* ZSTD_createDDictHashSet(ZSTD_customMem customMem) {
ZSTD_DDictHashSet* ret = (ZSTD_DDictHashSet*)ZSTD_customMalloc(sizeof(ZSTD_DDictHashSet), customMem);
DEBUGLOG(4, "Allocating new hash set");
if (!ret)
return NULL;
ret->ddictPtrTable = (const ZSTD_DDict**)ZSTD_customCalloc(DDICT_HASHSET_TABLE_BASE_SIZE * sizeof(ZSTD_DDict*), customMem);
if (!ret->ddictPtrTable) {
ZSTD_customFree(ret, customMem);
return NULL;
}
ret->ddictPtrTableSize = DDICT_HASHSET_TABLE_BASE_SIZE;
ret->ddictPtrCount = 0;
return ret;
}
/* Frees the table of ZSTD_DDict* within a hashset, then frees the hashset itself.
* Note: The ZSTD_DDict* within the table are NOT freed.
*/
static void ZSTD_freeDDictHashSet(ZSTD_DDictHashSet* hashSet, ZSTD_customMem customMem) {
DEBUGLOG(4, "Freeing ddict hash set");
if (hashSet && hashSet->ddictPtrTable) {
ZSTD_customFree((void*)hashSet->ddictPtrTable, customMem);
}
if (hashSet) {
ZSTD_customFree(hashSet, customMem);
}
}
/* Public function: Adds a DDict into the ZSTD_DDictHashSet, possibly triggering a resize of the hash set.
* Returns 0 on success, or a ZSTD error.
*/
static size_t ZSTD_DDictHashSet_addDDict(ZSTD_DDictHashSet* hashSet, const ZSTD_DDict* ddict, ZSTD_customMem customMem) {
DEBUGLOG(4, "Adding dict ID: %u to hashset with - Count: %zu Tablesize: %zu", ZSTD_getDictID_fromDDict(ddict), hashSet->ddictPtrCount, hashSet->ddictPtrTableSize);
if (hashSet->ddictPtrCount * DDICT_HASHSET_MAX_LOAD_FACTOR_COUNT_MULT / hashSet->ddictPtrTableSize * DDICT_HASHSET_MAX_LOAD_FACTOR_SIZE_MULT != 0) {
FORWARD_IF_ERROR(ZSTD_DDictHashSet_expand(hashSet, customMem), "");
}
FORWARD_IF_ERROR(ZSTD_DDictHashSet_emplaceDDict(hashSet, ddict), "");
return 0;
}
/*-*************************************************************
* Context management
***************************************************************/
size_t ZSTD_sizeof_DCtx (const ZSTD_DCtx* dctx)
{
if (dctx==NULL) return 0; /* support sizeof NULL */
return sizeof(*dctx)
+ ZSTD_sizeof_DDict(dctx->ddictLocal)
+ dctx->inBuffSize + dctx->outBuffSize;
}
size_t ZSTD_estimateDCtxSize(void) { return sizeof(ZSTD_DCtx); }
static size_t ZSTD_startingInputLength(ZSTD_format_e format)
{
size_t const startingInputLength = ZSTD_FRAMEHEADERSIZE_PREFIX(format);
/* only supports formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless */
assert( (format == ZSTD_f_zstd1) || (format == ZSTD_f_zstd1_magicless) );
return startingInputLength;
}
static void ZSTD_DCtx_resetParameters(ZSTD_DCtx* dctx)
{
assert(dctx->streamStage == zdss_init);
dctx->format = ZSTD_f_zstd1;
dctx->maxWindowSize = ZSTD_MAXWINDOWSIZE_DEFAULT;
dctx->outBufferMode = ZSTD_bm_buffered;
dctx->forceIgnoreChecksum = ZSTD_d_validateChecksum;
dctx->refMultipleDDicts = ZSTD_rmd_refSingleDDict;
dctx->disableHufAsm = 0;
dctx->maxBlockSizeParam = 0;
}
static void ZSTD_initDCtx_internal(ZSTD_DCtx* dctx)
{
dctx->staticSize = 0;
dctx->ddict = NULL;
dctx->ddictLocal = NULL;
dctx->dictEnd = NULL;
dctx->ddictIsCold = 0;
dctx->dictUses = ZSTD_dont_use;
dctx->inBuff = NULL;
dctx->inBuffSize = 0;
dctx->outBuffSize = 0;
dctx->streamStage = zdss_init;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
dctx->legacyContext = NULL;
dctx->previousLegacyVersion = 0;
out->hostage_byte = &dctx->hostageByte;
out->no_forward_progress = &dctx->noForwardProgress;
out->out_buffer_mode = &dctx->outBufferMode;
out->expected_out_buffer = &dctx->expectedOutBuffer;
out->lit_buffer = &dctx->litBuffer;
out->lit_buffer_end = &dctx->litBufferEnd;
out->lit_buffer_location = &dctx->litBufferLocation;
out->lit_extra_buffer = dctx->litExtraBuffer;
out->lit_extra_buffer_size = sizeof(dctx->litExtraBuffer);
out->header_buffer = dctx->headerBuffer;
out->header_buffer_size = sizeof(dctx->headerBuffer);
out->oversized_duration = &dctx->oversizedDuration;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
out->fuzz_begin = &dctx->dictContentBeginForFuzzing;
out->fuzz_end = &dctx->dictContentEndForFuzzing;
#endif
dctx->noForwardProgress = 0;
dctx->oversizedDuration = 0;
dctx->isFrameDecompression = 1;
out->dctx_size = sizeof(*dctx);
}
size_t ZSTD_rust_dctx_sizeof(void)
{
return sizeof(ZSTD_DCtx);
}
ZSTD_DCtx* ZSTD_rust_dctx_alloc(ZSTD_customMem customMem)
{
if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL;
return (ZSTD_DCtx*)ZSTD_customMalloc(sizeof(ZSTD_DCtx), customMem);
}
void ZSTD_rust_dctx_free_storage(ZSTD_DCtx* dctx, ZSTD_customMem customMem)
{
ZSTD_customFree(dctx, customMem);
}
void ZSTD_rust_dctx_init_platform(ZSTD_DCtx* dctx)
{
#if DYNAMIC_BMI2
dctx->bmi2 = ZSTD_cpuSupportsBmi2();
#endif
dctx->ddictSet = NULL;
ZSTD_DCtx_resetParameters(dctx);
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
dctx->dictContentEndForFuzzing = NULL;
#else
(void)dctx;
#endif
}
ZSTD_DCtx* ZSTD_initStaticDCtx(void *workspace, size_t workspaceSize)
size_t ZSTD_rust_dctx_default_max_window_size(void)
{
ZSTD_DCtx* const dctx = (ZSTD_DCtx*) workspace;
if ((size_t)workspace & 7) return NULL; /* 8-aligned */
if (workspaceSize < sizeof(ZSTD_DCtx)) return NULL; /* minimum size */
ZSTD_initDCtx_internal(dctx);
dctx->staticSize = workspaceSize;
dctx->inBuff = (char*)(dctx+1);
return dctx;
return ZSTD_MAXWINDOWSIZE_DEFAULT;
}
static ZSTD_DCtx* ZSTD_createDCtx_internal(ZSTD_customMem customMem) {
if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL;
{ ZSTD_DCtx* const dctx = (ZSTD_DCtx*)ZSTD_customMalloc(sizeof(*dctx), customMem);
if (!dctx) return NULL;
dctx->customMem = customMem;
ZSTD_initDCtx_internal(dctx);
return dctx;
}
int ZSTD_rust_no_forward_progress_max(void)
{
return ZSTD_NO_FORWARD_PROGRESS_MAX;
}
ZSTD_DCtx* ZSTD_createDCtx_advanced(ZSTD_customMem customMem)
int ZSTD_rust_heapmode(void)
{
return ZSTD_createDCtx_internal(customMem);
return ZSTD_HEAPMODE;
}
ZSTD_DCtx* ZSTD_createDCtx(void)
size_t ZSTD_rust_decompress_stack(void* dst, size_t dstCapacity,
const void* src, size_t srcSize)
{
DEBUGLOG(3, "ZSTD_createDCtx");
return ZSTD_createDCtx_internal(ZSTD_defaultCMem);
}
static void ZSTD_clearDict(ZSTD_DCtx* dctx)
{
ZSTD_freeDDict(dctx->ddictLocal);
dctx->ddictLocal = NULL;
dctx->ddict = NULL;
dctx->dictUses = ZSTD_dont_use;
}
size_t ZSTD_freeDCtx(ZSTD_DCtx* dctx)
{
if (dctx==NULL) return 0; /* support free on NULL */
RETURN_ERROR_IF(dctx->staticSize, memory_allocation, "not compatible with static DCtx");
{ ZSTD_customMem const cMem = dctx->customMem;
ZSTD_clearDict(dctx);
ZSTD_customFree(dctx->inBuff, cMem);
dctx->inBuff = NULL;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (dctx->legacyContext)
ZSTD_freeLegacyStreamContext(dctx->legacyContext, dctx->previousLegacyVersion);
#if ZSTD_HEAPMODE >= 1
(void)dst;
(void)dstCapacity;
(void)src;
(void)srcSize;
return ERROR(GENERIC);
#else
ZSTD_DCtx dctx;
ZSTD_DCtx* const initialized = ZSTD_initStaticDCtx(&dctx, sizeof(dctx));
if (initialized == NULL) return ERROR(memory_allocation);
/* This is a stack DCtx, not a user-provided static workspace. Keep the
* original heapmode=0 semantics so legacy decoding is permitted and no
* static-context allocation restrictions leak into the one-shot API. */
initialized->staticSize = 0;
return ZSTD_decompressDCtx(initialized, dst, dstCapacity, src, srcSize);
#endif
if (dctx->ddictSet) {
ZSTD_freeDDictHashSet(dctx->ddictSet, cMem);
dctx->ddictSet = NULL;
}
ZSTD_customFree(dctx, cMem);
return 0;
}
}
/* no longer useful */
void ZSTD_copyDCtx(ZSTD_DCtx* dstDCtx, const ZSTD_DCtx* srcDCtx)
void* ZSTD_rust_custom_malloc(size_t size, ZSTD_customMem customMem)
{
size_t const toCopy = (size_t)((char*)(&dstDCtx->inBuff) - (char*)dstDCtx);
ZSTD_memcpy(dstDCtx, srcDCtx, toCopy); /* no need to copy workspace */
return ZSTD_customMalloc(size, customMem);
}
/* Given a dctx with a digested frame params, re-selects the correct ZSTD_DDict based on
* the requested dict ID from the frame. If there exists a reference to the correct ZSTD_DDict, then
* accordingly sets the ddict to be used to decompress the frame.
*
* If no DDict is found, then no action is taken, and the ZSTD_DCtx::ddict remains as-is.
*
* ZSTD_d_refMultipleDDicts must be enabled for this function to be called.
*/
static void ZSTD_DCtx_selectFrameDDict(ZSTD_DCtx* dctx) {
assert(dctx->refMultipleDDicts && dctx->ddictSet);
DEBUGLOG(4, "Adjusting DDict based on requested dict ID from frame");
if (dctx->ddict) {
const ZSTD_DDict* frameDDict = ZSTD_DDictHashSet_getDDict(dctx->ddictSet, dctx->fParams.dictID);
if (frameDDict) {
DEBUGLOG(4, "DDict found!");
ZSTD_clearDict(dctx);
dctx->dictID = dctx->fParams.dictID;
dctx->ddict = frameDDict;
dctx->dictUses = ZSTD_use_indefinitely;
}
}
}
/*-*************************************************************
* Frame header decoding
***************************************************************/
/*! ZSTD_isFrame() :
* Tells if the content of `buffer` starts with a valid Frame Identifier.
* Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0.
* Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled.
* Note 3 : Skippable Frame Identifiers are considered valid. */
unsigned ZSTD_isFrame(const void* buffer, size_t size)
void* ZSTD_rust_custom_calloc(size_t size, ZSTD_customMem customMem)
{
if (size < ZSTD_FRAMEIDSIZE) return 0;
{ U32 const magic = MEM_readLE32(buffer);
if (magic == ZSTD_MAGICNUMBER) return 1;
if ((magic & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) return 1;
}
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (ZSTD_isLegacy(buffer, size)) return 1;
return ZSTD_customCalloc(size, customMem);
}
void ZSTD_rust_custom_free(void* allocation, ZSTD_customMem customMem)
{
ZSTD_customFree(allocation, customMem);
}
ZSTD_DDict* ZSTD_rust_create_ddict(const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType,
ZSTD_customMem customMem)
{
return ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod,
dictContentType, customMem);
}
void ZSTD_rust_dctx_copy_prefix(ZSTD_DCtx* dstDCtx, const ZSTD_DCtx* srcDCtx)
{
size_t const toCopy = (size_t)((const char*)(&dstDCtx->inBuff) - (const char*)dstDCtx);
ZSTD_memcpy(dstDCtx, srcDCtx, toCopy);
}
void ZSTD_rust_dctx_trace_begin(ZSTD_DCtx* dctx)
{
#if ZSTD_TRACE
dctx->traceCtx = (ZSTD_trace_decompress_begin != NULL)
? ZSTD_trace_decompress_begin(dctx) : 0;
#else
(void)dctx;
#endif
return 0;
}
/*! ZSTD_isSkippableFrame() :
* Tells if the content of `buffer` starts with a valid Frame Identifier for a skippable frame.
* Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0.
*/
unsigned ZSTD_isSkippableFrame(const void* buffer, size_t size)
{
if (size < ZSTD_FRAMEIDSIZE) return 0;
{ U32 const magic = MEM_readLE32(buffer);
if ((magic & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) return 1;
}
return 0;
}
/** ZSTD_frameHeaderSize_internal() :
* srcSize must be large enough to reach header size fields.
* note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless.
* @return : size of the Frame Header
* or an error code, which can be tested with ZSTD_isError() */
static size_t ZSTD_frameHeaderSize_internal(const void* src, size_t srcSize, ZSTD_format_e format)
{
size_t const minInputSize = ZSTD_startingInputLength(format);
RETURN_ERROR_IF(srcSize < minInputSize, srcSize_wrong, "");
{ BYTE const fhd = ((const BYTE*)src)[minInputSize-1];
U32 const dictID= fhd & 3;
U32 const singleSegment = (fhd >> 5) & 1;
U32 const fcsId = fhd >> 6;
return minInputSize + !singleSegment
+ ZSTD_did_fieldSize[dictID] + ZSTD_fcs_fieldSize[fcsId]
+ (singleSegment && !fcsId);
}
}
/** ZSTD_frameHeaderSize() :
* srcSize must be >= ZSTD_frameHeaderSize_prefix.
* @return : size of the Frame Header,
* or an error code (if srcSize is too small) */
size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize)
{
return ZSTD_frameHeaderSize_internal(src, srcSize, ZSTD_f_zstd1);
}
/** ZSTD_getFrameHeader_advanced() :
* decode Frame Header, or require larger `srcSize`.
* note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless
* @return : 0, `zfhPtr` is correctly filled,
* >0, `srcSize` is too small, value is wanted `srcSize` amount,
** or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_getFrameHeader_advanced(ZSTD_FrameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format)
{
const BYTE* ip = (const BYTE*)src;
size_t const minInputSize = ZSTD_startingInputLength(format);
DEBUGLOG(5, "ZSTD_getFrameHeader_advanced: minInputSize = %zu, srcSize = %zu", minInputSize, srcSize);
if (srcSize > 0) {
/* note : technically could be considered an assert(), since it's an invalid entry */
RETURN_ERROR_IF(src==NULL, GENERIC, "invalid parameter : src==NULL, but srcSize>0");
}
if (srcSize < minInputSize) {
if (srcSize > 0 && format != ZSTD_f_zstd1_magicless) {
/* when receiving less than @minInputSize bytes,
* control these bytes at least correspond to a supported magic number
* in order to error out early if they don't.
**/
size_t const toCopy = MIN(4, srcSize);
unsigned char hbuf[4]; MEM_writeLE32(hbuf, ZSTD_MAGICNUMBER);
assert(src != NULL);
ZSTD_memcpy(hbuf, src, toCopy);
if ( MEM_readLE32(hbuf) != ZSTD_MAGICNUMBER ) {
/* not a zstd frame : let's check if it's a skippable frame */
MEM_writeLE32(hbuf, ZSTD_MAGIC_SKIPPABLE_START);
ZSTD_memcpy(hbuf, src, toCopy);
if ((MEM_readLE32(hbuf) & ZSTD_MAGIC_SKIPPABLE_MASK) != ZSTD_MAGIC_SKIPPABLE_START) {
RETURN_ERROR(prefix_unknown,
"first bytes don't correspond to any supported magic number");
} } }
return minInputSize;
}
ZSTD_memset(zfhPtr, 0, sizeof(*zfhPtr)); /* not strictly necessary, but static analyzers may not understand that zfhPtr will be read only if return value is zero, since they are 2 different signals */
if ( (format != ZSTD_f_zstd1_magicless)
&& (MEM_readLE32(src) != ZSTD_MAGICNUMBER) ) {
if ((MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
/* skippable frame */
if (srcSize < ZSTD_SKIPPABLEHEADERSIZE)
return ZSTD_SKIPPABLEHEADERSIZE; /* magic number + frame length */
ZSTD_memset(zfhPtr, 0, sizeof(*zfhPtr));
zfhPtr->frameType = ZSTD_skippableFrame;
zfhPtr->dictID = MEM_readLE32(src) - ZSTD_MAGIC_SKIPPABLE_START;
zfhPtr->headerSize = ZSTD_SKIPPABLEHEADERSIZE;
zfhPtr->frameContentSize = MEM_readLE32((const char *)src + ZSTD_FRAMEIDSIZE);
return 0;
}
RETURN_ERROR(prefix_unknown, "");
}
/* ensure there is enough `srcSize` to fully read/decode frame header */
{ size_t const fhsize = ZSTD_frameHeaderSize_internal(src, srcSize, format);
if (srcSize < fhsize) return fhsize;
zfhPtr->headerSize = (U32)fhsize;
}
{ BYTE const fhdByte = ip[minInputSize-1];
size_t pos = minInputSize;
U32 const dictIDSizeCode = fhdByte&3;
U32 const checksumFlag = (fhdByte>>2)&1;
U32 const singleSegment = (fhdByte>>5)&1;
U32 const fcsID = fhdByte>>6;
U64 windowSize = 0;
U32 dictID = 0;
U64 frameContentSize = ZSTD_CONTENTSIZE_UNKNOWN;
RETURN_ERROR_IF((fhdByte & 0x08) != 0, frameParameter_unsupported,
"reserved bits, must be zero");
if (!singleSegment) {
BYTE const wlByte = ip[pos++];
U32 const windowLog = (wlByte >> 3) + ZSTD_WINDOWLOG_ABSOLUTEMIN;
RETURN_ERROR_IF(windowLog > ZSTD_WINDOWLOG_MAX, frameParameter_windowTooLarge, "");
windowSize = (1ULL << windowLog);
windowSize += (windowSize >> 3) * (wlByte&7);
}
switch(dictIDSizeCode)
{
default:
assert(0); /* impossible */
ZSTD_FALLTHROUGH;
case 0 : break;
case 1 : dictID = ip[pos]; pos++; break;
case 2 : dictID = MEM_readLE16(ip+pos); pos+=2; break;
case 3 : dictID = MEM_readLE32(ip+pos); pos+=4; break;
}
switch(fcsID)
{
default:
assert(0); /* impossible */
ZSTD_FALLTHROUGH;
case 0 : if (singleSegment) frameContentSize = ip[pos]; break;
case 1 : frameContentSize = MEM_readLE16(ip+pos)+256; break;
case 2 : frameContentSize = MEM_readLE32(ip+pos); break;
case 3 : frameContentSize = MEM_readLE64(ip+pos); break;
}
if (singleSegment) windowSize = frameContentSize;
zfhPtr->frameType = ZSTD_frame;
zfhPtr->frameContentSize = frameContentSize;
zfhPtr->windowSize = windowSize;
zfhPtr->blockSizeMax = (unsigned) MIN(windowSize, ZSTD_BLOCKSIZE_MAX);
zfhPtr->dictID = dictID;
zfhPtr->checksumFlag = checksumFlag;
}
return 0;
}
/** ZSTD_getFrameHeader() :
* decode Frame Header, or require larger `srcSize`.
* note : this function does not consume input, it only reads it.
* @return : 0, `zfhPtr` is correctly filled,
* >0, `srcSize` is too small, value is wanted `srcSize` amount,
* or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_getFrameHeader(ZSTD_FrameHeader* zfhPtr, const void* src, size_t srcSize)
{
return ZSTD_getFrameHeader_advanced(zfhPtr, src, srcSize, ZSTD_f_zstd1);
}
/** ZSTD_getFrameContentSize() :
* compatible with legacy mode
* @return : decompressed size of the single frame pointed to be `src` if known, otherwise
* - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined
* - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) */
unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize)
{
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (ZSTD_isLegacy(src, srcSize)) {
unsigned long long const ret = ZSTD_getDecompressedSize_legacy(src, srcSize);
return ret == 0 ? ZSTD_CONTENTSIZE_UNKNOWN : ret;
}
#endif
{ ZSTD_FrameHeader zfh;
if (ZSTD_getFrameHeader(&zfh, src, srcSize) != 0)
return ZSTD_CONTENTSIZE_ERROR;
if (zfh.frameType == ZSTD_skippableFrame) {
return 0;
} else {
return zfh.frameContentSize;
} }
}
static size_t readSkippableFrameSize(void const* src, size_t srcSize)
{
size_t const skippableHeaderSize = ZSTD_SKIPPABLEHEADERSIZE;
U32 sizeU32;
RETURN_ERROR_IF(srcSize < ZSTD_SKIPPABLEHEADERSIZE, srcSize_wrong, "");
sizeU32 = MEM_readLE32((BYTE const*)src + ZSTD_FRAMEIDSIZE);
RETURN_ERROR_IF((U32)(sizeU32 + ZSTD_SKIPPABLEHEADERSIZE) < sizeU32,
frameParameter_unsupported, "");
{ size_t const skippableSize = skippableHeaderSize + sizeU32;
RETURN_ERROR_IF(skippableSize > srcSize, srcSize_wrong, "");
return skippableSize;
}
}
/*! ZSTD_readSkippableFrame() :
* Retrieves content of a skippable frame, and writes it to dst buffer.
*
* The parameter magicVariant will receive the magicVariant that was supplied when the frame was written,
* i.e. magicNumber - ZSTD_MAGIC_SKIPPABLE_START. This can be NULL if the caller is not interested
* in the magicVariant.
*
* Returns an error if destination buffer is not large enough, or if this is not a valid skippable frame.
*
* @return : number of bytes written or a ZSTD error.
*/
size_t ZSTD_readSkippableFrame(void* dst, size_t dstCapacity,
unsigned* magicVariant, /* optional, can be NULL */
const void* src, size_t srcSize)
{
RETURN_ERROR_IF(srcSize < ZSTD_SKIPPABLEHEADERSIZE, srcSize_wrong, "");
{ U32 const magicNumber = MEM_readLE32(src);
size_t skippableFrameSize = readSkippableFrameSize(src, srcSize);
size_t skippableContentSize = skippableFrameSize - ZSTD_SKIPPABLEHEADERSIZE;
/* check input validity */
RETURN_ERROR_IF(!ZSTD_isSkippableFrame(src, srcSize), frameParameter_unsupported, "");
RETURN_ERROR_IF(skippableFrameSize < ZSTD_SKIPPABLEHEADERSIZE || skippableFrameSize > srcSize, srcSize_wrong, "");
RETURN_ERROR_IF(skippableContentSize > dstCapacity, dstSize_tooSmall, "");
/* deliver payload */
if (skippableContentSize > 0 && dst != NULL)
ZSTD_memcpy(dst, (const BYTE *)src + ZSTD_SKIPPABLEHEADERSIZE, skippableContentSize);
if (magicVariant != NULL)
*magicVariant = magicNumber - ZSTD_MAGIC_SKIPPABLE_START;
return skippableContentSize;
}
}
/** ZSTD_findDecompressedSize() :
* `srcSize` must be the exact length of some number of ZSTD compressed and/or
* skippable frames
* note: compatible with legacy mode
* @return : decompressed size of the frames contained */
unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize)
{
unsigned long long totalDstSize = 0;
while (srcSize >= ZSTD_startingInputLength(ZSTD_f_zstd1)) {
U32 const magicNumber = MEM_readLE32(src);
if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
size_t const skippableSize = readSkippableFrameSize(src, srcSize);
if (ZSTD_isError(skippableSize)) return ZSTD_CONTENTSIZE_ERROR;
assert(skippableSize <= srcSize);
src = (const BYTE *)src + skippableSize;
srcSize -= skippableSize;
continue;
}
{ unsigned long long const fcs = ZSTD_getFrameContentSize(src, srcSize);
if (fcs >= ZSTD_CONTENTSIZE_ERROR) return fcs;
if (totalDstSize + fcs < totalDstSize)
return ZSTD_CONTENTSIZE_ERROR; /* check for overflow */
totalDstSize += fcs;
}
/* skip to next frame */
{ size_t const frameSrcSize = ZSTD_findFrameCompressedSize(src, srcSize);
if (ZSTD_isError(frameSrcSize)) return ZSTD_CONTENTSIZE_ERROR;
assert(frameSrcSize <= srcSize);
src = (const BYTE *)src + frameSrcSize;
srcSize -= frameSrcSize;
}
} /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */
if (srcSize) return ZSTD_CONTENTSIZE_ERROR;
return totalDstSize;
}
/** ZSTD_getDecompressedSize() :
* compatible with legacy mode
* @return : decompressed size if known, 0 otherwise
note : 0 can mean any of the following :
- frame content is empty
- decompressed size field is not present in frame header
- frame header unknown / not supported
- frame header not complete (`srcSize` too small) */
unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize)
{
unsigned long long const ret = ZSTD_getFrameContentSize(src, srcSize);
ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_ERROR < ZSTD_CONTENTSIZE_UNKNOWN);
return (ret >= ZSTD_CONTENTSIZE_ERROR) ? 0 : ret;
}
/** ZSTD_decodeFrameHeader() :
* `headerSize` must be the size provided by ZSTD_frameHeaderSize().
* If multiple DDict references are enabled, also will choose the correct DDict to use.
* @return : 0 if success, or an error code, which can be tested using ZSTD_isError() */
static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx* dctx, const void* src, size_t headerSize)
{
size_t const result = ZSTD_getFrameHeader_advanced(&(dctx->fParams), src, headerSize, dctx->format);
if (ZSTD_isError(result)) return result; /* invalid header */
RETURN_ERROR_IF(result>0, srcSize_wrong, "headerSize too small");
/* Reference DDict requested by frame if dctx references multiple ddicts */
if (dctx->refMultipleDDicts == ZSTD_rmd_refMultipleDDicts && dctx->ddictSet) {
ZSTD_DCtx_selectFrameDDict(dctx);
}
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
/* Skip the dictID check in fuzzing mode, because it makes the search
* harder.
*/
RETURN_ERROR_IF(dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID),
dictionary_wrong, "");
#endif
dctx->validateChecksum = (dctx->fParams.checksumFlag && !dctx->forceIgnoreChecksum) ? 1 : 0;
if (dctx->validateChecksum) XXH64_reset(&dctx->xxhState, 0);
dctx->processedCSize += headerSize;
return 0;
}
static ZSTD_frameSizeInfo ZSTD_errorFrameSizeInfo(size_t ret)
{
ZSTD_frameSizeInfo frameSizeInfo;
frameSizeInfo.compressedSize = ret;
frameSizeInfo.decompressedBound = ZSTD_CONTENTSIZE_ERROR;
return frameSizeInfo;
}
static ZSTD_frameSizeInfo ZSTD_findFrameSizeInfo(const void* src, size_t srcSize, ZSTD_format_e format)
{
ZSTD_frameSizeInfo frameSizeInfo;
ZSTD_memset(&frameSizeInfo, 0, sizeof(ZSTD_frameSizeInfo));
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (format == ZSTD_f_zstd1 && ZSTD_isLegacy(src, srcSize))
return ZSTD_findFrameSizeInfoLegacy(src, srcSize);
#endif
if (format == ZSTD_f_zstd1 && (srcSize >= ZSTD_SKIPPABLEHEADERSIZE)
&& (MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
frameSizeInfo.compressedSize = readSkippableFrameSize(src, srcSize);
assert(ZSTD_isError(frameSizeInfo.compressedSize) ||
frameSizeInfo.compressedSize <= srcSize);
return frameSizeInfo;
} else {
const BYTE* ip = (const BYTE*)src;
const BYTE* const ipstart = ip;
size_t remainingSize = srcSize;
size_t nbBlocks = 0;
ZSTD_FrameHeader zfh;
/* Extract Frame Header */
{ size_t const ret = ZSTD_getFrameHeader_advanced(&zfh, src, srcSize, format);
if (ZSTD_isError(ret))
return ZSTD_errorFrameSizeInfo(ret);
if (ret > 0)
return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong));
}
ip += zfh.headerSize;
remainingSize -= zfh.headerSize;
/* Iterate over each block */
while (1) {
blockProperties_t blockProperties;
size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSize, &blockProperties);
if (ZSTD_isError(cBlockSize))
return ZSTD_errorFrameSizeInfo(cBlockSize);
if (ZSTD_blockHeaderSize + cBlockSize > remainingSize)
return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong));
ip += ZSTD_blockHeaderSize + cBlockSize;
remainingSize -= ZSTD_blockHeaderSize + cBlockSize;
nbBlocks++;
if (blockProperties.lastBlock) break;
}
/* Final frame content checksum */
if (zfh.checksumFlag) {
if (remainingSize < 4)
return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong));
ip += 4;
}
frameSizeInfo.nbBlocks = nbBlocks;
frameSizeInfo.compressedSize = (size_t)(ip - ipstart);
frameSizeInfo.decompressedBound = (zfh.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN)
? zfh.frameContentSize
: (unsigned long long)nbBlocks * zfh.blockSizeMax;
return frameSizeInfo;
}
}
static size_t ZSTD_findFrameCompressedSize_advanced(const void *src, size_t srcSize, ZSTD_format_e format) {
ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize, format);
return frameSizeInfo.compressedSize;
}
/** ZSTD_findFrameCompressedSize() :
* See docs in zstd.h
* Note: compatible with legacy mode */
size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize)
{
return ZSTD_findFrameCompressedSize_advanced(src, srcSize, ZSTD_f_zstd1);
}
/** ZSTD_decompressBound() :
* compatible with legacy mode
* `src` must point to the start of a ZSTD frame or a skippable frame
* `srcSize` must be at least as large as the frame contained
* @return : the maximum decompressed size of the compressed source
*/
unsigned long long ZSTD_decompressBound(const void* src, size_t srcSize)
{
unsigned long long bound = 0;
/* Iterate over each frame */
while (srcSize > 0) {
ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize, ZSTD_f_zstd1);
size_t const compressedSize = frameSizeInfo.compressedSize;
unsigned long long const decompressedBound = frameSizeInfo.decompressedBound;
if (ZSTD_isError(compressedSize) || decompressedBound == ZSTD_CONTENTSIZE_ERROR)
return ZSTD_CONTENTSIZE_ERROR;
assert(srcSize >= compressedSize);
src = (const BYTE*)src + compressedSize;
srcSize -= compressedSize;
bound += decompressedBound;
}
return bound;
}
size_t ZSTD_decompressionMargin(void const* src, size_t srcSize)
{
size_t margin = 0;
unsigned maxBlockSize = 0;
/* Iterate over each frame */
while (srcSize > 0) {
ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize, ZSTD_f_zstd1);
size_t const compressedSize = frameSizeInfo.compressedSize;
unsigned long long const decompressedBound = frameSizeInfo.decompressedBound;
ZSTD_FrameHeader zfh;
FORWARD_IF_ERROR(ZSTD_getFrameHeader(&zfh, src, srcSize), "");
if (ZSTD_isError(compressedSize) || decompressedBound == ZSTD_CONTENTSIZE_ERROR)
return ERROR(corruption_detected);
if (zfh.frameType == ZSTD_frame) {
/* Add the frame header to our margin */
margin += zfh.headerSize;
/* Add the checksum to our margin */
margin += zfh.checksumFlag ? 4 : 0;
/* Add 3 bytes per block */
margin += 3 * frameSizeInfo.nbBlocks;
/* Compute the max block size */
maxBlockSize = MAX(maxBlockSize, zfh.blockSizeMax);
} else {
assert(zfh.frameType == ZSTD_skippableFrame);
/* Add the entire skippable frame size to our margin. */
margin += compressedSize;
}
assert(srcSize >= compressedSize);
src = (const BYTE*)src + compressedSize;
srcSize -= compressedSize;
}
/* Add the max block size back to the margin. */
margin += maxBlockSize;
return margin;
}
/*-*************************************************************
* Frame decoding
***************************************************************/
/** ZSTD_insertBlock() :
* insert `src` block into `dctx` history. Useful to track uncompressed blocks. */
size_t ZSTD_insertBlock(ZSTD_DCtx* dctx, const void* blockStart, size_t blockSize)
{
DEBUGLOG(5, "ZSTD_insertBlock: %u bytes", (unsigned)blockSize);
ZSTD_checkContinuity(dctx, blockStart, blockSize);
dctx->previousDstEnd = (const char*)blockStart + blockSize;
return blockSize;
}
static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity,
const void* src, size_t srcSize)
{
DEBUGLOG(5, "ZSTD_copyRawBlock");
RETURN_ERROR_IF(srcSize > dstCapacity, dstSize_tooSmall, "");
if (dst == NULL) {
if (srcSize == 0) return 0;
RETURN_ERROR(dstBuffer_null, "");
}
ZSTD_memmove(dst, src, srcSize);
return srcSize;
}
static size_t ZSTD_setRleBlock(void* dst, size_t dstCapacity,
BYTE b,
size_t regenSize)
{
RETURN_ERROR_IF(regenSize > dstCapacity, dstSize_tooSmall, "");
if (dst == NULL) {
if (regenSize == 0) return 0;
RETURN_ERROR(dstBuffer_null, "");
}
ZSTD_memset(dst, b, regenSize);
return regenSize;
}
static void ZSTD_DCtx_trace_end(ZSTD_DCtx const* dctx, U64 uncompressedSize, U64 compressedSize, int streaming)
void ZSTD_rust_dctx_trace_end(ZSTD_DCtx* dctx, U64 uncompressedSize,
U64 compressedSize, int streaming)
{
#if ZSTD_TRACE
if (dctx->traceCtx && ZSTD_trace_decompress_end != NULL) {
@@ -945,1466 +356,127 @@ static void ZSTD_DCtx_trace_end(ZSTD_DCtx const* dctx, U64 uncompressedSize, U64
#endif
}
/*! ZSTD_decompressFrame() :
* @dctx must be properly initialized
* will update *srcPtr and *srcSizePtr,
* to make *srcPtr progress by one frame. */
static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity,
const void** srcPtr, size_t *srcSizePtr)
unsigned ZSTD_rust_legacy_is(const void* src, size_t srcSize)
{
const BYTE* const istart = (const BYTE*)(*srcPtr);
const BYTE* ip = istart;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = dstCapacity != 0 ? ostart + dstCapacity : ostart;
BYTE* op = ostart;
size_t remainingSrcSize = *srcSizePtr;
DEBUGLOG(4, "ZSTD_decompressFrame (srcSize:%i)", (int)*srcSizePtr);
/* check */
RETURN_ERROR_IF(
remainingSrcSize < ZSTD_FRAMEHEADERSIZE_MIN(dctx->format)+ZSTD_blockHeaderSize,
srcSize_wrong, "");
/* Frame Header */
{ size_t const frameHeaderSize = ZSTD_frameHeaderSize_internal(
ip, ZSTD_FRAMEHEADERSIZE_PREFIX(dctx->format), dctx->format);
if (ZSTD_isError(frameHeaderSize)) return frameHeaderSize;
RETURN_ERROR_IF(remainingSrcSize < frameHeaderSize+ZSTD_blockHeaderSize,
srcSize_wrong, "");
FORWARD_IF_ERROR( ZSTD_decodeFrameHeader(dctx, ip, frameHeaderSize) , "");
ip += frameHeaderSize; remainingSrcSize -= frameHeaderSize;
}
/* Shrink the blockSizeMax if enabled */
if (dctx->maxBlockSizeParam != 0)
dctx->fParams.blockSizeMax = MIN(dctx->fParams.blockSizeMax, (unsigned)dctx->maxBlockSizeParam);
/* Loop on each block */
while (1) {
BYTE* oBlockEnd = oend;
size_t decodedSize;
blockProperties_t blockProperties;
size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSrcSize, &blockProperties);
if (ZSTD_isError(cBlockSize)) return cBlockSize;
ip += ZSTD_blockHeaderSize;
remainingSrcSize -= ZSTD_blockHeaderSize;
RETURN_ERROR_IF(cBlockSize > remainingSrcSize, srcSize_wrong, "");
if (ip >= op && ip < oBlockEnd) {
/* We are decompressing in-place. Limit the output pointer so that we
* don't overwrite the block that we are currently reading. This will
* fail decompression if the input & output pointers aren't spaced
* far enough apart.
*
* This is important to set, even when the pointers are far enough
* apart, because ZSTD_decompressBlock_internal() can decide to store
* literals in the output buffer, after the block it is decompressing.
* Since we don't want anything to overwrite our input, we have to tell
* ZSTD_decompressBlock_internal to never write past ip.
*
* See ZSTD_allocateLiteralsBuffer() for reference.
*/
oBlockEnd = op + (ip - op);
}
switch(blockProperties.blockType)
{
case bt_compressed:
assert(dctx->isFrameDecompression == 1);
decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oBlockEnd-op), ip, cBlockSize, not_streaming);
break;
case bt_raw :
/* Use oend instead of oBlockEnd because this function is safe to overlap. It uses memmove. */
decodedSize = ZSTD_copyRawBlock(op, (size_t)(oend-op), ip, cBlockSize);
break;
case bt_rle :
decodedSize = ZSTD_setRleBlock(op, (size_t)(oBlockEnd-op), *ip, blockProperties.origSize);
break;
case bt_reserved :
default:
RETURN_ERROR(corruption_detected, "invalid block type");
}
FORWARD_IF_ERROR(decodedSize, "Block decompression failure");
DEBUGLOG(5, "Decompressed block of dSize = %u", (unsigned)decodedSize);
if (dctx->validateChecksum) {
XXH64_update(&dctx->xxhState, op, decodedSize);
}
if (decodedSize) /* support dst = NULL,0 */ {
op += decodedSize;
}
assert(ip != NULL);
ip += cBlockSize;
remainingSrcSize -= cBlockSize;
if (blockProperties.lastBlock) break;
}
if (dctx->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN) {
RETURN_ERROR_IF((U64)(op-ostart) != dctx->fParams.frameContentSize,
corruption_detected, "");
}
if (dctx->fParams.checksumFlag) { /* Frame content checksum verification */
RETURN_ERROR_IF(remainingSrcSize<4, checksum_wrong, "");
if (!dctx->forceIgnoreChecksum) {
U32 const checkCalc = (U32)XXH64_digest(&dctx->xxhState);
U32 checkRead;
checkRead = MEM_readLE32(ip);
RETURN_ERROR_IF(checkRead != checkCalc, checksum_wrong, "");
}
ip += 4;
remainingSrcSize -= 4;
}
ZSTD_DCtx_trace_end(dctx, (U64)(op-ostart), (U64)(ip-istart), /* streaming */ 0);
/* Allow caller to get size read */
DEBUGLOG(4, "ZSTD_decompressFrame: decompressed frame of size %i, consuming %i bytes of input", (int)(op-ostart), (int)(ip - (const BYTE*)*srcPtr));
*srcPtr = ip;
*srcSizePtr = remainingSrcSize;
return (size_t)(op-ostart);
}
static
ZSTD_ALLOW_POINTER_OVERFLOW_ATTR
size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const void* dict, size_t dictSize,
const ZSTD_DDict* ddict)
{
void* const dststart = dst;
int moreThan1Frame = 0;
DEBUGLOG(5, "ZSTD_decompressMultiFrame");
assert(dict==NULL || ddict==NULL); /* either dict or ddict set, not both */
if (ddict) {
dict = ZSTD_DDict_dictContent(ddict);
dictSize = ZSTD_DDict_dictSize(ddict);
}
while (srcSize >= ZSTD_startingInputLength(dctx->format)) {
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (dctx->format == ZSTD_f_zstd1 && ZSTD_isLegacy(src, srcSize)) {
size_t decodedSize;
size_t const frameSize = ZSTD_findFrameCompressedSizeLegacy(src, srcSize);
if (ZSTD_isError(frameSize)) return frameSize;
RETURN_ERROR_IF(dctx->staticSize, memory_allocation,
"legacy support is not compatible with static dctx");
decodedSize = ZSTD_decompressLegacy(dst, dstCapacity, src, frameSize, dict, dictSize);
if (ZSTD_isError(decodedSize)) return decodedSize;
{
unsigned long long const expectedSize = ZSTD_getFrameContentSize(src, srcSize);
RETURN_ERROR_IF(expectedSize == ZSTD_CONTENTSIZE_ERROR, corruption_detected, "Corrupted frame header!");
if (expectedSize != ZSTD_CONTENTSIZE_UNKNOWN) {
RETURN_ERROR_IF(expectedSize != decodedSize, corruption_detected,
"Frame header size does not match decoded size!");
}
}
assert(decodedSize <= dstCapacity);
dst = (BYTE*)dst + decodedSize;
dstCapacity -= decodedSize;
src = (const BYTE*)src + frameSize;
srcSize -= frameSize;
continue;
}
#endif
if (dctx->format == ZSTD_f_zstd1 && srcSize >= 4) {
U32 const magicNumber = MEM_readLE32(src);
DEBUGLOG(5, "reading magic number %08X", (unsigned)magicNumber);
if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) {
/* skippable frame detected : skip it */
size_t const skippableSize = readSkippableFrameSize(src, srcSize);
FORWARD_IF_ERROR(skippableSize, "invalid skippable frame");
assert(skippableSize <= srcSize);
src = (const BYTE *)src + skippableSize;
srcSize -= skippableSize;
continue; /* check next frame */
} }
if (ddict) {
/* we were called from ZSTD_decompress_usingDDict */
FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDDict(dctx, ddict), "");
} else {
/* this will initialize correctly with no dict if dict == NULL, so
* use this in all cases but ddict */
FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDict(dctx, dict, dictSize), "");
}
ZSTD_checkContinuity(dctx, dst, dstCapacity);
{ const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity,
&src, &srcSize);
RETURN_ERROR_IF(
(ZSTD_getErrorCode(res) == ZSTD_error_prefix_unknown)
&& (moreThan1Frame==1),
srcSize_wrong,
"At least one frame successfully completed, "
"but following bytes are garbage: "
"it's more likely to be a srcSize error, "
"specifying more input bytes than size of frame(s). "
"Note: one could be unlucky, it might be a corruption error instead, "
"happening right at the place where we expect zstd magic bytes. "
"But this is _much_ less likely than a srcSize field error.");
if (ZSTD_isError(res)) return res;
assert(res <= dstCapacity);
if (res != 0)
dst = (BYTE*)dst + res;
dstCapacity -= res;
}
moreThan1Frame = 1;
} /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */
RETURN_ERROR_IF(srcSize, srcSize_wrong, "input not entirely consumed");
return (size_t)((BYTE*)dst - (BYTE*)dststart);
}
size_t ZSTD_decompress_usingDict(ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const void* dict, size_t dictSize)
{
return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize, dict, dictSize, NULL);
}
static ZSTD_DDict const* ZSTD_getDDict(ZSTD_DCtx* dctx)
{
switch (dctx->dictUses) {
default:
assert(0 /* Impossible */);
ZSTD_FALLTHROUGH;
case ZSTD_dont_use:
ZSTD_clearDict(dctx);
return NULL;
case ZSTD_use_indefinitely:
return dctx->ddict;
case ZSTD_use_once:
dctx->dictUses = ZSTD_dont_use;
return dctx->ddict;
}
}
size_t ZSTD_decompressDCtx(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
return ZSTD_decompress_usingDDict(dctx, dst, dstCapacity, src, srcSize, ZSTD_getDDict(dctx));
}
size_t ZSTD_decompress(void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
#if defined(ZSTD_HEAPMODE) && (ZSTD_HEAPMODE>=1)
size_t regenSize;
ZSTD_DCtx* const dctx = ZSTD_createDCtx_internal(ZSTD_defaultCMem);
RETURN_ERROR_IF(dctx==NULL, memory_allocation, "NULL pointer!");
regenSize = ZSTD_decompressDCtx(dctx, dst, dstCapacity, src, srcSize);
ZSTD_freeDCtx(dctx);
return regenSize;
#else /* stack mode */
ZSTD_DCtx dctx;
ZSTD_initDCtx_internal(&dctx);
return ZSTD_decompressDCtx(&dctx, dst, dstCapacity, src, srcSize);
#endif
}
/*-**************************************
* Advanced Streaming Decompression API
* Bufferless and synchronous
****************************************/
size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx* dctx) { return dctx->expected; }
/**
* Similar to ZSTD_nextSrcSizeToDecompress(), but when a block input can be streamed, we
* allow taking a partial block as the input. Currently only raw uncompressed blocks can
* be streamed.
*
* For blocks that can be streamed, this allows us to reduce the latency until we produce
* output, and avoid copying the input.
*
* @param inputSize - The total amount of input that the caller currently has.
*/
static size_t ZSTD_nextSrcSizeToDecompressWithInputSize(ZSTD_DCtx* dctx, size_t inputSize) {
if (!(dctx->stage == ZSTDds_decompressBlock || dctx->stage == ZSTDds_decompressLastBlock))
return dctx->expected;
if (dctx->bType != bt_raw)
return dctx->expected;
return BOUNDED(1, inputSize, dctx->expected);
}
ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx* dctx) {
switch(dctx->stage)
{
default: /* should not happen */
assert(0);
ZSTD_FALLTHROUGH;
case ZSTDds_getFrameHeaderSize:
ZSTD_FALLTHROUGH;
case ZSTDds_decodeFrameHeader:
return ZSTDnit_frameHeader;
case ZSTDds_decodeBlockHeader:
return ZSTDnit_blockHeader;
case ZSTDds_decompressBlock:
return ZSTDnit_block;
case ZSTDds_decompressLastBlock:
return ZSTDnit_lastBlock;
case ZSTDds_checkChecksum:
return ZSTDnit_checksum;
case ZSTDds_decodeSkippableHeader:
ZSTD_FALLTHROUGH;
case ZSTDds_skipFrame:
return ZSTDnit_skippableFrame;
}
}
static int ZSTD_isSkipFrame(ZSTD_DCtx* dctx) { return dctx->stage == ZSTDds_skipFrame; }
/** ZSTD_decompressContinue() :
* srcSize : must be the exact nb of bytes expected (see ZSTD_nextSrcSizeToDecompress())
* @return : nb of bytes generated into `dst` (necessarily <= `dstCapacity)
* or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
DEBUGLOG(5, "ZSTD_decompressContinue (srcSize:%u)", (unsigned)srcSize);
/* Sanity check */
RETURN_ERROR_IF(srcSize != ZSTD_nextSrcSizeToDecompressWithInputSize(dctx, srcSize), srcSize_wrong, "not allowed");
ZSTD_checkContinuity(dctx, dst, dstCapacity);
dctx->processedCSize += srcSize;
switch (dctx->stage)
{
case ZSTDds_getFrameHeaderSize :
assert(src != NULL);
if (dctx->format == ZSTD_f_zstd1) { /* allows header */
assert(srcSize >= ZSTD_FRAMEIDSIZE); /* to read skippable magic number */
if ((MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { /* skippable frame */
ZSTD_memcpy(dctx->headerBuffer, src, srcSize);
dctx->expected = ZSTD_SKIPPABLEHEADERSIZE - srcSize; /* remaining to load to get full skippable frame header */
dctx->stage = ZSTDds_decodeSkippableHeader;
return 0;
} }
dctx->headerSize = ZSTD_frameHeaderSize_internal(src, srcSize, dctx->format);
if (ZSTD_isError(dctx->headerSize)) return dctx->headerSize;
ZSTD_memcpy(dctx->headerBuffer, src, srcSize);
dctx->expected = dctx->headerSize - srcSize;
dctx->stage = ZSTDds_decodeFrameHeader;
return 0;
case ZSTDds_decodeFrameHeader:
assert(src != NULL);
ZSTD_memcpy(dctx->headerBuffer + (dctx->headerSize - srcSize), src, srcSize);
FORWARD_IF_ERROR(ZSTD_decodeFrameHeader(dctx, dctx->headerBuffer, dctx->headerSize), "");
dctx->expected = ZSTD_blockHeaderSize;
dctx->stage = ZSTDds_decodeBlockHeader;
return 0;
case ZSTDds_decodeBlockHeader:
{ blockProperties_t bp;
size_t const cBlockSize = ZSTD_getcBlockSize(src, ZSTD_blockHeaderSize, &bp);
if (ZSTD_isError(cBlockSize)) return cBlockSize;
RETURN_ERROR_IF(cBlockSize > dctx->fParams.blockSizeMax, corruption_detected, "Block Size Exceeds Maximum");
dctx->expected = cBlockSize;
dctx->bType = bp.blockType;
dctx->rleSize = bp.origSize;
if (cBlockSize) {
dctx->stage = bp.lastBlock ? ZSTDds_decompressLastBlock : ZSTDds_decompressBlock;
return 0;
}
/* empty block */
if (bp.lastBlock) {
if (dctx->fParams.checksumFlag) {
dctx->expected = 4;
dctx->stage = ZSTDds_checkChecksum;
} else {
dctx->expected = 0; /* end of frame */
dctx->stage = ZSTDds_getFrameHeaderSize;
}
} else {
dctx->expected = ZSTD_blockHeaderSize; /* jump to next header */
dctx->stage = ZSTDds_decodeBlockHeader;
}
return 0;
}
case ZSTDds_decompressLastBlock:
case ZSTDds_decompressBlock:
DEBUGLOG(5, "ZSTD_decompressContinue: case ZSTDds_decompressBlock");
{ size_t rSize;
switch(dctx->bType)
{
case bt_compressed:
DEBUGLOG(5, "ZSTD_decompressContinue: case bt_compressed");
assert(dctx->isFrameDecompression == 1);
rSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize, is_streaming);
dctx->expected = 0; /* Streaming not supported */
break;
case bt_raw :
assert(srcSize <= dctx->expected);
rSize = ZSTD_copyRawBlock(dst, dstCapacity, src, srcSize);
FORWARD_IF_ERROR(rSize, "ZSTD_copyRawBlock failed");
assert(rSize == srcSize);
dctx->expected -= rSize;
break;
case bt_rle :
rSize = ZSTD_setRleBlock(dst, dstCapacity, *(const BYTE*)src, dctx->rleSize);
dctx->expected = 0; /* Streaming not supported */
break;
case bt_reserved : /* should never happen */
default:
RETURN_ERROR(corruption_detected, "invalid block type");
}
FORWARD_IF_ERROR(rSize, "");
RETURN_ERROR_IF(rSize > dctx->fParams.blockSizeMax, corruption_detected, "Decompressed Block Size Exceeds Maximum");
DEBUGLOG(5, "ZSTD_decompressContinue: decoded size from block : %u", (unsigned)rSize);
dctx->decodedSize += rSize;
if (dctx->validateChecksum) XXH64_update(&dctx->xxhState, dst, rSize);
dctx->previousDstEnd = (char*)dst + rSize;
/* Stay on the same stage until we are finished streaming the block. */
if (dctx->expected > 0) {
return rSize;
}
if (dctx->stage == ZSTDds_decompressLastBlock) { /* end of frame */
DEBUGLOG(4, "ZSTD_decompressContinue: decoded size from frame : %u", (unsigned)dctx->decodedSize);
RETURN_ERROR_IF(
dctx->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN
&& dctx->decodedSize != dctx->fParams.frameContentSize,
corruption_detected, "");
if (dctx->fParams.checksumFlag) { /* another round for frame checksum */
dctx->expected = 4;
dctx->stage = ZSTDds_checkChecksum;
} else {
ZSTD_DCtx_trace_end(dctx, dctx->decodedSize, dctx->processedCSize, /* streaming */ 1);
dctx->expected = 0; /* ends here */
dctx->stage = ZSTDds_getFrameHeaderSize;
}
} else {
dctx->stage = ZSTDds_decodeBlockHeader;
dctx->expected = ZSTD_blockHeaderSize;
}
return rSize;
}
case ZSTDds_checkChecksum:
assert(srcSize == 4); /* guaranteed by dctx->expected */
{
if (dctx->validateChecksum) {
U32 const h32 = (U32)XXH64_digest(&dctx->xxhState);
U32 const check32 = MEM_readLE32(src);
DEBUGLOG(4, "ZSTD_decompressContinue: checksum : calculated %08X :: %08X read", (unsigned)h32, (unsigned)check32);
RETURN_ERROR_IF(check32 != h32, checksum_wrong, "");
}
ZSTD_DCtx_trace_end(dctx, dctx->decodedSize, dctx->processedCSize, /* streaming */ 1);
dctx->expected = 0;
dctx->stage = ZSTDds_getFrameHeaderSize;
return 0;
}
case ZSTDds_decodeSkippableHeader:
assert(src != NULL);
assert(srcSize <= ZSTD_SKIPPABLEHEADERSIZE);
assert(dctx->format != ZSTD_f_zstd1_magicless);
ZSTD_memcpy(dctx->headerBuffer + (ZSTD_SKIPPABLEHEADERSIZE - srcSize), src, srcSize); /* complete skippable header */
dctx->expected = MEM_readLE32(dctx->headerBuffer + ZSTD_FRAMEIDSIZE); /* note : dctx->expected can grow seriously large, beyond local buffer size */
dctx->stage = ZSTDds_skipFrame;
return 0;
case ZSTDds_skipFrame:
dctx->expected = 0;
dctx->stage = ZSTDds_getFrameHeaderSize;
return 0;
default:
assert(0); /* impossible */
RETURN_ERROR(GENERIC, "impossible to reach"); /* some compilers require default to do something */
}
}
static size_t ZSTD_refDictContent(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
{
dctx->dictEnd = dctx->previousDstEnd;
dctx->virtualStart = (const char*)dict - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->prefixStart));
dctx->prefixStart = dict;
dctx->previousDstEnd = (const char*)dict + dictSize;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
dctx->dictContentBeginForFuzzing = dctx->prefixStart;
dctx->dictContentEndForFuzzing = dctx->previousDstEnd;
#endif
return 0;
}
/*! ZSTD_loadDEntropy() :
* dict : must point at beginning of a valid zstd dictionary.
* @return : size of entropy tables read */
size_t
ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy,
const void* const dict, size_t const dictSize)
{
const BYTE* dictPtr = (const BYTE*)dict;
const BYTE* const dictEnd = dictPtr + dictSize;
RETURN_ERROR_IF(dictSize <= 8, dictionary_corrupted, "dict is too small");
assert(MEM_readLE32(dict) == ZSTD_MAGIC_DICTIONARY); /* dict must be valid */
dictPtr += 8; /* skip header = magic + dictID */
ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, OFTable) == offsetof(ZSTD_entropyDTables_t, LLTable) + sizeof(entropy->LLTable));
ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, MLTable) == offsetof(ZSTD_entropyDTables_t, OFTable) + sizeof(entropy->OFTable));
ZSTD_STATIC_ASSERT(sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable) >= HUF_DECOMPRESS_WORKSPACE_SIZE);
{ void* const workspace = &entropy->LLTable; /* use fse tables as temporary workspace; implies fse tables are grouped together */
size_t const workspaceSize = sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable);
#ifdef HUF_FORCE_DECOMPRESS_X1
/* in minimal huffman, we always use X1 variants */
size_t const hSize = HUF_readDTableX1_wksp(entropy->hufTable,
dictPtr, dictEnd - dictPtr,
workspace, workspaceSize, /* flags */ 0);
return ZSTD_isLegacy(src, srcSize);
#else
size_t const hSize = HUF_readDTableX2_wksp(entropy->hufTable,
dictPtr, (size_t)(dictEnd - dictPtr),
workspace, workspaceSize, /* flags */ 0);
(void)src;
(void)srcSize;
return 0;
#endif
RETURN_ERROR_IF(HUF_isError(hSize), dictionary_corrupted, "");
dictPtr += hSize;
}
{ short offcodeNCount[MaxOff+1];
unsigned offcodeMaxValue = MaxOff, offcodeLog;
size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, (size_t)(dictEnd-dictPtr));
RETURN_ERROR_IF(FSE_isError(offcodeHeaderSize), dictionary_corrupted, "");
RETURN_ERROR_IF(offcodeMaxValue > MaxOff, dictionary_corrupted, "");
RETURN_ERROR_IF(offcodeLog > OffFSELog, dictionary_corrupted, "");
ZSTD_buildFSETable( entropy->OFTable,
offcodeNCount, offcodeMaxValue,
OF_base, OF_bits,
offcodeLog,
entropy->workspace, sizeof(entropy->workspace),
/* bmi2 */0);
dictPtr += offcodeHeaderSize;
}
{ short matchlengthNCount[MaxML+1];
unsigned matchlengthMaxValue = MaxML, matchlengthLog;
size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, (size_t)(dictEnd-dictPtr));
RETURN_ERROR_IF(FSE_isError(matchlengthHeaderSize), dictionary_corrupted, "");
RETURN_ERROR_IF(matchlengthMaxValue > MaxML, dictionary_corrupted, "");
RETURN_ERROR_IF(matchlengthLog > MLFSELog, dictionary_corrupted, "");
ZSTD_buildFSETable( entropy->MLTable,
matchlengthNCount, matchlengthMaxValue,
ML_base, ML_bits,
matchlengthLog,
entropy->workspace, sizeof(entropy->workspace),
/* bmi2 */ 0);
dictPtr += matchlengthHeaderSize;
}
{ short litlengthNCount[MaxLL+1];
unsigned litlengthMaxValue = MaxLL, litlengthLog;
size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, (size_t)(dictEnd-dictPtr));
RETURN_ERROR_IF(FSE_isError(litlengthHeaderSize), dictionary_corrupted, "");
RETURN_ERROR_IF(litlengthMaxValue > MaxLL, dictionary_corrupted, "");
RETURN_ERROR_IF(litlengthLog > LLFSELog, dictionary_corrupted, "");
ZSTD_buildFSETable( entropy->LLTable,
litlengthNCount, litlengthMaxValue,
LL_base, LL_bits,
litlengthLog,
entropy->workspace, sizeof(entropy->workspace),
/* bmi2 */ 0);
dictPtr += litlengthHeaderSize;
}
RETURN_ERROR_IF(dictPtr+12 > dictEnd, dictionary_corrupted, "");
{ int i;
size_t const dictContentSize = (size_t)(dictEnd - (dictPtr+12));
for (i=0; i<3; i++) {
U32 const rep = MEM_readLE32(dictPtr); dictPtr += 4;
RETURN_ERROR_IF(rep==0 || rep > dictContentSize,
dictionary_corrupted, "");
entropy->rep[i] = rep;
} }
return (size_t)(dictPtr - (const BYTE*)dict);
}
static size_t ZSTD_decompress_insertDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
unsigned long long ZSTD_rust_legacy_get_decompressed_size(const void* src, size_t srcSize)
{
if (dictSize < 8) return ZSTD_refDictContent(dctx, dict, dictSize);
{ U32 const magic = MEM_readLE32(dict);
if (magic != ZSTD_MAGIC_DICTIONARY) {
return ZSTD_refDictContent(dctx, dict, dictSize); /* pure content mode */
} }
dctx->dictID = MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE);
/* load entropy tables */
{ size_t const eSize = ZSTD_loadDEntropy(&dctx->entropy, dict, dictSize);
RETURN_ERROR_IF(ZSTD_isError(eSize), dictionary_corrupted, "");
dict = (const char*)dict + eSize;
dictSize -= eSize;
}
dctx->litEntropy = dctx->fseEntropy = 1;
/* reference dictionary content */
return ZSTD_refDictContent(dctx, dict, dictSize);
}
size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx)
{
assert(dctx != NULL);
#if ZSTD_TRACE
dctx->traceCtx = (ZSTD_trace_decompress_begin != NULL) ? ZSTD_trace_decompress_begin(dctx) : 0;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
return ZSTD_getDecompressedSize_legacy(src, srcSize);
#else
(void)src;
(void)srcSize;
return 0;
#endif
dctx->expected = ZSTD_startingInputLength(dctx->format); /* dctx->format must be properly set */
dctx->stage = ZSTDds_getFrameHeaderSize;
dctx->processedCSize = 0;
dctx->decodedSize = 0;
dctx->previousDstEnd = NULL;
dctx->prefixStart = NULL;
dctx->virtualStart = NULL;
dctx->dictEnd = NULL;
dctx->entropy.hufTable[0] = (HUF_DTable)((ZSTD_HUFFDTABLE_CAPACITY_LOG)*0x1000001); /* cover both little and big endian */
dctx->litEntropy = dctx->fseEntropy = 0;
dctx->dictID = 0;
dctx->bType = bt_reserved;
dctx->isFrameDecompression = 1;
ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue));
ZSTD_memcpy(dctx->entropy.rep, repStartValue, sizeof(repStartValue)); /* initial repcodes */
dctx->LLTptr = dctx->entropy.LLTable;
dctx->MLTptr = dctx->entropy.MLTable;
dctx->OFTptr = dctx->entropy.OFTable;
dctx->HUFptr = dctx->entropy.hufTable;
return 0;
}
size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
size_t ZSTD_rust_legacy_find_compressed_size(const void* src, size_t srcSize)
{
FORWARD_IF_ERROR( ZSTD_decompressBegin(dctx) , "");
if (dict && dictSize)
RETURN_ERROR_IF(
ZSTD_isError(ZSTD_decompress_insertDictionary(dctx, dict, dictSize)),
dictionary_corrupted, "");
return 0;
}
/* ====== ZSTD_DDict ====== */
size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
DEBUGLOG(4, "ZSTD_decompressBegin_usingDDict");
assert(dctx != NULL);
if (ddict) {
const char* const dictStart = (const char*)ZSTD_DDict_dictContent(ddict);
size_t const dictSize = ZSTD_DDict_dictSize(ddict);
const void* const dictEnd = dictStart + dictSize;
dctx->ddictIsCold = (dctx->dictEnd != dictEnd);
DEBUGLOG(4, "DDict is %s",
dctx->ddictIsCold ? "~cold~" : "hot!");
}
FORWARD_IF_ERROR( ZSTD_decompressBegin(dctx) , "");
if (ddict) { /* NULL ddict is equivalent to no dictionary */
ZSTD_copyDDictParameters(dctx, ddict);
}
return 0;
}
/*! ZSTD_getDictID_fromDict() :
* Provides the dictID stored within dictionary.
* if @return == 0, the dictionary is not conformant with Zstandard specification.
* It can still be loaded, but as a content-only dictionary. */
unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize)
{
if (dictSize < 8) return 0;
if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) return 0;
return MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE);
}
/*! ZSTD_getDictID_fromFrame() :
* Provides the dictID required to decompress frame stored within `src`.
* If @return == 0, the dictID could not be decoded.
* This could for one of the following reasons :
* - The frame does not require a dictionary (most common case).
* - The frame was built with dictID intentionally removed.
* Needed dictionary is a hidden piece of information.
* Note : this use case also happens when using a non-conformant dictionary.
* - `srcSize` is too small, and as a result, frame header could not be decoded.
* Note : possible if `srcSize < ZSTD_FRAMEHEADERSIZE_MAX`.
* - This is not a Zstandard frame.
* When identifying the exact failure cause, it's possible to use
* ZSTD_getFrameHeader(), which will provide a more precise error code. */
unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize)
{
ZSTD_FrameHeader zfp = { 0, 0, 0, ZSTD_frame, 0, 0, 0, 0, 0 };
size_t const hError = ZSTD_getFrameHeader(&zfp, src, srcSize);
if (ZSTD_isError(hError)) return 0;
return zfp.dictID;
}
/*! ZSTD_decompress_usingDDict() :
* Decompression using a pre-digested Dictionary
* Use dictionary without significant overhead. */
size_t ZSTD_decompress_usingDDict(ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const ZSTD_DDict* ddict)
{
/* pass content and size in case legacy frames are encountered */
return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize,
NULL, 0,
ddict);
}
/*=====================================
* Streaming decompression
*====================================*/
ZSTD_DStream* ZSTD_createDStream(void)
{
DEBUGLOG(3, "ZSTD_createDStream");
return ZSTD_createDCtx_internal(ZSTD_defaultCMem);
}
ZSTD_DStream* ZSTD_initStaticDStream(void *workspace, size_t workspaceSize)
{
return ZSTD_initStaticDCtx(workspace, workspaceSize);
}
ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem)
{
return ZSTD_createDCtx_internal(customMem);
}
size_t ZSTD_freeDStream(ZSTD_DStream* zds)
{
return ZSTD_freeDCtx(zds);
}
/* *** Initialization *** */
size_t ZSTD_DStreamInSize(void) { return ZSTD_BLOCKSIZE_MAX + ZSTD_blockHeaderSize; }
size_t ZSTD_DStreamOutSize(void) { return ZSTD_BLOCKSIZE_MAX; }
size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx,
const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType)
{
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, "");
ZSTD_clearDict(dctx);
if (dict && dictSize != 0) {
dctx->ddictLocal = ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod, dictContentType, dctx->customMem);
RETURN_ERROR_IF(dctx->ddictLocal == NULL, memory_allocation, "NULL pointer!");
dctx->ddict = dctx->ddictLocal;
dctx->dictUses = ZSTD_use_indefinitely;
}
return 0;
}
size_t ZSTD_DCtx_loadDictionary_byReference(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
{
return ZSTD_DCtx_loadDictionary_advanced(dctx, dict, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto);
}
size_t ZSTD_DCtx_loadDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
{
return ZSTD_DCtx_loadDictionary_advanced(dctx, dict, dictSize, ZSTD_dlm_byCopy, ZSTD_dct_auto);
}
size_t ZSTD_DCtx_refPrefix_advanced(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType)
{
FORWARD_IF_ERROR(ZSTD_DCtx_loadDictionary_advanced(dctx, prefix, prefixSize, ZSTD_dlm_byRef, dictContentType), "");
dctx->dictUses = ZSTD_use_once;
return 0;
}
size_t ZSTD_DCtx_refPrefix(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize)
{
return ZSTD_DCtx_refPrefix_advanced(dctx, prefix, prefixSize, ZSTD_dct_rawContent);
}
/* ZSTD_initDStream_usingDict() :
* return : expected size, aka ZSTD_startingInputLength().
* this function cannot fail */
size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize)
{
DEBUGLOG(4, "ZSTD_initDStream_usingDict");
FORWARD_IF_ERROR( ZSTD_DCtx_reset(zds, ZSTD_reset_session_only) , "");
FORWARD_IF_ERROR( ZSTD_DCtx_loadDictionary(zds, dict, dictSize) , "");
return ZSTD_startingInputLength(zds->format);
}
/* note : this variant can't fail */
size_t ZSTD_initDStream(ZSTD_DStream* zds)
{
DEBUGLOG(4, "ZSTD_initDStream");
FORWARD_IF_ERROR(ZSTD_DCtx_reset(zds, ZSTD_reset_session_only), "");
FORWARD_IF_ERROR(ZSTD_DCtx_refDDict(zds, NULL), "");
return ZSTD_startingInputLength(zds->format);
}
/* ZSTD_initDStream_usingDDict() :
* ddict will just be referenced, and must outlive decompression session
* this function cannot fail */
size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* dctx, const ZSTD_DDict* ddict)
{
DEBUGLOG(4, "ZSTD_initDStream_usingDDict");
FORWARD_IF_ERROR( ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only) , "");
FORWARD_IF_ERROR( ZSTD_DCtx_refDDict(dctx, ddict) , "");
return ZSTD_startingInputLength(dctx->format);
}
/* ZSTD_resetDStream() :
* return : expected size, aka ZSTD_startingInputLength().
* this function cannot fail */
size_t ZSTD_resetDStream(ZSTD_DStream* dctx)
{
DEBUGLOG(4, "ZSTD_resetDStream");
FORWARD_IF_ERROR(ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only), "");
return ZSTD_startingInputLength(dctx->format);
}
size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, "");
ZSTD_clearDict(dctx);
if (ddict) {
dctx->ddict = ddict;
dctx->dictUses = ZSTD_use_indefinitely;
if (dctx->refMultipleDDicts == ZSTD_rmd_refMultipleDDicts) {
if (dctx->ddictSet == NULL) {
dctx->ddictSet = ZSTD_createDDictHashSet(dctx->customMem);
if (!dctx->ddictSet) {
RETURN_ERROR(memory_allocation, "Failed to allocate memory for hash set!");
}
}
assert(!dctx->staticSize); /* Impossible: ddictSet cannot have been allocated if static dctx */
FORWARD_IF_ERROR(ZSTD_DDictHashSet_addDDict(dctx->ddictSet, ddict, dctx->customMem), "");
}
}
return 0;
}
/* ZSTD_DCtx_setMaxWindowSize() :
* note : no direct equivalence in ZSTD_DCtx_setParameter,
* since this version sets windowSize, and the other sets windowLog */
size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize)
{
ZSTD_bounds const bounds = ZSTD_dParam_getBounds(ZSTD_d_windowLogMax);
size_t const min = (size_t)1 << bounds.lowerBound;
size_t const max = (size_t)1 << bounds.upperBound;
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, "");
RETURN_ERROR_IF(maxWindowSize < min, parameter_outOfBound, "");
RETURN_ERROR_IF(maxWindowSize > max, parameter_outOfBound, "");
dctx->maxWindowSize = maxWindowSize;
return 0;
}
size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format)
{
return ZSTD_DCtx_setParameter(dctx, ZSTD_d_format, (int)format);
}
ZSTD_bounds ZSTD_dParam_getBounds(ZSTD_dParameter dParam)
{
ZSTD_bounds bounds = { 0, 0, 0 };
switch(dParam) {
case ZSTD_d_windowLogMax:
bounds.lowerBound = ZSTD_WINDOWLOG_ABSOLUTEMIN;
bounds.upperBound = ZSTD_WINDOWLOG_MAX;
return bounds;
case ZSTD_d_format:
bounds.lowerBound = (int)ZSTD_f_zstd1;
bounds.upperBound = (int)ZSTD_f_zstd1_magicless;
ZSTD_STATIC_ASSERT(ZSTD_f_zstd1 < ZSTD_f_zstd1_magicless);
return bounds;
case ZSTD_d_stableOutBuffer:
bounds.lowerBound = (int)ZSTD_bm_buffered;
bounds.upperBound = (int)ZSTD_bm_stable;
return bounds;
case ZSTD_d_forceIgnoreChecksum:
bounds.lowerBound = (int)ZSTD_d_validateChecksum;
bounds.upperBound = (int)ZSTD_d_ignoreChecksum;
return bounds;
case ZSTD_d_refMultipleDDicts:
bounds.lowerBound = (int)ZSTD_rmd_refSingleDDict;
bounds.upperBound = (int)ZSTD_rmd_refMultipleDDicts;
return bounds;
case ZSTD_d_disableHuffmanAssembly:
bounds.lowerBound = 0;
bounds.upperBound = 1;
return bounds;
case ZSTD_d_maxBlockSize:
bounds.lowerBound = ZSTD_BLOCKSIZE_MAX_MIN;
bounds.upperBound = ZSTD_BLOCKSIZE_MAX;
return bounds;
default:;
}
bounds.error = ERROR(parameter_unsupported);
return bounds;
}
/* ZSTD_dParam_withinBounds:
* @return 1 if value is within dParam bounds,
* 0 otherwise */
static int ZSTD_dParam_withinBounds(ZSTD_dParameter dParam, int value)
{
ZSTD_bounds const bounds = ZSTD_dParam_getBounds(dParam);
if (ZSTD_isError(bounds.error)) return 0;
if (value < bounds.lowerBound) return 0;
if (value > bounds.upperBound) return 0;
return 1;
}
#define CHECK_DBOUNDS(p,v) { \
RETURN_ERROR_IF(!ZSTD_dParam_withinBounds(p, v), parameter_outOfBound, ""); \
}
size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value)
{
switch (param) {
case ZSTD_d_windowLogMax:
*value = (int)ZSTD_highbit32((U32)dctx->maxWindowSize);
return 0;
case ZSTD_d_format:
*value = (int)dctx->format;
return 0;
case ZSTD_d_stableOutBuffer:
*value = (int)dctx->outBufferMode;
return 0;
case ZSTD_d_forceIgnoreChecksum:
*value = (int)dctx->forceIgnoreChecksum;
return 0;
case ZSTD_d_refMultipleDDicts:
*value = (int)dctx->refMultipleDDicts;
return 0;
case ZSTD_d_disableHuffmanAssembly:
*value = (int)dctx->disableHufAsm;
return 0;
case ZSTD_d_maxBlockSize:
*value = dctx->maxBlockSizeParam;
return 0;
default:;
}
RETURN_ERROR(parameter_unsupported, "");
}
size_t ZSTD_DCtx_setParameter(ZSTD_DCtx* dctx, ZSTD_dParameter dParam, int value)
{
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, "");
switch(dParam) {
case ZSTD_d_windowLogMax:
if (value == 0) value = ZSTD_WINDOWLOG_LIMIT_DEFAULT;
CHECK_DBOUNDS(ZSTD_d_windowLogMax, value);
dctx->maxWindowSize = ((size_t)1) << value;
return 0;
case ZSTD_d_format:
CHECK_DBOUNDS(ZSTD_d_format, value);
dctx->format = (ZSTD_format_e)value;
return 0;
case ZSTD_d_stableOutBuffer:
CHECK_DBOUNDS(ZSTD_d_stableOutBuffer, value);
dctx->outBufferMode = (ZSTD_bufferMode_e)value;
return 0;
case ZSTD_d_forceIgnoreChecksum:
CHECK_DBOUNDS(ZSTD_d_forceIgnoreChecksum, value);
dctx->forceIgnoreChecksum = (ZSTD_forceIgnoreChecksum_e)value;
return 0;
case ZSTD_d_refMultipleDDicts:
CHECK_DBOUNDS(ZSTD_d_refMultipleDDicts, value);
if (dctx->staticSize != 0) {
RETURN_ERROR(parameter_unsupported, "Static dctx does not support multiple DDicts!");
}
dctx->refMultipleDDicts = (ZSTD_refMultipleDDicts_e)value;
return 0;
case ZSTD_d_disableHuffmanAssembly:
CHECK_DBOUNDS(ZSTD_d_disableHuffmanAssembly, value);
dctx->disableHufAsm = value != 0;
return 0;
case ZSTD_d_maxBlockSize:
if (value != 0) CHECK_DBOUNDS(ZSTD_d_maxBlockSize, value);
dctx->maxBlockSizeParam = value;
return 0;
default:;
}
RETURN_ERROR(parameter_unsupported, "");
}
size_t ZSTD_DCtx_reset(ZSTD_DCtx* dctx, ZSTD_ResetDirective reset)
{
if ( (reset == ZSTD_reset_session_only)
|| (reset == ZSTD_reset_session_and_parameters) ) {
dctx->streamStage = zdss_init;
dctx->noForwardProgress = 0;
dctx->isFrameDecompression = 1;
}
if ( (reset == ZSTD_reset_parameters)
|| (reset == ZSTD_reset_session_and_parameters) ) {
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, "");
ZSTD_clearDict(dctx);
ZSTD_DCtx_resetParameters(dctx);
}
return 0;
}
size_t ZSTD_sizeof_DStream(const ZSTD_DStream* dctx)
{
return ZSTD_sizeof_DCtx(dctx);
}
static size_t ZSTD_decodingBufferSize_internal(unsigned long long windowSize, unsigned long long frameContentSize, size_t blockSizeMax)
{
size_t const blockSize = MIN((size_t)MIN(windowSize, ZSTD_BLOCKSIZE_MAX), blockSizeMax);
/* We need blockSize + WILDCOPY_OVERLENGTH worth of buffer so that if a block
* ends at windowSize + WILDCOPY_OVERLENGTH + 1 bytes, we can start writing
* the block at the beginning of the output buffer, and maintain a full window.
*
* We need another blockSize worth of buffer so that we can store split
* literals at the end of the block without overwriting the extDict window.
*/
unsigned long long const neededRBSize = windowSize + (blockSize * 2) + (WILDCOPY_OVERLENGTH * 2);
unsigned long long const neededSize = MIN(frameContentSize, neededRBSize);
size_t const minRBSize = (size_t) neededSize;
RETURN_ERROR_IF((unsigned long long)minRBSize != neededSize,
frameParameter_windowTooLarge, "");
return minRBSize;
}
size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize)
{
return ZSTD_decodingBufferSize_internal(windowSize, frameContentSize, ZSTD_BLOCKSIZE_MAX);
}
size_t ZSTD_estimateDStreamSize(size_t windowSize)
{
size_t const blockSize = MIN(windowSize, ZSTD_BLOCKSIZE_MAX);
size_t const inBuffSize = blockSize; /* no block can be larger */
size_t const outBuffSize = ZSTD_decodingBufferSize_min(windowSize, ZSTD_CONTENTSIZE_UNKNOWN);
return ZSTD_estimateDCtxSize() + inBuffSize + outBuffSize;
}
size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize)
{
U32 const windowSizeMax = 1U << ZSTD_WINDOWLOG_MAX; /* note : should be user-selectable, but requires an additional parameter (or a dctx) */
ZSTD_FrameHeader zfh;
size_t const err = ZSTD_getFrameHeader(&zfh, src, srcSize);
if (ZSTD_isError(err)) return err;
RETURN_ERROR_IF(err>0, srcSize_wrong, "");
RETURN_ERROR_IF(zfh.windowSize > windowSizeMax,
frameParameter_windowTooLarge, "");
return ZSTD_estimateDStreamSize((size_t)zfh.windowSize);
}
/* ***** Decompression ***** */
static int ZSTD_DCtx_isOverflow(ZSTD_DStream* zds, size_t const neededInBuffSize, size_t const neededOutBuffSize)
{
return (zds->inBuffSize + zds->outBuffSize) >= (neededInBuffSize + neededOutBuffSize) * ZSTD_WORKSPACETOOLARGE_FACTOR;
}
static void ZSTD_DCtx_updateOversizedDuration(ZSTD_DStream* zds, size_t const neededInBuffSize, size_t const neededOutBuffSize)
{
if (ZSTD_DCtx_isOverflow(zds, neededInBuffSize, neededOutBuffSize))
zds->oversizedDuration++;
else
zds->oversizedDuration = 0;
}
static int ZSTD_DCtx_isOversizedTooLong(ZSTD_DStream* zds)
{
return zds->oversizedDuration >= ZSTD_WORKSPACETOOLARGE_MAXDURATION;
}
/* Checks that the output buffer hasn't changed if ZSTD_obm_stable is used. */
static size_t ZSTD_checkOutBuffer(ZSTD_DStream const* zds, ZSTD_outBuffer const* output)
{
ZSTD_outBuffer const expect = zds->expectedOutBuffer;
/* No requirement when ZSTD_obm_stable is not enabled. */
if (zds->outBufferMode != ZSTD_bm_stable)
return 0;
/* Any buffer is allowed in zdss_init, this must be the same for every other call until
* the context is reset.
*/
if (zds->streamStage == zdss_init)
return 0;
/* The buffer must match our expectation exactly. */
if (expect.dst == output->dst && expect.pos == output->pos && expect.size == output->size)
return 0;
RETURN_ERROR(dstBuffer_wrong, "ZSTD_d_stableOutBuffer enabled but output differs!");
}
/* Calls ZSTD_decompressContinue() with the right parameters for ZSTD_decompressStream()
* and updates the stage and the output buffer state. This call is extracted so it can be
* used both when reading directly from the ZSTD_inBuffer, and in buffered input mode.
* NOTE: You must break after calling this function since the streamStage is modified.
*/
static size_t ZSTD_decompressContinueStream(
ZSTD_DStream* zds, char** op, char* oend,
void const* src, size_t srcSize) {
int const isSkipFrame = ZSTD_isSkipFrame(zds);
if (zds->outBufferMode == ZSTD_bm_buffered) {
size_t const dstSize = isSkipFrame ? 0 : zds->outBuffSize - zds->outStart;
size_t const decodedSize = ZSTD_decompressContinue(zds,
zds->outBuff + zds->outStart, dstSize, src, srcSize);
FORWARD_IF_ERROR(decodedSize, "");
if (!decodedSize && !isSkipFrame) {
zds->streamStage = zdss_read;
} else {
zds->outEnd = zds->outStart + decodedSize;
zds->streamStage = zdss_flush;
}
} else {
/* Write directly into the output buffer */
size_t const dstSize = isSkipFrame ? 0 : (size_t)(oend - *op);
size_t const decodedSize = ZSTD_decompressContinue(zds, *op, dstSize, src, srcSize);
FORWARD_IF_ERROR(decodedSize, "");
*op += decodedSize;
/* Flushing is not needed. */
zds->streamStage = zdss_read;
assert(*op <= oend);
assert(zds->outBufferMode == ZSTD_bm_stable);
}
return 0;
}
size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inBuffer* input)
{
const char* const src = (const char*)input->src;
const char* const istart = input->pos != 0 ? src + input->pos : src;
const char* const iend = input->size != 0 ? src + input->size : src;
const char* ip = istart;
char* const dst = (char*)output->dst;
char* const ostart = output->pos != 0 ? dst + output->pos : dst;
char* const oend = output->size != 0 ? dst + output->size : dst;
char* op = ostart;
U32 someMoreWork = 1;
DEBUGLOG(5, "ZSTD_decompressStream");
assert(zds != NULL);
RETURN_ERROR_IF(
input->pos > input->size,
srcSize_wrong,
"forbidden. in: pos: %u vs size: %u",
(U32)input->pos, (U32)input->size);
RETURN_ERROR_IF(
output->pos > output->size,
dstSize_tooSmall,
"forbidden. out: pos: %u vs size: %u",
(U32)output->pos, (U32)output->size);
DEBUGLOG(5, "input size : %u", (U32)(input->size - input->pos));
FORWARD_IF_ERROR(ZSTD_checkOutBuffer(zds, output), "");
while (someMoreWork) {
switch(zds->streamStage)
{
case zdss_init :
DEBUGLOG(5, "stage zdss_init => transparent reset ");
zds->streamStage = zdss_loadHeader;
zds->lhSize = zds->inPos = zds->outStart = zds->outEnd = 0;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
zds->legacyVersion = 0;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
return ZSTD_findFrameCompressedSizeLegacy(src, srcSize);
#else
(void)src;
(void)srcSize;
return ERROR(prefix_unknown);
#endif
zds->hostageByte = 0;
zds->expectedOutBuffer = *output;
ZSTD_FALLTHROUGH;
case zdss_loadHeader :
DEBUGLOG(5, "stage zdss_loadHeader (srcSize : %u)", (U32)(iend - ip));
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
if (zds->legacyVersion) {
RETURN_ERROR_IF(zds->staticSize, memory_allocation,
"legacy support is incompatible with static dctx");
{ size_t const hint = ZSTD_decompressLegacyStream(zds->legacyContext, zds->legacyVersion, output, input);
if (hint==0) zds->streamStage = zdss_init;
return hint;
} }
#endif
{ size_t const hSize = ZSTD_getFrameHeader_advanced(&zds->fParams, zds->headerBuffer, zds->lhSize, zds->format);
if (zds->refMultipleDDicts && zds->ddictSet) {
ZSTD_DCtx_selectFrameDDict(zds);
}
if (ZSTD_isError(hSize)) {
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
U32 const legacyVersion = ZSTD_isLegacy(istart, iend-istart);
if (legacyVersion) {
ZSTD_DDict const* const ddict = ZSTD_getDDict(zds);
const void* const dict = ddict ? ZSTD_DDict_dictContent(ddict) : NULL;
size_t const dictSize = ddict ? ZSTD_DDict_dictSize(ddict) : 0;
DEBUGLOG(5, "ZSTD_decompressStream: detected legacy version v0.%u", legacyVersion);
RETURN_ERROR_IF(zds->staticSize, memory_allocation,
"legacy support is incompatible with static dctx");
FORWARD_IF_ERROR(ZSTD_initLegacyStream(&zds->legacyContext,
zds->previousLegacyVersion, legacyVersion,
dict, dictSize), "");
zds->legacyVersion = zds->previousLegacyVersion = legacyVersion;
{ size_t const hint = ZSTD_decompressLegacyStream(zds->legacyContext, legacyVersion, output, input);
if (hint==0) zds->streamStage = zdss_init; /* or stay in stage zdss_loadHeader */
return hint;
} }
#endif
return hSize; /* error */
}
if (hSize != 0) { /* need more input */
size_t const toLoad = hSize - zds->lhSize; /* if hSize!=0, hSize > zds->lhSize */
size_t const remainingInput = (size_t)(iend-ip);
assert(iend >= ip);
if (toLoad > remainingInput) { /* not enough input to load full header */
if (remainingInput > 0) {
ZSTD_memcpy(zds->headerBuffer + zds->lhSize, ip, remainingInput);
zds->lhSize += remainingInput;
}
input->pos = input->size;
/* check first few bytes */
FORWARD_IF_ERROR(
ZSTD_getFrameHeader_advanced(&zds->fParams, zds->headerBuffer, zds->lhSize, zds->format),
"First few bytes detected incorrect" );
/* return hint input size */
return (MAX((size_t)ZSTD_FRAMEHEADERSIZE_MIN(zds->format), hSize) - zds->lhSize) + ZSTD_blockHeaderSize; /* remaining header bytes + next block header */
}
assert(ip != NULL);
ZSTD_memcpy(zds->headerBuffer + zds->lhSize, ip, toLoad); zds->lhSize = hSize; ip += toLoad;
break;
} }
/* check for single-pass mode opportunity */
if (zds->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN
&& zds->fParams.frameType != ZSTD_skippableFrame
&& (U64)(size_t)(oend-op) >= zds->fParams.frameContentSize) {
size_t const cSize = ZSTD_findFrameCompressedSize_advanced(istart, (size_t)(iend-istart), zds->format);
if (cSize <= (size_t)(iend-istart)) {
/* shortcut : using single-pass mode */
size_t const decompressedSize = ZSTD_decompress_usingDDict(zds, op, (size_t)(oend-op), istart, cSize, ZSTD_getDDict(zds));
if (ZSTD_isError(decompressedSize)) return decompressedSize;
DEBUGLOG(4, "shortcut to single-pass ZSTD_decompress_usingDDict()");
assert(istart != NULL);
ip = istart + cSize;
op = op ? op + decompressedSize : op; /* can occur if frameContentSize = 0 (empty frame) */
zds->expected = 0;
zds->streamStage = zdss_init;
someMoreWork = 0;
break;
} }
/* Check output buffer is large enough for ZSTD_odm_stable. */
if (zds->outBufferMode == ZSTD_bm_stable
&& zds->fParams.frameType != ZSTD_skippableFrame
&& zds->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN
&& (U64)(size_t)(oend-op) < zds->fParams.frameContentSize) {
RETURN_ERROR(dstSize_tooSmall, "ZSTD_obm_stable passed but ZSTD_outBuffer is too small");
}
/* Consume header (see ZSTDds_decodeFrameHeader) */
DEBUGLOG(4, "Consume header");
FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDDict(zds, ZSTD_getDDict(zds)), "");
if (zds->format == ZSTD_f_zstd1
&& (MEM_readLE32(zds->headerBuffer) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { /* skippable frame */
zds->expected = MEM_readLE32(zds->headerBuffer + ZSTD_FRAMEIDSIZE);
zds->stage = ZSTDds_skipFrame;
} else {
FORWARD_IF_ERROR(ZSTD_decodeFrameHeader(zds, zds->headerBuffer, zds->lhSize), "");
zds->expected = ZSTD_blockHeaderSize;
zds->stage = ZSTDds_decodeBlockHeader;
}
/* control buffer memory usage */
DEBUGLOG(4, "Control max memory usage (%u KB <= max %u KB)",
(U32)(zds->fParams.windowSize >>10),
(U32)(zds->maxWindowSize >> 10) );
zds->fParams.windowSize = MAX(zds->fParams.windowSize, 1U << ZSTD_WINDOWLOG_ABSOLUTEMIN);
RETURN_ERROR_IF(zds->fParams.windowSize > zds->maxWindowSize,
frameParameter_windowTooLarge, "");
if (zds->maxBlockSizeParam != 0)
zds->fParams.blockSizeMax = MIN(zds->fParams.blockSizeMax, (unsigned)zds->maxBlockSizeParam);
/* Adapt buffer sizes to frame header instructions */
{ size_t const neededInBuffSize = MAX(zds->fParams.blockSizeMax, 4 /* frame checksum */);
size_t const neededOutBuffSize = zds->outBufferMode == ZSTD_bm_buffered
? ZSTD_decodingBufferSize_internal(zds->fParams.windowSize, zds->fParams.frameContentSize, zds->fParams.blockSizeMax)
: 0;
ZSTD_DCtx_updateOversizedDuration(zds, neededInBuffSize, neededOutBuffSize);
{ int const tooSmall = (zds->inBuffSize < neededInBuffSize) || (zds->outBuffSize < neededOutBuffSize);
int const tooLarge = ZSTD_DCtx_isOversizedTooLong(zds);
if (tooSmall || tooLarge) {
size_t const bufferSize = neededInBuffSize + neededOutBuffSize;
DEBUGLOG(4, "inBuff : from %u to %u",
(U32)zds->inBuffSize, (U32)neededInBuffSize);
DEBUGLOG(4, "outBuff : from %u to %u",
(U32)zds->outBuffSize, (U32)neededOutBuffSize);
if (zds->staticSize) { /* static DCtx */
DEBUGLOG(4, "staticSize : %u", (U32)zds->staticSize);
assert(zds->staticSize >= sizeof(ZSTD_DCtx)); /* controlled at init */
RETURN_ERROR_IF(
bufferSize > zds->staticSize - sizeof(ZSTD_DCtx),
memory_allocation, "");
} else {
ZSTD_customFree(zds->inBuff, zds->customMem);
zds->inBuffSize = 0;
zds->outBuffSize = 0;
zds->inBuff = (char*)ZSTD_customMalloc(bufferSize, zds->customMem);
RETURN_ERROR_IF(zds->inBuff == NULL, memory_allocation, "");
}
zds->inBuffSize = neededInBuffSize;
zds->outBuff = zds->inBuff + zds->inBuffSize;
zds->outBuffSize = neededOutBuffSize;
} } }
zds->streamStage = zdss_read;
ZSTD_FALLTHROUGH;
case zdss_read:
DEBUGLOG(5, "stage zdss_read");
{ size_t const neededInSize = ZSTD_nextSrcSizeToDecompressWithInputSize(zds, (size_t)(iend - ip));
DEBUGLOG(5, "neededInSize = %u", (U32)neededInSize);
if (neededInSize==0) { /* end of frame */
zds->streamStage = zdss_init;
someMoreWork = 0;
break;
}
if ((size_t)(iend-ip) >= neededInSize) { /* decode directly from src */
FORWARD_IF_ERROR(ZSTD_decompressContinueStream(zds, &op, oend, ip, neededInSize), "");
assert(ip != NULL);
ip += neededInSize;
/* Function modifies the stage so we must break */
break;
} }
if (ip==iend) { someMoreWork = 0; break; } /* no more input */
zds->streamStage = zdss_load;
ZSTD_FALLTHROUGH;
case zdss_load:
{ size_t const neededInSize = ZSTD_nextSrcSizeToDecompress(zds);
size_t const toLoad = neededInSize - zds->inPos;
int const isSkipFrame = ZSTD_isSkipFrame(zds);
size_t loadedSize;
/* At this point we shouldn't be decompressing a block that we can stream. */
assert(neededInSize == ZSTD_nextSrcSizeToDecompressWithInputSize(zds, (size_t)(iend - ip)));
if (isSkipFrame) {
loadedSize = MIN(toLoad, (size_t)(iend-ip));
} else {
RETURN_ERROR_IF(toLoad > zds->inBuffSize - zds->inPos,
corruption_detected,
"should never happen");
loadedSize = ZSTD_limitCopy(zds->inBuff + zds->inPos, toLoad, ip, (size_t)(iend-ip));
}
if (loadedSize != 0) {
/* ip may be NULL */
ip += loadedSize;
zds->inPos += loadedSize;
}
if (loadedSize < toLoad) { someMoreWork = 0; break; } /* not enough input, wait for more */
/* decode loaded input */
zds->inPos = 0; /* input is consumed */
FORWARD_IF_ERROR(ZSTD_decompressContinueStream(zds, &op, oend, zds->inBuff, neededInSize), "");
/* Function modifies the stage so we must break */
break;
}
case zdss_flush:
{
size_t const toFlushSize = zds->outEnd - zds->outStart;
size_t const flushedSize = ZSTD_limitCopy(op, (size_t)(oend-op), zds->outBuff + zds->outStart, toFlushSize);
op = op ? op + flushedSize : op;
zds->outStart += flushedSize;
if (flushedSize == toFlushSize) { /* flush completed */
zds->streamStage = zdss_read;
if ( (zds->outBuffSize < zds->fParams.frameContentSize)
&& (zds->outStart + zds->fParams.blockSizeMax > zds->outBuffSize) ) {
DEBUGLOG(5, "restart filling outBuff from beginning (left:%i, needed:%u)",
(int)(zds->outBuffSize - zds->outStart),
(U32)zds->fParams.blockSizeMax);
zds->outStart = zds->outEnd = 0;
}
break;
} }
/* cannot complete flush */
someMoreWork = 0;
break;
default:
assert(0); /* impossible */
RETURN_ERROR(GENERIC, "impossible to reach"); /* some compilers require default to do something */
} }
/* result */
input->pos = (size_t)(ip - (const char*)(input->src));
output->pos = (size_t)(op - (char*)(output->dst));
/* Update the expected output buffer for ZSTD_obm_stable. */
zds->expectedOutBuffer = *output;
if ((ip==istart) && (op==ostart)) { /* no forward progress */
zds->noForwardProgress ++;
if (zds->noForwardProgress >= ZSTD_NO_FORWARD_PROGRESS_MAX) {
RETURN_ERROR_IF(op==oend, noForwardProgress_destFull, "");
RETURN_ERROR_IF(ip==iend, noForwardProgress_inputEmpty, "");
assert(0);
}
} else {
zds->noForwardProgress = 0;
}
{ size_t nextSrcSizeHint = ZSTD_nextSrcSizeToDecompress(zds);
if (!nextSrcSizeHint) { /* frame fully decoded */
if (zds->outEnd == zds->outStart) { /* output fully flushed */
if (zds->hostageByte) {
if (input->pos >= input->size) {
/* can't release hostage (not present) */
zds->streamStage = zdss_read;
return 1;
}
input->pos++; /* release hostage */
} /* zds->hostageByte */
return 0;
} /* zds->outEnd == zds->outStart */
if (!zds->hostageByte) { /* output not fully flushed; keep last byte as hostage; will be released when all output is flushed */
input->pos--; /* note : pos > 0, otherwise, impossible to finish reading last block */
zds->hostageByte=1;
}
return 1;
} /* nextSrcSizeHint==0 */
nextSrcSizeHint += ZSTD_blockHeaderSize * (ZSTD_nextInputType(zds) == ZSTDnit_block); /* preload header of next block */
assert(zds->inPos <= nextSrcSizeHint);
nextSrcSizeHint -= zds->inPos; /* part already loaded*/
return nextSrcSizeHint;
}
}
size_t ZSTD_decompressStream_simpleArgs (
ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity, size_t* dstPos,
const void* src, size_t srcSize, size_t* srcPos)
size_t ZSTD_rust_legacy_frame_size_info(const void* src, size_t srcSize,
size_t* compressedSize,
unsigned long long* decompressedBound,
size_t* nbBlocks)
{
ZSTD_outBuffer output;
ZSTD_inBuffer input;
output.dst = dst;
output.size = dstCapacity;
output.pos = *dstPos;
input.src = src;
input.size = srcSize;
input.pos = *srcPos;
{ size_t const cErr = ZSTD_decompressStream(dctx, &output, &input);
*dstPos = output.pos;
*srcPos = input.pos;
return cErr;
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
ZSTD_frameSizeInfo const info = ZSTD_findFrameSizeInfoLegacy(src, srcSize);
*compressedSize = info.compressedSize;
*decompressedBound = info.decompressedBound;
*nbBlocks = info.decompressedBound == ZSTD_CONTENTSIZE_ERROR
? 0 : (size_t)(info.decompressedBound / ZSTD_BLOCKSIZE_MAX);
return ZSTD_isError(info.compressedSize) ? info.compressedSize : 0;
#else
(void)src;
(void)srcSize;
*compressedSize = ERROR(prefix_unknown);
*decompressedBound = ZSTD_CONTENTSIZE_ERROR;
*nbBlocks = 0;
return ERROR(prefix_unknown);
#endif
}
size_t ZSTD_rust_legacy_decompress(void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const void* dict, size_t dictSize)
{
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
return ZSTD_decompressLegacy(dst, dstCapacity, src, srcSize, dict, dictSize);
#else
(void)dst;
(void)dstCapacity;
(void)src;
(void)srcSize;
(void)dict;
(void)dictSize;
return ERROR(prefix_unknown);
#endif
}
size_t ZSTD_rust_legacy_decompress_stream(ZSTD_DCtx* dctx,
ZSTD_outBuffer* output,
ZSTD_inBuffer* input,
const void* dict, size_t dictSize)
{
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
size_t hint;
if (dctx->legacyVersion) {
hint = ZSTD_decompressLegacyStream(dctx->legacyContext,
dctx->legacyVersion, output, input);
if (hint == 0) dctx->streamStage = zdss_init;
return hint;
}
{
const char* const istart = input->pos != 0
? (const char*)input->src + input->pos
: (const char*)input->src;
size_t const inputSize = input->size - input->pos;
U32 const legacyVersion = ZSTD_isLegacy(istart, inputSize);
if (!legacyVersion) return ERROR(prefix_unknown);
if (dctx->staticSize) return ERROR(memory_allocation);
FORWARD_IF_ERROR(ZSTD_initLegacyStream(&dctx->legacyContext,
dctx->previousLegacyVersion,
legacyVersion, dict, dictSize), "");
dctx->legacyVersion = dctx->previousLegacyVersion = legacyVersion;
hint = ZSTD_decompressLegacyStream(dctx->legacyContext, legacyVersion,
output, input);
if (hint == 0) dctx->streamStage = zdss_init;
return hint;
}
#else
(void)dctx;
(void)output;
(void)input;
(void)dict;
(void)dictSize;
return ERROR(prefix_unknown);
#endif
}
void ZSTD_rust_legacy_free_stream(ZSTD_DCtx* dctx)
{
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
if (dctx->legacyContext) {
ZSTD_freeLegacyStreamContext(dctx->legacyContext,
dctx->previousLegacyVersion);
dctx->legacyContext = NULL;
}
#else
(void)dctx;
#endif
}