comparison contrib/python-zstandard/zstd/compress/zstd_lazy.c @ 40121:73fef626dae3

zstandard: vendor python-zstandard 0.10.1

This was just released.

The upstream source distribution from PyPI was extracted. Unwanted
files were removed.

The clang-format ignore list was updated to reflect the new source
of files.

setup.py was updated to pass a new argument to python-zstandard's
function for returning an Extension instance. Upstream had to change
to use relative paths because Python 3.7's packaging doesn't
seem to like absolute paths when defining sources, includes, etc.
The default relative path calculation is relative to setup_zstd.py
which is different from the directory of Mercurial's setup.py.

The project contains a vendored copy of zstandard 1.3.6. The old
version was 1.3.4.

The API should be backwards compatible and nothing in core should
need adjusted. However, there is a new "chunker" API that we
may find useful in places where we want to emit compressed chunks
of a fixed size.

There are a pair of bug fixes in 0.10.0 with regards to
compressobj() and decompressobj() when block flushing is used. I
actually found these bugs when introducing these APIs in Mercurial!
But existing Mercurial code is not affected because we don't
perform block flushing.

# no-check-commit because 3rd party code has different style guidelines

Differential Revision: https://phab.mercurial-scm.org/D4911

40120:89742f1fa6cb

/*-*************************************
*  Binary Tree search
***************************************/

void ZSTD_updateDUBT(
                ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                const BYTE* ip, const BYTE* iend,
                U32 mls)
{
    U32* const hashTable = ms->hashTable;
    U32  const hashLog = cParams->hashLog;

    U32* const bt = ms->chainTable;
    U32  const btLog  = cParams->chainLog - 1;

/** ZSTD_insertDUBT1() :
 *  sort one already inserted but unsorted position
 *  assumption : current >= btlow == (current - btmask)
 *  doesn't fail */
static void ZSTD_insertDUBT1(
                 ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                 U32 current, const BYTE* inputEnd,
                 U32 nbCompares, U32 btLow, int extDict)
{
    U32*   const bt = ms->chainTable;
    U32    const btLog  = cParams->chainLog - 1;
    U32    const btMask = (1 << btLog) - 1;
    size_t commonLengthSmaller=0, commonLengthLarger=0;
    const BYTE* const base = ms->window.base;

    while (nbCompares-- && (matchIndex > windowLow)) {
        U32* const nextPtr = bt + 2*(matchIndex & btMask);
        size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger);   /* guaranteed minimum nb of common bytes */
        assert(matchIndex < current);

        if ( (!extDict)
          || (matchIndex+matchLength >= dictLimit)  /* both in current segment*/
          || (current < dictLimit) /* both in extDict */) {
            const BYTE* const mBase = !extDict || ((matchIndex+matchLength) >= dictLimit) ? base : dictBase;
            assert( (matchIndex+matchLength >= dictLimit)   /* might be wrong if extDict is incorrectly set to 0 */
                 || (current < dictLimit) );
            match = mBase + matchIndex;
            matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend);
        } else {

    *smallerPtr = *largerPtr = 0;
}


static size_t ZSTD_DUBT_findBestMatch (
                            ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                            const BYTE* const ip, const BYTE* const iend,
                            size_t* offsetPtr,
                            U32 const mls,
                            U32 const extDict)
{
    U32*   const hashTable = ms->hashTable;
    U32    const hashLog = cParams->hashLog;
    size_t const h  = ZSTD_hashPtr(ip, hashLog, mls);
    U32          matchIndex  = hashTable[h];

    /* 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, cParams, matchIndex, iend,
                         nbCandidates, unsortLimit, extDict);
        matchIndex = nextCandidateIdx;
        nbCandidates++;
    }

    /* find longest match */

        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;
                matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend);
            } else {
                match = dictBase + matchIndex;
                matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart);

                matchIndex = nextPtr[0];
        }   }

        *smallerPtr = *largerPtr = 0;

        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)",

    }
}


/** ZSTD_BtFindBestMatch() : Tree updater, providing best match */
static size_t ZSTD_BtFindBestMatch (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* const ip, const BYTE* const iLimit,
                        size_t* offsetPtr,
                        const U32 mls /* template */)
{
    DEBUGLOG(7, "ZSTD_BtFindBestMatch");
    if (ip < ms->window.base + ms->nextToUpdate) return 0;   /* skipped area */
    ZSTD_updateDUBT(ms, cParams, ip, iLimit, mls);
    return ZSTD_DUBT_findBestMatch(ms, cParams, ip, iLimit, offsetPtr, mls, 0);
}


static size_t ZSTD_BtFindBestMatch_selectMLS (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(cParams->searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_BtFindBestMatch(ms, cParams, ip, iLimit, offsetPtr, 4);
    case 5 : return ZSTD_BtFindBestMatch(ms, cParams, ip, iLimit, offsetPtr, 5);
    case 7 :
    case 6 : return ZSTD_BtFindBestMatch(ms, cParams, ip, iLimit, offsetPtr, 6);
    }
}


/** Tree updater, providing best match */
static size_t ZSTD_BtFindBestMatch_extDict (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* const ip, const BYTE* const iLimit,
                        size_t* offsetPtr,
                        const U32 mls)
{
    DEBUGLOG(7, "ZSTD_BtFindBestMatch_extDict");
    if (ip < ms->window.base + ms->nextToUpdate) return 0;   /* skipped area */
    ZSTD_updateDUBT(ms, cParams, ip, iLimit, mls);
    return ZSTD_DUBT_findBestMatch(ms, cParams, ip, iLimit, offsetPtr, mls, 1);
}


static size_t ZSTD_BtFindBestMatch_selectMLS_extDict (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(cParams->searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_BtFindBestMatch_extDict(ms, cParams, ip, iLimit, offsetPtr, 4);
    case 5 : return ZSTD_BtFindBestMatch_extDict(ms, cParams, ip, iLimit, offsetPtr, 5);
    case 7 :
    case 6 : return ZSTD_BtFindBestMatch_extDict(ms, cParams, ip, iLimit, offsetPtr, 6);
    }
}

#define NEXT_IN_CHAIN(d, mask)   chainTable[(d) & mask]

/* Update chains up to ip (excluded)
   Assumption : always within prefix (i.e. not within extDict) */
static U32 ZSTD_insertAndFindFirstIndex_internal(
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, U32 const mls)
{
    U32* const hashTable  = ms->hashTable;
    const U32 hashLog = cParams->hashLog;
    U32* const chainTable = ms->chainTable;

    ms->nextToUpdate = target;
    return hashTable[ZSTD_hashPtr(ip, hashLog, mls)];
}

U32 ZSTD_insertAndFindFirstIndex(
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip)
{
    return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, cParams->searchLength);
}


/* inlining is important to hardwire a hot branch (template emulation) */
FORCE_INLINE_TEMPLATE
size_t ZSTD_HcFindBestMatch_generic (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* const ip, const BYTE* const iLimit,
                        size_t* offsetPtr,
                        const U32 mls, const U32 extDict)
{
    U32* const chainTable = ms->chainTable;
    const U32 chainSize = (1 << cParams->chainLog);
    const U32 chainMask = chainSize-1;
    const BYTE* const base = ms->window.base;
    const BYTE* const dictBase = ms->window.dictBase;

    /* HC4 match finder */
    U32 matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);

    for ( ; (matchIndex>lowLimit) & (nbAttempts>0) ; nbAttempts--) {
        size_t currentMl=0;
        if ((!extDict) || matchIndex >= dictLimit) {
            const BYTE* const match = base + matchIndex;
            if (match[ml] == ip[ml])   /* potentially better */
                currentMl = ZSTD_count(ip, match, iLimit);
        } else {
            const BYTE* const match = dictBase + matchIndex;

        if (matchIndex <= minChain) break;
        matchIndex = NEXT_IN_CHAIN(matchIndex, chainMask);
    }

    return ml;
}


FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(cParams->searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 4, 0);
    case 5 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 5, 0);
    case 7 :
    case 6 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 6, 0);
    }
}


FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* const offsetPtr)
{
    switch(cParams->searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 4, 1);
    case 5 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 5, 1);
    case 7 :
    case 6 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 6, 1);
    }
}


/* *******************************

*********************************/
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_generic(
                        ZSTD_matchState_t* ms, seqStore_t* seqStore,
                        U32 rep[ZSTD_REP_NUM],
                        ZSTD_compressionParameters const* cParams,
                        const void* src, size_t srcSize,
                        const U32 searchMethod, const U32 depth)
{
    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 + ms->window.dictLimit;

    typedef size_t (*searchMax_f)(
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
    searchMax_f const searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS;
    U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0;

    /* init */
    ip += (ip==base);
    ms->nextToUpdate3 = ms->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(ms, cParams, ip, iend, &offsetFound);
            if (ml2 > matchLength)
                matchLength = ml2, start = ip, offset=offsetFound;
        }

        if (matchLength < 4) {

        /* 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(ms, cParams, 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 ((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(ms, cParams, 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;

         * (-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 - (offset-ZSTD_REP_MOVE) > base))
                 && (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(seqStore, 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(seqStore, 0, anchor, 0, matchLength-MINMATCH);
            ip += matchLength;
            anchor = ip;
            continue;   /* faster when present ... (?) */
    }   }

    /* Save reps for next block */
    rep[0] = offset_1 ? offset_1 : savedOffset;
    rep[1] = offset_2 ? offset_2 : savedOffset;

}


size_t ZSTD_compressBlock_btlazy2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 1, 2);
}

size_t ZSTD_compressBlock_lazy2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 2);
}

size_t ZSTD_compressBlock_lazy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 1);
}

size_t ZSTD_compressBlock_greedy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 0);
}


FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_extDict_generic(
                        ZSTD_matchState_t* ms, seqStore_t* seqStore,
                        U32 rep[ZSTD_REP_NUM],
                        ZSTD_compressionParameters const* cParams,
                        const void* src, size_t srcSize,
                        const U32 searchMethod, const U32 depth)
{
    const BYTE* const istart = (const BYTE*)src;
    const BYTE* ip = istart;

    const BYTE* const dictBase = ms->window.dictBase;
    const BYTE* const dictEnd  = dictBase + dictLimit;
    const BYTE* const dictStart  = dictBase + lowestIndex;

    typedef size_t (*searchMax_f)(
                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                        const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
    searchMax_f searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS_extDict : ZSTD_HcFindBestMatch_extDict_selectMLS;

    U32 offset_1 = rep[0], offset_2 = rep[1];

    /* init */
    ms->nextToUpdate3 = ms->nextToUpdate;

                matchLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repEnd, prefixStart) + 4;
                if (depth==0) goto _storeSequence;
        }   }

        /* first search (depth 0) */
        {   size_t offsetFound = 99999999;
            size_t const ml2 = searchMax(ms, cParams, ip, iend, &offsetFound);
            if (ml2 > matchLength)
                matchLength = ml2, start = ip, offset=offsetFound;
        }

         if (matchLength < 4) {

                    if ((repLength >= 4) && (gain2 > gain1))
                        matchLength = repLength, offset = 0, start = ip;
            }   }

            /* search match, depth 1 */
            {   size_t offset2=99999999;
                size_t const ml2 = searchMax(ms, cParams, 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 */

                        if ((repLength >= 4) && (gain2 > gain1))
                            matchLength = repLength, offset = 0, start = ip;
                }   }

                /* search match, depth 2 */
                {   size_t offset2=99999999;
                    size_t const ml2 = searchMax(ms, cParams, 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;

}


size_t ZSTD_compressBlock_greedy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 0);
}

size_t ZSTD_compressBlock_lazy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 1);
}

size_t ZSTD_compressBlock_lazy2_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 2);
}

size_t ZSTD_compressBlock_btlazy2_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 1, 2);
}

40121:73fef626dae3

/*-*************************************
*  Binary Tree search
***************************************/

static void
ZSTD_updateDUBT(ZSTD_matchState_t* ms,
                const BYTE* ip, const BYTE* iend,
                U32 mls)
{
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
    U32* const hashTable = ms->hashTable;
    U32  const hashLog = cParams->hashLog;

    U32* const bt = ms->chainTable;
    U32  const btLog  = cParams->chainLog - 1;

/** ZSTD_insertDUBT1() :
 *  sort one already inserted but unsorted position
 *  assumption : current >= btlow == (current - btmask)
 *  doesn't fail */
static void
ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
                 U32 current, const BYTE* inputEnd,
                 U32 nbCompares, U32 btLow, const ZSTD_dictMode_e dictMode)
{
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
    U32*   const bt = ms->chainTable;
    U32    const btLog  = cParams->chainLog - 1;
    U32    const btMask = (1 << btLog) - 1;
    size_t commonLengthSmaller=0, commonLengthLarger=0;
    const BYTE* const base = ms->window.base;

    while (nbCompares-- && (matchIndex > windowLow)) {
        U32* const nextPtr = bt + 2*(matchIndex & btMask);
        size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger);   /* guaranteed minimum nb of common bytes */
        assert(matchIndex < current);

        if ( (dictMode != ZSTD_extDict)
          || (matchIndex+matchLength >= dictLimit)  /* both in current segment*/
          || (current < dictLimit) /* both in extDict */) {
            const BYTE* const mBase = ( (dictMode != ZSTD_extDict)
                                     || (matchIndex+matchLength >= dictLimit)) ?
                                        base : dictBase;
            assert( (matchIndex+matchLength >= dictLimit)   /* might be wrong if extDict is incorrectly set to 0 */
                 || (current < dictLimit) );
            match = mBase + matchIndex;
            matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend);
        } else {

    *smallerPtr = *largerPtr = 0;
}


static size_t
ZSTD_DUBT_findBetterDictMatch (
        ZSTD_matchState_t* ms,
        const BYTE* const ip, const BYTE* const iend,
        size_t* offsetPtr,
        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, bestLength=0;
    U32 matchEndIdx = current+8+1;

    (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 (matchLength > matchEndIdx - matchIndex)
                matchEndIdx = matchIndex + (U32)matchLength;
            if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) {
                DEBUGLOG(2, "ZSTD_DUBT_findBestDictMatch(%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) */
            }
        }

        DEBUGLOG(2, "matchLength:%6zu, match:%p, prefixStart:%p, ip:%p", matchLength, match, prefixStart, ip);
        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(2, "ZSTD_DUBT_findBestDictMatch(%u) : found match of length %u and offsetCode %u (pos %u)",
                    current, (U32)bestLength, (U32)*offsetPtr, mIndex);
    }
    return bestLength;

}


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];

    /* 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 */

        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);

                matchIndex = nextPtr[0];
        }   }

        *smallerPtr = *largerPtr = 0;

        if (dictMode == ZSTD_dictMatchState && nbCompares) {
            bestLength = ZSTD_DUBT_findBetterDictMatch(ms, ip, iend, offsetPtr, 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)",

    }
}


/** ZSTD_BtFindBestMatch() : Tree updater, providing best match */
FORCE_INLINE_TEMPLATE size_t
ZSTD_BtFindBestMatch( ZSTD_matchState_t* ms,
                const BYTE* const ip, const BYTE* const iLimit,
                      size_t* offsetPtr,
                const U32 mls /* template */,
                const ZSTD_dictMode_e dictMode)
{
    DEBUGLOG(7, "ZSTD_BtFindBestMatch");
    if (ip < ms->window.base + ms->nextToUpdate) return 0;   /* skipped area */
    ZSTD_updateDUBT(ms, ip, iLimit, mls);
    return ZSTD_DUBT_findBestMatch(ms, ip, iLimit, offsetPtr, mls, dictMode);
}


static size_t
ZSTD_BtFindBestMatch_selectMLS (  ZSTD_matchState_t* ms,
                            const BYTE* ip, const BYTE* const iLimit,
                                  size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict);
    case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_noDict);
    case 7 :
    case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_noDict);
    }
}


static size_t ZSTD_BtFindBestMatch_dictMatchState_selectMLS (
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState);
    case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_dictMatchState);
    case 7 :
    case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_dictMatchState);
    }
}


static size_t ZSTD_BtFindBestMatch_extDict_selectMLS (
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict);
    case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_extDict);
    case 7 :
    case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_extDict);
    }
}

#define NEXT_IN_CHAIN(d, mask)   chainTable[(d) & mask]

/* Update chains up to ip (excluded)
   Assumption : always within prefix (i.e. not within extDict) */
static U32 ZSTD_insertAndFindFirstIndex_internal(
                        ZSTD_matchState_t* ms,
                        const ZSTD_compressionParameters* const cParams,
                        const BYTE* ip, U32 const mls)
{
    U32* const hashTable  = ms->hashTable;
    const U32 hashLog = cParams->hashLog;
    U32* const chainTable = ms->chainTable;

    ms->nextToUpdate = target;
    return hashTable[ZSTD_hashPtr(ip, hashLog, mls)];
}

U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip) {
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
    return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.searchLength);
}


/* inlining is important to hardwire a hot branch (template emulation) */
FORCE_INLINE_TEMPLATE
size_t ZSTD_HcFindBestMatch_generic (
                        ZSTD_matchState_t* ms,
                        const BYTE* const ip, const BYTE* const iLimit,
                        size_t* offsetPtr,
                        const U32 mls, const ZSTD_dictMode_e dictMode)
{
    const ZSTD_compressionParameters* const cParams = &ms->cParams;
    U32* const chainTable = ms->chainTable;
    const U32 chainSize = (1 << cParams->chainLog);
    const U32 chainMask = chainSize-1;
    const BYTE* const base = ms->window.base;
    const BYTE* const dictBase = ms->window.dictBase;

    /* HC4 match finder */
    U32 matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);

    for ( ; (matchIndex>lowLimit) & (nbAttempts>0) ; nbAttempts--) {
        size_t currentMl=0;
        if ((dictMode != ZSTD_extDict) || matchIndex >= dictLimit) {
            const BYTE* const match = base + matchIndex;
            if (match[ml] == ip[ml])   /* potentially better */
                currentMl = ZSTD_count(ip, match, iLimit);
        } else {
            const BYTE* const match = dictBase + matchIndex;

        if (matchIndex <= minChain) break;
        matchIndex = NEXT_IN_CHAIN(matchIndex, chainMask);
    }

    if (dictMode == ZSTD_dictMatchState) {
        const ZSTD_matchState_t* const dms = ms->dictMatchState;
        const U32* const dmsChainTable = dms->chainTable;
        const U32 dmsChainSize         = (1 << dms->cParams.chainLog);
        const U32 dmsChainMask         = dmsChainSize - 1;
        const U32 dmsLowestIndex       = dms->window.dictLimit;
        const BYTE* const dmsBase      = dms->window.base;
        const BYTE* const dmsEnd       = dms->window.nextSrc;
        const U32 dmsSize              = (U32)(dmsEnd - dmsBase);
        const U32 dmsIndexDelta        = dictLimit - dmsSize;
        const U32 dmsMinChain = dmsSize > dmsChainSize ? dmsSize - dmsChainSize : 0;

        matchIndex = dms->hashTable[ZSTD_hashPtr(ip, dms->cParams.hashLog, mls)];

        for ( ; (matchIndex>dmsLowestIndex) & (nbAttempts>0) ; nbAttempts--) {
            size_t currentMl=0;
            const BYTE* const match = dmsBase + matchIndex;
            assert(match+4 <= dmsEnd);
            if (MEM_read32(match) == MEM_read32(ip))   /* assumption : matchIndex <= dictLimit-4 (by table construction) */
                currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, dmsEnd, prefixStart) + 4;

            /* save best solution */
            if (currentMl > ml) {
                ml = currentMl;
                *offsetPtr = current - (matchIndex + dmsIndexDelta) + ZSTD_REP_MOVE;
                if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */
            }

            if (matchIndex <= dmsMinChain) break;
            matchIndex = dmsChainTable[matchIndex & dmsChainMask];
        }
    }

    return ml;
}


FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS (
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict);
    case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_noDict);
    case 7 :
    case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_noDict);
    }
}


static size_t ZSTD_HcFindBestMatch_dictMatchState_selectMLS (
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState);
    case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_dictMatchState);
    case 7 :
    case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_dictMatchState);
    }
}


FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* const iLimit,
                        size_t* offsetPtr)
{
    switch(ms->cParams.searchLength)
    {
    default : /* includes case 3 */
    case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict);
    case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_extDict);
    case 7 :
    case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_extDict);
    }
}


/* *******************************

*********************************/
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) {

        /* 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;

         * (-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 ... (?) */
    }   }   }

    /* Save reps for next block */
    rep[0] = offset_1 ? offset_1 : savedOffset;
    rep[1] = offset_2 ? offset_2 : savedOffset;

}


size_t ZSTD_compressBlock_btlazy2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 1, 2, ZSTD_noDict);
}

size_t ZSTD_compressBlock_lazy2(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 2, ZSTD_noDict);
}

size_t ZSTD_compressBlock_lazy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 1, ZSTD_noDict);
}

size_t ZSTD_compressBlock_greedy(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 0, ZSTD_noDict);
}

size_t ZSTD_compressBlock_btlazy2_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 1, 2, ZSTD_dictMatchState);
}

size_t ZSTD_compressBlock_lazy2_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 2, ZSTD_dictMatchState);
}

size_t ZSTD_compressBlock_lazy_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 1, ZSTD_dictMatchState);
}

size_t ZSTD_compressBlock_greedy_dictMatchState(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 0, ZSTD_dictMatchState);
}


FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_extDict_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)
{
    const BYTE* const istart = (const BYTE*)src;
    const BYTE* ip = istart;

    const BYTE* const dictBase = ms->window.dictBase;
    const BYTE* const dictEnd  = dictBase + dictLimit;
    const BYTE* const dictStart  = dictBase + lowestIndex;

    typedef size_t (*searchMax_f)(
                        ZSTD_matchState_t* ms,
                        const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
    searchMax_f searchMax = searchMethod ? ZSTD_BtFindBestMatch_extDict_selectMLS : ZSTD_HcFindBestMatch_extDict_selectMLS;

    U32 offset_1 = rep[0], offset_2 = rep[1];

    /* init */
    ms->nextToUpdate3 = ms->nextToUpdate;

                matchLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repEnd, prefixStart) + 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) {

                    if ((repLength >= 4) && (gain2 > gain1))
                        matchLength = repLength, offset = 0, start = ip;
            }   }

            /* search match, depth 1 */
            {   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 */

                        if ((repLength >= 4) && (gain2 > gain1))
                            matchLength = repLength, offset = 0, start = ip;
                }   }

                /* search match, depth 2 */
                {   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;

}


size_t ZSTD_compressBlock_greedy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)
{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 0);
}

size_t ZSTD_compressBlock_lazy_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 1);
}

size_t ZSTD_compressBlock_lazy2_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 2);
}

size_t ZSTD_compressBlock_btlazy2_extDict(
        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
        void const* src, size_t srcSize)

{
    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 1, 2);
}