facebook/zstd
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Change 1.3.4 bugfix to be cross-compatible with superblocks and normal compression

Browse Source
This commit is contained in:
senhuang42 2021-01-11 08:30:16 -05:00 committed by Sen Huang
parent eb1ee8686d
commit 0633bf17c3
3 changed files with 25 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
lib/compress/zstd_compress.c
View File

@ -2208,7 +2208,7 @@ static int ZSTD_useBlockSplitting(const ZSTD_CCtx_params* cctxParams)
*/ */
MEM_STATIC size_t MEM_STATIC size_t
ZSTD_buildSequencesStatistics(seqStore_t* seqStorePtr, size_t nbSeq, ZSTD_buildSequencesStatistics(seqStore_t* seqStorePtr, size_t nbSeq,
U32* LLtype, U32* Offtype, U32* MLtype, BYTE** lastNCount, U32* LLtype, U32* Offtype, U32* MLtype, size_t* lastCountSize,
const ZSTD_fseCTables_t* prevEntropy, ZSTD_fseCTables_t* nextEntropy, const ZSTD_fseCTables_t* prevEntropy, ZSTD_fseCTables_t* nextEntropy,
BYTE* dst, const BYTE* const dstEnd, BYTE* dst, const BYTE* const dstEnd,
ZSTD_strategy strategy, ZSTD_strategy strategy,
@ -2252,7 +2252,7 @@ ZSTD_buildSequencesStatistics(seqStore_t* seqStorePtr, size_t nbSeq,
entropyWorkspace, entropyWkspSize); entropyWorkspace, entropyWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed"); FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed");
if (*LLtype == set_compressed) if (*LLtype == set_compressed)
*lastNCount = op; *lastCountSize = countSize;
op += countSize; op += countSize;
assert(op <= oend); assert(op <= oend);
} } } }
@ -2280,7 +2280,7 @@ ZSTD_buildSequencesStatistics(seqStore_t* seqStorePtr, size_t nbSeq,
entropyWorkspace, entropyWkspSize); entropyWorkspace, entropyWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed"); FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed");
if (*Offtype == set_compressed) if (*Offtype == set_compressed)
*lastNCount = op; *lastCountSize = countSize;
op += countSize; op += countSize;
assert(op <= oend); assert(op <= oend);
} } } }
@ -2306,7 +2306,7 @@ ZSTD_buildSequencesStatistics(seqStore_t* seqStorePtr, size_t nbSeq,
entropyWorkspace, entropyWkspSize); entropyWorkspace, entropyWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed"); FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed");
if (*MLtype == set_compressed) if (*MLtype == set_compressed)
*lastNCount = op; *lastCountSize = countSize;
op += countSize; op += countSize;
assert(op <= oend); assert(op <= oend);
} } } }
@ -2339,8 +2339,7 @@ ZSTD_entropyCompressSequences_internal(seqStore_t* seqStorePtr,
BYTE* const ostart = (BYTE*)dst; BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity; BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart; BYTE* op = ostart;
BYTE* lastNCount = NULL; size_t lastCountSize = 0;
BYTE* seqHead;
entropyWorkspace = count + (MaxSeq + 1); entropyWorkspace = count + (MaxSeq + 1);
entropyWkspSize -= (MaxSeq + 1) * sizeof(*count); entropyWkspSize -= (MaxSeq + 1) * sizeof(*count);
@ -2385,15 +2384,14 @@ ZSTD_entropyCompressSequences_internal(seqStore_t* seqStorePtr,
ZSTD_memcpy(&nextEntropy->fse, &prevEntropy->fse, sizeof(prevEntropy->fse)); ZSTD_memcpy(&nextEntropy->fse, &prevEntropy->fse, sizeof(prevEntropy->fse));
return (size_t)(op - ostart); return (size_t)(op - ostart);
} }
{ {
U32 LLtype; U32 LLtype;
U32 Offtype; U32 Offtype;
U32 MLtype; U32 MLtype;
seqHead = op++; BYTE* seqHead = op++;
/* build stats for sequences */ /* build stats for sequences */
entropyStatisticsSize = ZSTD_buildSequencesStatistics(seqStorePtr, nbSeq, entropyStatisticsSize = ZSTD_buildSequencesStatistics(seqStorePtr, nbSeq,
&LLtype, &Offtype, &MLtype, &lastNCount, &LLtype, &Offtype, &MLtype, &lastCountSize,
&prevEntropy->fse, &nextEntropy->fse, &prevEntropy->fse, &nextEntropy->fse,
op, oend, op, oend,
strategy, strategy,
@ -2421,9 +2419,9 @@ ZSTD_entropyCompressSequences_internal(seqStore_t* seqStorePtr,
* In this exceedingly rare case, we will simply emit an uncompressed * In this exceedingly rare case, we will simply emit an uncompressed
* block, since it isn't worth optimizing. * block, since it isn't worth optimizing.
*/ */
if (lastNCount && (op - lastNCount) < 4) { if (lastCountSize && (lastCountSize + bitstreamSize) < 4) {
/* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */ /* lastCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
assert(op - lastNCount == 3); assert(lastCountSize + bitstreamSize == 3);
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by " DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by "
"emitting an uncompressed block."); "emitting an uncompressed block.");
return 0; return 0;
@ -2886,20 +2884,20 @@ static size_t ZSTD_buildBlockEntropyStats_sequences(seqStore_t* seqStorePtr,
BYTE* const ostart = fseMetadata->fseTablesBuffer; BYTE* const ostart = fseMetadata->fseTablesBuffer;
BYTE* const oend = ostart + sizeof(fseMetadata->fseTablesBuffer); BYTE* const oend = ostart + sizeof(fseMetadata->fseTablesBuffer);
BYTE* op = ostart; BYTE* op = ostart;
BYTE* lastNCount = NULL;
size_t hSize; size_t hSize;
U32 LLtype;
U32 Offtype; /* ZSTD_buildSequencesStatistics() is guaranteed to overwrite these values */
U32 MLtype; U32 LLtype = set_basic;
U32 Offtype = set_basic;
U32 MLtype = set_basic;
DEBUGLOG(5, "ZSTD_buildBlockEntropyStats_sequences (nbSeq=%zu)", nbSeq); DEBUGLOG(5, "ZSTD_buildBlockEntropyStats_sequences (nbSeq=%zu)", nbSeq);
ZSTD_memset(workspace, 0, wkspSize); ZSTD_memset(workspace, 0, wkspSize);
hSize = ZSTD_buildSequencesStatistics(seqStorePtr, nbSeq, hSize = ZSTD_buildSequencesStatistics(seqStorePtr, nbSeq,
&LLtype, &Offtype, &MLtype, &lastNCount, &LLtype, &Offtype, &MLtype, &fseMetadata->lastCountSize,
prevEntropy, nextEntropy, op, oend, prevEntropy, nextEntropy, op, oend,
strategy, strategy,
workspace, wkspSize); workspace, wkspSize);
fseMetadata->lastNCount = lastNCount;
fseMetadata->llType = (symbolEncodingType_e) LLtype; fseMetadata->llType = (symbolEncodingType_e) LLtype;
fseMetadata->ofType = (symbolEncodingType_e) Offtype; fseMetadata->ofType = (symbolEncodingType_e) Offtype;
fseMetadata->mlType = (symbolEncodingType_e) MLtype; fseMetadata->mlType = (symbolEncodingType_e) MLtype;
@ -3104,8 +3102,8 @@ static size_t ZSTD_countSeqStoreMatchBytes(const seqStore_t* seqStore) {
static void ZSTD_deriveSeqStoreChunk(seqStore_t* resultSeqStore, const seqStore_t* originalSeqStore, static void ZSTD_deriveSeqStoreChunk(seqStore_t* resultSeqStore, const seqStore_t* originalSeqStore,
size_t startIdx, size_t endIdx) { size_t startIdx, size_t endIdx) {
BYTE* const litEnd = originalSeqStore->lit; BYTE* const litEnd = originalSeqStore->lit;
U32 literalsBytes; size_t literalsBytes;
U32 literalsBytesPreceding = 0; size_t literalsBytesPreceding = 0;
*resultSeqStore = *originalSeqStore; *resultSeqStore = *originalSeqStore;
if (startIdx > 0) { if (startIdx > 0) {
@ -3118,7 +3116,7 @@ static void ZSTD_deriveSeqStoreChunk(seqStore_t* resultSeqStore, const seqStore_
if (originalSeqStore->longLengthPos < startIdx || originalSeqStore->longLengthPos > endIdx) { if (originalSeqStore->longLengthPos < startIdx || originalSeqStore->longLengthPos > endIdx) {
resultSeqStore->longLengthID = 0; resultSeqStore->longLengthID = 0;
} else { } else {
resultSeqStore->longLengthPos -= startIdx; resultSeqStore->longLengthPos -= (U32)startIdx;
} }
} }
resultSeqStore->sequencesStart = originalSeqStore->sequencesStart + startIdx; resultSeqStore->sequencesStart = originalSeqStore->sequencesStart + startIdx;
@ -3240,7 +3238,7 @@ static void ZSTD_deriveBlockSplitsHelper(seqStoreSplits* splits, size_t startIdx
} }
if (estimatedFirstHalfSize + estimatedSecondHalfSize < estimatedOriginalSize) { if (estimatedFirstHalfSize + estimatedSecondHalfSize < estimatedOriginalSize) {
ZSTD_deriveBlockSplitsHelper(splits, startIdx, midIdx, zc, origSeqStore); ZSTD_deriveBlockSplitsHelper(splits, startIdx, midIdx, zc, origSeqStore);
splits->splitLocations[splits->idx] = midIdx; splits->splitLocations[splits->idx] = (U32)midIdx;
splits->idx++; splits->idx++;
ZSTD_deriveBlockSplitsHelper(splits, midIdx, endIdx, zc, origSeqStore); ZSTD_deriveBlockSplitsHelper(splits, midIdx, endIdx, zc, origSeqStore);
} }
@ -3322,21 +3320,20 @@ static size_t ZSTD_compressBlock_splitBlock(ZSTD_CCtx* zc,
const void* src, size_t srcSize, U32 lastBlock) { const void* src, size_t srcSize, U32 lastBlock) {
const BYTE* ip = (const BYTE*)src; const BYTE* ip = (const BYTE*)src;
BYTE* op = (BYTE*)dst; BYTE* op = (BYTE*)dst;
size_t nbSeq; U32 nbSeq;
size_t cSize; size_t cSize;
DEBUGLOG(4, "ZSTD_compressBlock_splitBlock"); DEBUGLOG(4, "ZSTD_compressBlock_splitBlock");
{ const size_t bss = ZSTD_buildSeqStore(zc, src, srcSize); { const size_t bss = ZSTD_buildSeqStore(zc, src, srcSize);
FORWARD_IF_ERROR(bss, "ZSTD_buildSeqStore failed"); FORWARD_IF_ERROR(bss, "ZSTD_buildSeqStore failed");
if (bss == ZSTDbss_noCompress) { if (bss == ZSTDbss_noCompress) {
cSize = 0;
if (zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid) if (zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid)
zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check; zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check;
cSize = ZSTD_noCompressBlock(op, dstCapacity, ip, srcSize, lastBlock); cSize = ZSTD_noCompressBlock(op, dstCapacity, ip, srcSize, lastBlock);
FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed"); FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed");
return cSize; return cSize;
} }
nbSeq = (size_t)(zc->seqStore.sequences - zc->seqStore.sequencesStart); nbSeq = (U32)(zc->seqStore.sequences - zc->seqStore.sequencesStart);
} }
assert(zc->appliedParams.splitBlocks == 1); assert(zc->appliedParams.splitBlocks == 1);

2
lib/compress/zstd_compress_internal.h
View File

@ -107,7 +107,7 @@ typedef struct {
symbolEncodingType_e mlType; symbolEncodingType_e mlType;
BYTE fseTablesBuffer[ZSTD_MAX_FSE_HEADERS_SIZE]; BYTE fseTablesBuffer[ZSTD_MAX_FSE_HEADERS_SIZE];
size_t fseTablesSize; size_t fseTablesSize;
BYTE* lastNCount; /* This is to account for bug in 1.3.4. More detail in ZSTD_compressSubBlock_sequences() */ size_t lastCountSize; /* This is to account for bug in 1.3.4. More detail in ZSTD_entropyCompressSequences_internal() */
} ZSTD_fseCTablesMetadata_t; } ZSTD_fseCTablesMetadata_t;
typedef struct { typedef struct {

4
lib/compress/zstd_compress_superblock.c
View File

@ -500,9 +500,9 @@ static size_t ZSTD_compressSubBlock_sequences(const ZSTD_fseCTables_t* fseTables
* block, since it isn't worth optimizing. * block, since it isn't worth optimizing.
*/ */
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
if (writeEntropy && fseMetadata->lastNCount && (op - fseMetadata->lastNCount) < 4) { if (writeEntropy && fseMetadata->lastCountSize && (bitstreamSize + fseMetadata->lastCountSize) < 4) {
/* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */ /* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
assert(op - fseMetadata->lastNCount == 3); assert(bitstreamSize + fseMetadata->lastCountSize == 3);
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by " DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by "
"emitting an uncompressed block."); "emitting an uncompressed block.");
return 0; return 0;