/*-*************************************
* Price functions for optimal parser
***************************************/
static void ZSTD_setLog2Prices(optState_t* optPtr)
{
optPtr->log2matchLengthSum = ZSTD_highbit32(optPtr->matchLengthSum+1);
optPtr->log2litLengthSum = ZSTD_highbit32(optPtr->litLengthSum+1);
optPtr->log2litSum = ZSTD_highbit32(optPtr->litSum+1);
optPtr->log2offCodeSum = ZSTD_highbit32(optPtr->offCodeSum+1);
optPtr->factor = 1 + ((optPtr->litSum>>5) / optPtr->litLengthSum) + ((optPtr->litSum<<1) / (optPtr->litSum + optPtr->matchSum));
}
ZSTDMT_CCtx *ZSTDMT_createCCtx(unsigned nbThreads)
{
ZSTDMT_CCtx* cctx;
U32 const minNbJobs = nbThreads + 2;
U32 const nbJobsLog2 = ZSTD_highbit32(minNbJobs) + 1;
U32 const nbJobs = 1 << nbJobsLog2;
DEBUGLOG(5, "nbThreads : %u ; minNbJobs : %u ; nbJobsLog2 : %u ; nbJobs : %u \n",
nbThreads, minNbJobs, nbJobsLog2, nbJobs);
if ((nbThreads < 1) | (nbThreads > ZSTDMT_NBTHREADS_MAX)) return NULL;
cctx = (ZSTDMT_CCtx*) calloc(1, sizeof(ZSTDMT_CCtx) + nbJobs*sizeof(ZSTDMT_jobDescription));
if (!cctx) return NULL;
cctx->nbThreads = nbThreads;
cctx->jobIDMask = nbJobs - 1;
cctx->allJobsCompleted = 1;
cctx->sectionSize = 0;
cctx->overlapRLog = 3;
cctx->factory = POOL_create(nbThreads, 1);
cctx->buffPool = ZSTDMT_createBufferPool(nbThreads);
cctx->cctxPool = ZSTDMT_createCCtxPool(nbThreads);
if (!cctx->factory | !cctx->buffPool | !cctx->cctxPool) { /* one object was not created */
ZSTDMT_freeCCtx(cctx);
return NULL;
}
if (nbThreads==1) {
cctx->cstream = ZSTD_createCStream();
if (!cctx->cstream) {
ZSTDMT_freeCCtx(cctx); return NULL;
} }
pthread_mutex_init(&cctx->jobCompleted_mutex, NULL); /* Todo : check init function return */
pthread_cond_init(&cctx->jobCompleted_cond, NULL);
DEBUGLOG(4, "mt_cctx created, for %u threads \n", nbThreads);
return cctx;
}
/**
* Initializes a map of the given size.
* Returns 1 on success and 0 on failure.
* The map must be destroyed with COVER_map_destroy().
* The map is only guaranteed to be large enough to hold size elements.
*/
static int COVER_map_init(COVER_map_t *map, U32 size) {
map->sizeLog = ZSTD_highbit32(size) + 2;
map->size = (U32)1 << map->sizeLog;
map->sizeMask = map->size - 1;
map->data = (COVER_map_pair_t *)malloc(map->size * sizeof(COVER_map_pair_t));
if (!map->data) {
map->sizeLog = 0;
map->size = 0;
return 0;
}
COVER_map_clear(map);
return 1;
}
/* *******************************
* Common parser - lazy strategy
*********************************/
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore,
U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize,
const U32 searchMethod, const U32 depth,
ZSTD_dictMode_e const dictMode)
{
const BYTE* const istart = (const BYTE*)src;
const BYTE* ip = istart;
const BYTE* anchor = istart;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - 8;
const BYTE* const base = ms->window.base;
const U32 prefixLowestIndex = ms->window.dictLimit;
const BYTE* const prefixLowest = base + prefixLowestIndex;
typedef size_t (*searchMax_f)(
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
searchMax_f const searchMax = dictMode == ZSTD_dictMatchState ?
(searchMethod ? ZSTD_BtFindBestMatch_dictMatchState_selectMLS : ZSTD_HcFindBestMatch_dictMatchState_selectMLS) :
(searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS);
U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0;
const ZSTD_matchState_t* const dms = ms->dictMatchState;
const U32 dictLowestIndex = dictMode == ZSTD_dictMatchState ?
dms->window.dictLimit : 0;
const BYTE* const dictBase = dictMode == ZSTD_dictMatchState ?
dms->window.base : NULL;
const BYTE* const dictLowest = dictMode == ZSTD_dictMatchState ?
dictBase + dictLowestIndex : NULL;
const BYTE* const dictEnd = dictMode == ZSTD_dictMatchState ?
dms->window.nextSrc : NULL;
const U32 dictIndexDelta = dictMode == ZSTD_dictMatchState ?
prefixLowestIndex - (U32)(dictEnd - dictBase) :
0;
const U32 dictAndPrefixLength = (U32)(ip - prefixLowest + dictEnd - dictLowest);
/* init */
ip += (dictAndPrefixLength == 0);
ms->nextToUpdate3 = ms->nextToUpdate;
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixLowest);
if (offset_2 > maxRep) savedOffset = offset_2, offset_2 = 0;
if (offset_1 > maxRep) savedOffset = offset_1, offset_1 = 0;
}
if (dictMode == ZSTD_dictMatchState) {
/* dictMatchState repCode checks don't currently handle repCode == 0
* disabling. */
assert(offset_1 <= dictAndPrefixLength);
assert(offset_2 <= dictAndPrefixLength);
}
/* Match Loop */
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
const BYTE* start=ip+1;
/* check repCode */
if (dictMode == ZSTD_dictMatchState) {
const U32 repIndex = (U32)(ip - base) + 1 - offset_1;
const BYTE* repMatch = (dictMode == ZSTD_dictMatchState
&& repIndex < prefixLowestIndex) ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */)
&& (MEM_read32(repMatch) == MEM_read32(ip+1)) ) {
const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend;
matchLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4;
if (depth==0) goto _storeSequence;
}
}
if ( dictMode == ZSTD_noDict
&& ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1)))) {
matchLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4;
if (depth==0) goto _storeSequence;
}
/* first search (depth 0) */
{ size_t offsetFound = 999999999;
size_t const ml2 = searchMax(ms, ip, iend, &offsetFound);
if (ml2 > matchLength)
matchLength = ml2, start = ip, offset=offsetFound;
}
if (matchLength < 4) {
ip += ((ip-anchor) >> kSearchStrength) + 1; /* jump faster over incompressible sections */
continue;
}
/* let's try to find a better solution */
if (depth>=1)
while (ip<ilimit) {
ip ++;
if ( (dictMode == ZSTD_noDict)
&& (offset) && ((offset_1>0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) {
size_t const mlRep = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4;
int const gain2 = (int)(mlRep * 3);
int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1);
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
if (dictMode == ZSTD_dictMatchState) {
const U32 repIndex = (U32)(ip - base) - offset_1;
const BYTE* repMatch = repIndex < prefixLowestIndex ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */)
&& (MEM_read32(repMatch) == MEM_read32(ip)) ) {
const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend;
size_t const mlRep = ZSTD_count_2segments(ip+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4;
int const gain2 = (int)(mlRep * 3);
int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1);
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
}
{ size_t offset2=999999999;
size_t const ml2 = searchMax(ms, ip, iend, &offset2);
int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 4);
if ((ml2 >= 4) && (gain2 > gain1)) {
matchLength = ml2, offset = offset2, start = ip;
continue; /* search a better one */
} }
/* let's find an even better one */
if ((depth==2) && (ip<ilimit)) {
ip ++;
if ( (dictMode == ZSTD_noDict)
&& (offset) && ((offset_1>0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) {
size_t const mlRep = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4;
int const gain2 = (int)(mlRep * 4);
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1);
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
if (dictMode == ZSTD_dictMatchState) {
const U32 repIndex = (U32)(ip - base) - offset_1;
const BYTE* repMatch = repIndex < prefixLowestIndex ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */)
&& (MEM_read32(repMatch) == MEM_read32(ip)) ) {
const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend;
size_t const mlRep = ZSTD_count_2segments(ip+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4;
int const gain2 = (int)(mlRep * 4);
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1);
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
}
{ size_t offset2=999999999;
size_t const ml2 = searchMax(ms, ip, iend, &offset2);
int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 7);
if ((ml2 >= 4) && (gain2 > gain1)) {
matchLength = ml2, offset = offset2, start = ip;
continue;
} } }
break; /* nothing found : store previous solution */
}
/* NOTE:
* start[-offset+ZSTD_REP_MOVE-1] is undefined behavior.
* (-offset+ZSTD_REP_MOVE-1) is unsigned, and is added to start, which
* overflows the pointer, which is undefined behavior.
*/
/* catch up */
if (offset) {
if (dictMode == ZSTD_noDict) {
while ( ((start > anchor) & (start - (offset-ZSTD_REP_MOVE) > prefixLowest))
&& (start[-1] == (start-(offset-ZSTD_REP_MOVE))[-1]) ) /* only search for offset within prefix */
{ start--; matchLength++; }
}
if (dictMode == ZSTD_dictMatchState) {
U32 const matchIndex = (U32)((start-base) - (offset - ZSTD_REP_MOVE));
const BYTE* match = (matchIndex < prefixLowestIndex) ? dictBase + matchIndex - dictIndexDelta : base + matchIndex;
const BYTE* const mStart = (matchIndex < prefixLowestIndex) ? dictLowest : prefixLowest;
while ((start>anchor) && (match>mStart) && (start[-1] == match[-1])) { start--; match--; matchLength++; } /* catch up */
}
offset_2 = offset_1; offset_1 = (U32)(offset - ZSTD_REP_MOVE);
}
/* store sequence */
_storeSequence:
{ size_t const litLength = start - anchor;
ZSTD_storeSeq(seqStore, litLength, anchor, (U32)offset, matchLength-MINMATCH);
anchor = ip = start + matchLength;
}
/* check immediate repcode */
if (dictMode == ZSTD_dictMatchState) {
while (ip <= ilimit) {
U32 const current2 = (U32)(ip-base);
U32 const repIndex = current2 - offset_2;
const BYTE* repMatch = dictMode == ZSTD_dictMatchState
&& repIndex < prefixLowestIndex ?
dictBase - dictIndexDelta + repIndex :
base + repIndex;
if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex) >= 3 /* intentional overflow */)
&& (MEM_read32(repMatch) == MEM_read32(ip)) ) {
const BYTE* const repEnd2 = repIndex < prefixLowestIndex ? dictEnd : iend;
matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd2, prefixLowest) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue;
}
break;
}
}
if (dictMode == ZSTD_noDict) {
while ( ((ip <= ilimit) & (offset_2>0))
&& (MEM_read32(ip) == MEM_read32(ip - offset_2)) ) {
/* store sequence */
matchLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap repcodes */
ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue; /* faster when present ... (?) */
} } }
static size_t
ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
const BYTE* const ip, const BYTE* const iend,
size_t* offsetPtr,
U32 const mls,
const ZSTD_dictMode_e dictMode)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const hashTable = ms->hashTable;
U32 const hashLog = cParams->hashLog;
size_t const h = ZSTD_hashPtr(ip, hashLog, mls);
U32 matchIndex = hashTable[h];
const BYTE* const base = ms->window.base;
U32 const current = (U32)(ip-base);
U32 const windowLow = ms->window.lowLimit;
U32* const bt = ms->chainTable;
U32 const btLog = cParams->chainLog - 1;
U32 const btMask = (1 << btLog) - 1;
U32 const btLow = (btMask >= current) ? 0 : current - btMask;
U32 const unsortLimit = MAX(btLow, windowLow);
U32* nextCandidate = bt + 2*(matchIndex&btMask);
U32* unsortedMark = bt + 2*(matchIndex&btMask) + 1;
U32 nbCompares = 1U << cParams->searchLog;
U32 nbCandidates = nbCompares;
U32 previousCandidate = 0;
DEBUGLOG(7, "ZSTD_DUBT_findBestMatch (%u) ", current);
assert(ip <= iend-8); /* required for h calculation */
/* reach end of unsorted candidates list */
while ( (matchIndex > unsortLimit)
&& (*unsortedMark == ZSTD_DUBT_UNSORTED_MARK)
&& (nbCandidates > 1) ) {
DEBUGLOG(8, "ZSTD_DUBT_findBestMatch: candidate %u is unsorted",
matchIndex);
*unsortedMark = previousCandidate; /* the unsortedMark becomes a reversed chain, to move up back to original position */
previousCandidate = matchIndex;
matchIndex = *nextCandidate;
nextCandidate = bt + 2*(matchIndex&btMask);
unsortedMark = bt + 2*(matchIndex&btMask) + 1;
nbCandidates --;
}
/* nullify last candidate if it's still unsorted
* simplification, detrimental to compression ratio, beneficial for speed */
if ( (matchIndex > unsortLimit)
&& (*unsortedMark==ZSTD_DUBT_UNSORTED_MARK) ) {
DEBUGLOG(7, "ZSTD_DUBT_findBestMatch: nullify last unsorted candidate %u",
matchIndex);
*nextCandidate = *unsortedMark = 0;
}
/* batch sort stacked candidates */
matchIndex = previousCandidate;
while (matchIndex) { /* will end on matchIndex == 0 */
U32* const nextCandidateIdxPtr = bt + 2*(matchIndex&btMask) + 1;
U32 const nextCandidateIdx = *nextCandidateIdxPtr;
ZSTD_insertDUBT1(ms, matchIndex, iend,
nbCandidates, unsortLimit, dictMode);
matchIndex = nextCandidateIdx;
nbCandidates++;
}
/* find longest match */
{ size_t commonLengthSmaller = 0, commonLengthLarger = 0;
const BYTE* const dictBase = ms->window.dictBase;
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
U32 matchEndIdx = current + 8 + 1;
U32 dummy32; /* to be nullified at the end */
size_t bestLength = 0;
matchIndex = hashTable[h];
hashTable[h] = current; /* Update Hash Table */
while (nbCompares-- && (matchIndex > windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
const BYTE* match;
if ((dictMode != ZSTD_extDict) || (matchIndex+matchLength >= dictLimit)) {
match = base + matchIndex;
matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend);
} else {
match = dictBase + matchIndex;
matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart);
if (matchIndex+matchLength >= dictLimit)
match = base + matchIndex; /* to prepare for next usage of match[matchLength] */
}
if (matchLength > bestLength) {
if (matchLength > matchEndIdx - matchIndex)
matchEndIdx = matchIndex + (U32)matchLength;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) )
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + current - matchIndex;
if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */
if (dictMode == ZSTD_dictMatchState) {
nbCompares = 0; /* in addition to avoiding checking any
* further in this loop, make sure we
* skip checking in the dictionary. */
}
break; /* drop, to guarantee consistency (miss a little bit of compression) */
}
}
if (match[matchLength] < ip[matchLength]) {
/* match is smaller than current */
*smallerPtr = matchIndex; /* update smaller idx */
commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */
if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop the search */
smallerPtr = nextPtr+1; /* new "smaller" => larger of match */
matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */
} else {
/* match is larger than current */
*largerPtr = matchIndex;
commonLengthLarger = matchLength;
if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop the search */
largerPtr = nextPtr;
matchIndex = nextPtr[0];
} }
*smallerPtr = *largerPtr = 0;
if (dictMode == ZSTD_dictMatchState && nbCompares) {
bestLength = ZSTD_DUBT_findBetterDictMatch(
ms, ip, iend,
offsetPtr, bestLength, nbCompares,
mls, dictMode);
}
assert(matchEndIdx > current+8); /* ensure nextToUpdate is increased */
ms->nextToUpdate = matchEndIdx - 8; /* skip repetitive patterns */
if (bestLength >= MINMATCH) {
U32 const mIndex = current - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
DEBUGLOG(8, "ZSTD_DUBT_findBestMatch(%u) : found match of length %u and offsetCode %u (pos %u)",
current, (U32)bestLength, (U32)*offsetPtr, mIndex);
}
return bestLength;
}
}
static size_t
ZSTD_DUBT_findBetterDictMatch (
ZSTD_matchState_t* ms,
const BYTE* const ip, const BYTE* const iend,
size_t* offsetPtr,
size_t bestLength,
U32 nbCompares,
U32 const mls,
const ZSTD_dictMode_e dictMode)
{
const ZSTD_matchState_t * const dms = ms->dictMatchState;
const ZSTD_compressionParameters* const dmsCParams = &dms->cParams;
const U32 * const dictHashTable = dms->hashTable;
U32 const hashLog = dmsCParams->hashLog;
size_t const h = ZSTD_hashPtr(ip, hashLog, mls);
U32 dictMatchIndex = dictHashTable[h];
const BYTE* const base = ms->window.base;
const BYTE* const prefixStart = base + ms->window.dictLimit;
U32 const current = (U32)(ip-base);
const BYTE* const dictBase = dms->window.base;
const BYTE* const dictEnd = dms->window.nextSrc;
U32 const dictHighLimit = (U32)(dms->window.nextSrc - dms->window.base);
U32 const dictLowLimit = dms->window.lowLimit;
U32 const dictIndexDelta = ms->window.lowLimit - dictHighLimit;
U32* const dictBt = dms->chainTable;
U32 const btLog = dmsCParams->chainLog - 1;
U32 const btMask = (1 << btLog) - 1;
U32 const btLow = (btMask >= dictHighLimit - dictLowLimit) ? dictLowLimit : dictHighLimit - btMask;
size_t commonLengthSmaller=0, commonLengthLarger=0;
(void)dictMode;
assert(dictMode == ZSTD_dictMatchState);
while (nbCompares-- && (dictMatchIndex > dictLowLimit)) {
U32* const nextPtr = dictBt + 2*(dictMatchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
const BYTE* match = dictBase + dictMatchIndex;
matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart);
if (dictMatchIndex+matchLength >= dictHighLimit)
match = base + dictMatchIndex + dictIndexDelta; /* to prepare for next usage of match[matchLength] */
if (matchLength > bestLength) {
U32 matchIndex = dictMatchIndex + dictIndexDelta;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) {
DEBUGLOG(9, "ZSTD_DUBT_findBetterDictMatch(%u) : found better match length %u -> %u and offsetCode %u -> %u (dictMatchIndex %u, matchIndex %u)",
current, (U32)bestLength, (U32)matchLength, (U32)*offsetPtr, ZSTD_REP_MOVE + current - matchIndex, dictMatchIndex, matchIndex);
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + current - matchIndex;
}
if (ip+matchLength == iend) { /* reached end of input : ip[matchLength] is not valid, no way to know if it's larger or smaller than match */
break; /* drop, to guarantee consistency (miss a little bit of compression) */
}
}
if (match[matchLength] < ip[matchLength]) {
if (dictMatchIndex <= btLow) { break; } /* beyond tree size, stop the search */
commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */
dictMatchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */
} else {
/* match is larger than current */
if (dictMatchIndex <= btLow) { break; } /* beyond tree size, stop the search */
commonLengthLarger = matchLength;
dictMatchIndex = nextPtr[0];
}
}
if (bestLength >= MINMATCH) {
U32 const mIndex = current - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
DEBUGLOG(8, "ZSTD_DUBT_findBetterDictMatch(%u) : found match of length %u and offsetCode %u (pos %u)",
current, (U32)bestLength, (U32)*offsetPtr, mIndex);
}
return bestLength;
}
/* *******************************
* Common parser - lazy strategy
*********************************/
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx,
const void* src, size_t srcSize,
const U32 searchMethod, const U32 depth)
{
seqStore_t* seqStorePtr = &(ctx->seqStore);
const BYTE* const istart = (const BYTE*)src;
const BYTE* ip = istart;
const BYTE* anchor = istart;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - 8;
const BYTE* const base = ctx->base + ctx->dictLimit;
U32 const maxSearches = 1 << ctx->appliedParams.cParams.searchLog;
U32 const mls = ctx->appliedParams.cParams.searchLength;
typedef size_t (*searchMax_f)(ZSTD_CCtx* zc, const BYTE* ip, const BYTE* iLimit,
size_t* offsetPtr,
U32 maxNbAttempts, U32 matchLengthSearch);
searchMax_f const searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS;
U32 offset_1 = seqStorePtr->rep[0], offset_2 = seqStorePtr->rep[1], savedOffset=0;
/* init */
ip += (ip==base);
ctx->nextToUpdate3 = ctx->nextToUpdate;
{ U32 const maxRep = (U32)(ip-base);
if (offset_2 > maxRep) savedOffset = offset_2, offset_2 = 0;
if (offset_1 > maxRep) savedOffset = offset_1, offset_1 = 0;
}
/* Match Loop */
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
const BYTE* start=ip+1;
/* check repCode */
if ((offset_1>0) & (MEM_read32(ip+1) == MEM_read32(ip+1 - offset_1))) {
/* repcode : we take it */
matchLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4;
if (depth==0) goto _storeSequence;
}
/* first search (depth 0) */
{ size_t offsetFound = 99999999;
size_t const ml2 = searchMax(ctx, ip, iend, &offsetFound, maxSearches, mls);
if (ml2 > matchLength)
matchLength = ml2, start = ip, offset=offsetFound;
}
if (matchLength < 4) {
ip += ((ip-anchor) >> g_searchStrength) + 1; /* jump faster over incompressible sections */
continue;
}
/* let's try to find a better solution */
if (depth>=1)
while (ip<ilimit) {
ip ++;
if ((offset) && ((offset_1>0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) {
size_t const mlRep = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4;
int const gain2 = (int)(mlRep * 3);
int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1);
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
{ size_t offset2=99999999;
size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 4);
if ((ml2 >= 4) && (gain2 > gain1)) {
matchLength = ml2, offset = offset2, start = ip;
continue; /* search a better one */
} }
/* let's find an even better one */
if ((depth==2) && (ip<ilimit)) {
ip ++;
if ((offset) && ((offset_1>0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) {
size_t const ml2 = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4;
int const gain2 = (int)(ml2 * 4);
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1);
if ((ml2 >= 4) && (gain2 > gain1))
matchLength = ml2, offset = 0, start = ip;
}
{ size_t offset2=99999999;
size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */
int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 7);
if ((ml2 >= 4) && (gain2 > gain1)) {
matchLength = ml2, offset = offset2, start = ip;
continue;
} } }
break; /* nothing found : store previous solution */
}
/* NOTE:
* start[-offset+ZSTD_REP_MOVE-1] is undefined behavior.
* (-offset+ZSTD_REP_MOVE-1) is unsigned, and is added to start, which
* overflows the pointer, which is undefined behavior.
*/
/* catch up */
if (offset) {
while ( (start > anchor)
&& (start > base+offset-ZSTD_REP_MOVE)
&& (start[-1] == (start-offset+ZSTD_REP_MOVE)[-1]) ) /* only search for offset within prefix */
{ start--; matchLength++; }
offset_2 = offset_1; offset_1 = (U32)(offset - ZSTD_REP_MOVE);
}
/* store sequence */
_storeSequence:
{ size_t const litLength = start - anchor;
ZSTD_storeSeq(seqStorePtr, litLength, anchor, (U32)offset, matchLength-MINMATCH);
anchor = ip = start + matchLength;
}
/* check immediate repcode */
while ( (ip <= ilimit)
&& ((offset_2>0)
& (MEM_read32(ip) == MEM_read32(ip - offset_2)) )) {
/* store sequence */
matchLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap repcodes */
ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue; /* faster when present ... (?) */
} }
static size_t ZSTD_insertBtAndFindBestMatch (
ZSTD_CCtx* zc,
const BYTE* const ip, const BYTE* const iend,
size_t* offsetPtr,
U32 nbCompares, const U32 mls,
U32 extDict)
{
U32* const hashTable = zc->hashTable;
U32 const hashLog = zc->appliedParams.cParams.hashLog;
size_t const h = ZSTD_hashPtr(ip, hashLog, mls);
U32* const bt = zc->chainTable;
U32 const btLog = zc->appliedParams.cParams.chainLog - 1;
U32 const btMask = (1 << btLog) - 1;
U32 matchIndex = hashTable[h];
size_t commonLengthSmaller=0, commonLengthLarger=0;
const BYTE* const base = zc->base;
const BYTE* const dictBase = zc->dictBase;
const U32 dictLimit = zc->dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
const U32 current = (U32)(ip-base);
const U32 btLow = btMask >= current ? 0 : current - btMask;
const U32 windowLow = zc->lowLimit;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
U32 matchEndIdx = current+8;
U32 dummy32; /* to be nullified at the end */
size_t bestLength = 0;
assert(ip <= iend-8); /* required for h calculation */
hashTable[h] = current; /* Update Hash Table */
while (nbCompares-- && (matchIndex > windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
const BYTE* match;
if ((!extDict) || (matchIndex+matchLength >= dictLimit)) {
match = base + matchIndex;
if (match[matchLength] == ip[matchLength])
matchLength += ZSTD_count(ip+matchLength+1, match+matchLength+1, iend) +1;
} else {
match = dictBase + matchIndex;
matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart);
if (matchIndex+matchLength >= dictLimit)
match = base + matchIndex; /* to prepare for next usage of match[matchLength] */
}
if (matchLength > bestLength) {
if (matchLength > matchEndIdx - matchIndex)
matchEndIdx = matchIndex + (U32)matchLength;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) )
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + current - matchIndex;
if (ip+matchLength == iend) /* equal : no way to know if inf or sup */
break; /* drop, to guarantee consistency (miss a little bit of compression) */
}
if (match[matchLength] < ip[matchLength]) {
/* match is smaller than current */
*smallerPtr = matchIndex; /* update smaller idx */
commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */
if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop the search */
smallerPtr = nextPtr+1; /* new "smaller" => larger of match */
matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */
} else {
/* match is larger than current */
*largerPtr = matchIndex;
commonLengthLarger = matchLength;
if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop the search */
largerPtr = nextPtr;
matchIndex = nextPtr[0];
} }
*smallerPtr = *largerPtr = 0;
zc->nextToUpdate = (matchEndIdx > current + 8) ? matchEndIdx - 8 : current+1;
return bestLength;
}
