zstd-ruby 1.3.3.0 → 1.3.4.0

data/ext/zstdruby/libzstd/compress/zstd_double_fast.h

@@ -16,11 +16,18 @@ extern "C" {
 #endif
 
 #include "mem.h"      /* U32 */
-#include "zstd.h"     /* ZSTD_CCtx, size_t */
+#include "zstd_compress_internal.h"     /* ZSTD_CCtx, size_t */
+
+void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms,
+                              ZSTD_compressionParameters const* cParams,
+                              void const* end);
+size_t ZSTD_compressBlock_doubleFast(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_doubleFast_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize);
 
-void ZSTD_fillDoubleHashTable(ZSTD_CCtx* cctx, const void* end, const U32 mls);
-size_t ZSTD_compressBlock_doubleFast(ZSTD_CCtx* ctx, const void* src, size_t srcSize);
-size_t ZSTD_compressBlock_doubleFast_extDict(ZSTD_CCtx* ctx, const void* src, size_t srcSize);
 
 #if defined (__cplusplus)
 }

data/ext/zstdruby/libzstd/compress/zstd_fast.c

@@ -12,39 +12,48 @@
 #include "zstd_fast.h"
 
 
-void ZSTD_fillHashTable (ZSTD_CCtx* zc, const void* end, const U32 mls)
+void ZSTD_fillHashTable(ZSTD_matchState_t* ms,
+                        ZSTD_compressionParameters const* cParams,
+                        void const* end)
 {
-    U32* const hashTable = zc->hashTable;
-    U32  const hBits = zc->appliedParams.cParams.hashLog;
-    const BYTE* const base = zc->base;
-    const BYTE* ip = base + zc->nextToUpdate;
+    U32* const hashTable = ms->hashTable;
+    U32  const hBits = cParams->hashLog;
+    U32  const mls = cParams->searchLength;
+    const BYTE* const base = ms->window.base;
+    const BYTE* ip = base + ms->nextToUpdate;
     const BYTE* const iend = ((const BYTE*)end) - HASH_READ_SIZE;
-    const size_t fastHashFillStep = 3;
+    const U32 fastHashFillStep = 3;
 
-    while(ip <= iend) {
-        hashTable[ZSTD_hashPtr(ip, hBits, mls)] = (U32)(ip - base);
-        ip += fastHashFillStep;
+    /* Always insert every fastHashFillStep position into the hash table.
+     * Insert the other positions if their hash entry is empty.
+     */
+    for (; ip + fastHashFillStep - 1 <= iend; ip += fastHashFillStep) {
+        U32 const current = (U32)(ip - base);
+        U32 i;
+        for (i = 0; i < fastHashFillStep; ++i) {
+            size_t const hash = ZSTD_hashPtr(ip + i, hBits, mls);
+            if (i == 0 || hashTable[hash] == 0)
+                hashTable[hash] = current + i;
+        }
     }
 }
 
-
 FORCE_INLINE_TEMPLATE
-size_t ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx,
-                                 const void* src, size_t srcSize,
-                                 const U32 mls)
+size_t ZSTD_compressBlock_fast_generic(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize,
+        U32 const hlog, U32 const stepSize, U32 const mls)
 {
-    U32* const hashTable = cctx->hashTable;
-    U32  const hBits = cctx->appliedParams.cParams.hashLog;
-    seqStore_t* seqStorePtr = &(cctx->seqStore);
-    const BYTE* const base = cctx->base;
+    U32* const hashTable = ms->hashTable;
+    const BYTE* const base = ms->window.base;
     const BYTE* const istart = (const BYTE*)src;
     const BYTE* ip = istart;
     const BYTE* anchor = istart;
-    const U32   lowestIndex = cctx->dictLimit;
+    const U32   lowestIndex = ms->window.dictLimit;
     const BYTE* const lowest = base + lowestIndex;
     const BYTE* const iend = istart + srcSize;
     const BYTE* const ilimit = iend - HASH_READ_SIZE;
-    U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1];
+    U32 offset_1=rep[0], offset_2=rep[1];
     U32 offsetSaved = 0;
 
     /* init */
@@ -57,7 +66,7 @@ size_t ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx,
     /* Main Search Loop */
     while (ip < ilimit) {   /* < instead of <=, because repcode check at (ip+1) */
         size_t mLength;
-        size_t const h = ZSTD_hashPtr(ip, hBits, mls);
+        size_t const h = ZSTD_hashPtr(ip, hlog, mls);
         U32 const current = (U32)(ip-base);
         U32 const matchIndex = hashTable[h];
         const BYTE* match = base + matchIndex;
@@ -66,21 +75,21 @@ size_t ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx,
         if ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1))) {
             mLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4;
             ip++;
-            ZSTD_storeSeq(seqStorePtr, ip-anchor, anchor, 0, mLength-MINMATCH);
+            ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
         } else {
-            U32 offset;
-            if ( (matchIndex <= lowestIndex) || (MEM_read32(match) != MEM_read32(ip)) ) {
-                ip += ((ip-anchor) >> g_searchStrength) + 1;
+            if ( (matchIndex <= lowestIndex)
+              || (MEM_read32(match) != MEM_read32(ip)) ) {
+                assert(stepSize >= 1);
+                ip += ((ip-anchor) >> kSearchStrength) + stepSize;
                 continue;
             }
             mLength = ZSTD_count(ip+4, match+4, iend) + 4;
-            offset = (U32)(ip-match);
-            while (((ip>anchor) & (match>lowest)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
-            offset_2 = offset_1;
-            offset_1 = offset;
-
-            ZSTD_storeSeq(seqStorePtr, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
-        }
+            {   U32 const offset = (U32)(ip-match);
+                while (((ip>anchor) & (match>lowest)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
+                offset_2 = offset_1;
+                offset_1 = offset;
+                ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
+        }   }
 
         /* match found */
         ip += mLength;
@@ -88,8 +97,8 @@ size_t ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx,
 
         if (ip <= ilimit) {
             /* Fill Table */
-            hashTable[ZSTD_hashPtr(base+current+2, hBits, mls)] = current+2;  /* here because current+2 could be > iend-8 */
-            hashTable[ZSTD_hashPtr(ip-2, hBits, mls)] = (U32)(ip-2-base);
+            hashTable[ZSTD_hashPtr(base+current+2, hlog, mls)] = current+2;  /* here because current+2 could be > iend-8 */
+            hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base);
             /* check immediate repcode */
             while ( (ip <= ilimit)
                  && ( (offset_2>0)
@@ -97,65 +106,67 @@ size_t ZSTD_compressBlock_fast_generic(ZSTD_CCtx* cctx,
                 /* store sequence */
                 size_t const rLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4;
                 { U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; }  /* swap offset_2 <=> offset_1 */
-                hashTable[ZSTD_hashPtr(ip, hBits, mls)] = (U32)(ip-base);
-                ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, rLength-MINMATCH);
+                hashTable[ZSTD_hashPtr(ip, hlog, mls)] = (U32)(ip-base);
+                ZSTD_storeSeq(seqStore, 0, anchor, 0, rLength-MINMATCH);
                 ip += rLength;
                 anchor = ip;
                 continue;   /* faster when present ... (?) */
     }   }   }
 
     /* save reps for next block */
-    seqStorePtr->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved;
-    seqStorePtr->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved;
+    rep[0] = offset_1 ? offset_1 : offsetSaved;
+    rep[1] = offset_2 ? offset_2 : offsetSaved;
 
     /* Return the last literals size */
     return iend - anchor;
 }
 
 
-size_t ZSTD_compressBlock_fast(ZSTD_CCtx* ctx,
-                       const void* src, size_t srcSize)
+size_t ZSTD_compressBlock_fast(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
 {
-    const U32 mls = ctx->appliedParams.cParams.searchLength;
+    U32 const hlog = cParams->hashLog;
+    U32 const mls = cParams->searchLength;
+    U32 const stepSize = cParams->targetLength;
     switch(mls)
     {
     default: /* includes case 3 */
     case 4 :
-        return ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 4);
+        return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 4);
     case 5 :
-        return ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 5);
+        return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 5);
     case 6 :
-        return ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 6);
+        return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 6);
     case 7 :
-        return ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 7);
+        return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 7);
     }
 }
 
 
-static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
-                                 const void* src, size_t srcSize,
-                                 const U32 mls)
+static size_t ZSTD_compressBlock_fast_extDict_generic(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        void const* src, size_t srcSize,
+        U32 const hlog, U32 const stepSize, U32 const mls)
 {
-    U32* hashTable = ctx->hashTable;
-    const U32 hBits = ctx->appliedParams.cParams.hashLog;
-    seqStore_t* seqStorePtr = &(ctx->seqStore);
-    const BYTE* const base = ctx->base;
-    const BYTE* const dictBase = ctx->dictBase;
+    U32* hashTable = ms->hashTable;
+    const BYTE* const base = ms->window.base;
+    const BYTE* const dictBase = ms->window.dictBase;
     const BYTE* const istart = (const BYTE*)src;
     const BYTE* ip = istart;
     const BYTE* anchor = istart;
-    const U32   lowestIndex = ctx->lowLimit;
+    const U32   lowestIndex = ms->window.lowLimit;
     const BYTE* const dictStart = dictBase + lowestIndex;
-    const U32   dictLimit = ctx->dictLimit;
+    const U32   dictLimit = ms->window.dictLimit;
     const BYTE* const lowPrefixPtr = base + dictLimit;
     const BYTE* const dictEnd = dictBase + dictLimit;
     const BYTE* const iend = istart + srcSize;
     const BYTE* const ilimit = iend - 8;
-    U32 offset_1=seqStorePtr->rep[0], offset_2=seqStorePtr->rep[1];
+    U32 offset_1=rep[0], offset_2=rep[1];
 
     /* Search Loop */
     while (ip < ilimit) {  /* < instead of <=, because (ip+1) */
-        const size_t h = ZSTD_hashPtr(ip, hBits, mls);
+        const size_t h = ZSTD_hashPtr(ip, hlog, mls);
         const U32 matchIndex = hashTable[h];
         const BYTE* matchBase = matchIndex < dictLimit ? dictBase : base;
         const BYTE* match = matchBase + matchIndex;
@@ -171,11 +182,12 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
             const BYTE* repMatchEnd = repIndex < dictLimit ? dictEnd : iend;
             mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, lowPrefixPtr) + 4;
             ip++;
-            ZSTD_storeSeq(seqStorePtr, ip-anchor, anchor, 0, mLength-MINMATCH);
+            ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
         } else {
             if ( (matchIndex < lowestIndex) ||
                  (MEM_read32(match) != MEM_read32(ip)) ) {
-                ip += ((ip-anchor) >> g_searchStrength) + 1;
+                assert(stepSize >= 1);
+                ip += ((ip-anchor) >> kSearchStrength) + stepSize;
                 continue;
             }
             {   const BYTE* matchEnd = matchIndex < dictLimit ? dictEnd : iend;
@@ -186,7 +198,7 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
                 offset = current - matchIndex;
                 offset_2 = offset_1;
                 offset_1 = offset;
-                ZSTD_storeSeq(seqStorePtr, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
+                ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
         }   }
 
         /* found a match : store it */
@@ -195,8 +207,8 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
 
         if (ip <= ilimit) {
             /* Fill Table */
-            hashTable[ZSTD_hashPtr(base+current+2, hBits, mls)] = current+2;
-            hashTable[ZSTD_hashPtr(ip-2, hBits, mls)] = (U32)(ip-2-base);
+            hashTable[ZSTD_hashPtr(base+current+2, hlog, mls)] = current+2;
+            hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base);
             /* check immediate repcode */
             while (ip <= ilimit) {
                 U32 const current2 = (U32)(ip-base);
@@ -207,8 +219,8 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
                     const BYTE* const repEnd2 = repIndex2 < dictLimit ? dictEnd : iend;
                     size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, lowPrefixPtr) + 4;
                     U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset;   /* swap offset_2 <=> offset_1 */
-                    ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, repLength2-MINMATCH);
-                    hashTable[ZSTD_hashPtr(ip, hBits, mls)] = current2;
+                    ZSTD_storeSeq(seqStore, 0, anchor, 0, repLength2-MINMATCH);
+                    hashTable[ZSTD_hashPtr(ip, hlog, mls)] = current2;
                     ip += repLength2;
                     anchor = ip;
                     continue;
@@ -217,27 +229,31 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx* ctx,
     }   }   }
 
     /* save reps for next block */
-    seqStorePtr->repToConfirm[0] = offset_1; seqStorePtr->repToConfirm[1] = offset_2;
+    rep[0] = offset_1;
+    rep[1] = offset_2;
 
     /* Return the last literals size */
     return iend - anchor;
 }
 
 
-size_t ZSTD_compressBlock_fast_extDict(ZSTD_CCtx* ctx,
-                         const void* src, size_t srcSize)
+size_t ZSTD_compressBlock_fast_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize)
 {
-    U32 const mls = ctx->appliedParams.cParams.searchLength;
+    U32 const hlog = cParams->hashLog;
+    U32 const mls = cParams->searchLength;
+    U32 const stepSize = cParams->targetLength;
     switch(mls)
     {
     default: /* includes case 3 */
     case 4 :
-        return ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 4);
+        return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 4);
     case 5 :
-        return ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 5);
+        return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 5);
     case 6 :
-        return ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 6);
+        return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 6);
     case 7 :
-        return ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 7);
+        return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, hlog, stepSize, 7);
     }
 }

data/ext/zstdruby/libzstd/compress/zstd_fast.h

@@ -16,13 +16,17 @@ extern "C" {
 #endif
 
 #include "mem.h"      /* U32 */
-#include "zstd.h"     /* ZSTD_CCtx, size_t */
+#include "zstd_compress_internal.h"
 
-void ZSTD_fillHashTable(ZSTD_CCtx* zc, const void* end, const U32 mls);
-size_t ZSTD_compressBlock_fast(ZSTD_CCtx* ctx,
-                         const void* src, size_t srcSize);
-size_t ZSTD_compressBlock_fast_extDict(ZSTD_CCtx* ctx,
-                         const void* src, size_t srcSize);
+void ZSTD_fillHashTable(ZSTD_matchState_t* ms,
+                        ZSTD_compressionParameters const* cParams,
+                        void const* end);
+size_t ZSTD_compressBlock_fast(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize);
+size_t ZSTD_compressBlock_fast_extDict(
+        ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
+        ZSTD_compressionParameters const* cParams, void const* src, size_t srcSize);
 
 #if defined (__cplusplus)
 }

data/ext/zstdruby/libzstd/compress/zstd_lazy.c

@@ -15,76 +15,90 @@
 /*-*************************************
 *  Binary Tree search
 ***************************************/
-/** ZSTD_insertBt1() : add one or multiple positions to tree.
- *  ip : assumed <= iend-8 .
- * @return : nb of positions added */
-static U32 ZSTD_insertBt1(ZSTD_CCtx* zc,
-                const BYTE* const ip, const BYTE* const iend,
-                U32 nbCompares, U32 const mls, U32 const extDict)
+
+void ZSTD_updateDUBT(
+                ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
+                const BYTE* ip, const BYTE* iend,
+                U32 mls)
 {
-    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 hashTable = ms->hashTable;
+    U32  const hashLog = cParams->hashLog;
+
+    U32* const bt = ms->chainTable;
+    U32  const btLog  = cParams->chainLog - 1;
+    U32  const btMask = (1 << btLog) - 1;
+
+    const BYTE* const base = ms->window.base;
+    U32 const target = (U32)(ip - base);
+    U32 idx = ms->nextToUpdate;
+
+    if (idx != target)
+        DEBUGLOG(7, "ZSTD_updateDUBT, from %u to %u (dictLimit:%u)",
+                    idx, target, ms->window.dictLimit);
+    assert(ip + 8 <= iend);   /* condition for ZSTD_hashPtr */
+    (void)iend;
+
+    assert(idx >= ms->window.dictLimit);   /* condition for valid base+idx */
+    for ( ; idx < target ; idx++) {
+        size_t const h  = ZSTD_hashPtr(base + idx, hashLog, mls);   /* assumption : ip + 8 <= iend */
+        U32    const matchIndex = hashTable[h];
+
+        U32*   const nextCandidatePtr = bt + 2*(idx&btMask);
+        U32*   const sortMarkPtr  = nextCandidatePtr + 1;
+
+        DEBUGLOG(8, "ZSTD_updateDUBT: insert %u", idx);
+        hashTable[h] = idx;   /* Update Hash Table */
+        *nextCandidatePtr = matchIndex;   /* update BT like a chain */
+        *sortMarkPtr = ZSTD_DUBT_UNSORTED_MARK;
+    }
+    ms->nextToUpdate = target;
+}
+
+
+/** 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;
-    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 base = ms->window.base;
+    const BYTE* const dictBase = ms->window.dictBase;
+    const U32 dictLimit = ms->window.dictLimit;
+    const BYTE* const ip = (current>=dictLimit) ? base + current : dictBase + current;
+    const BYTE* const iend = (current>=dictLimit) ? inputEnd : dictBase + dictLimit;
     const BYTE* const dictEnd = dictBase + dictLimit;
     const BYTE* const prefixStart = base + dictLimit;
     const BYTE* match;
-    const U32 current = (U32)(ip-base);
-    const U32 btLow = btMask >= current ? 0 : current - btMask;
     U32* smallerPtr = bt + 2*(current&btMask);
     U32* largerPtr  = smallerPtr + 1;
+    U32 matchIndex = *smallerPtr;
     U32 dummy32;   /* to be nullified at the end */
-    U32 const windowLow = zc->lowLimit;
-    U32 matchEndIdx = current+8+1;
-    size_t bestLength = 8;
-#ifdef ZSTD_C_PREDICT
-    U32 predictedSmall = *(bt + 2*((current-1)&btMask) + 0);
-    U32 predictedLarge = *(bt + 2*((current-1)&btMask) + 1);
-    predictedSmall += (predictedSmall>0);
-    predictedLarge += (predictedLarge>0);
-#endif /* ZSTD_C_PREDICT */
-
-    DEBUGLOG(8, "ZSTD_insertBt1 (%u)", current);
+    U32 const windowLow = ms->window.lowLimit;
 
-    assert(ip <= iend-8);   /* required for h calculation */
-    hashTable[h] = current;   /* Update Hash Table */
+    DEBUGLOG(8, "ZSTD_insertDUBT1(%u) (dictLimit=%u, lowLimit=%u)",
+                current, dictLimit, windowLow);
+    assert(current >= btLow);
+    assert(ip < iend);   /* condition for ZSTD_count */
 
     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);
 
-#ifdef ZSTD_C_PREDICT   /* note : can create issues when hlog small <= 11 */
-        const U32* predictPtr = bt + 2*((matchIndex-1) & btMask);   /* written this way, as bt is a roll buffer */
-        if (matchIndex == predictedSmall) {
-            /* no need to check length, result known */
-            *smallerPtr = matchIndex;
-            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) */
-            predictedSmall = predictPtr[1] + (predictPtr[1]>0);
-            continue;
-        }
-        if (matchIndex == predictedLarge) {
-            *largerPtr = matchIndex;
-            if (matchIndex <= btLow) { largerPtr=&dummy32; break; }   /* beyond tree size, stop the search */
-            largerPtr = nextPtr;
-            matchIndex = nextPtr[0];
-            predictedLarge = predictPtr[0] + (predictPtr[0]>0);
-            continue;
-        }
-#endif
-
-        if ((!extDict) || (matchIndex+matchLength >= dictLimit)) {
-            assert(matchIndex+matchLength >= dictLimit);   /* might be wrong if extDict is incorrectly set to 0 */
-            match = base + matchIndex;
+        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 {
             match = dictBase + matchIndex;
@@ -93,11 +107,8 @@ static U32 ZSTD_insertBt1(ZSTD_CCtx* zc,
                 match = base + matchIndex;   /* to prepare for next usage of match[matchLength] */
         }
 
-        if (matchLength > bestLength) {
-            bestLength = matchLength;
-            if (matchLength > matchEndIdx - matchIndex)
-                matchEndIdx = matchIndex + (U32)matchLength;
-        }
+        DEBUGLOG(8, "ZSTD_insertDUBT1: comparing %u with %u : found %u common bytes ",
+                    current, matchIndex, (U32)matchLength);
 
         if (ip+matchLength == iend) {   /* equal : no way to know if inf or sup */
             break;   /* drop , to guarantee consistency ; miss a bit of compression, but other solutions can corrupt tree */
@@ -108,6 +119,8 @@ static U32 ZSTD_insertBt1(ZSTD_CCtx* zc,
             *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 searching */
+            DEBUGLOG(8, "ZSTD_insertDUBT1: %u (>btLow=%u) is smaller : next => %u",
+                        matchIndex, btLow, nextPtr[1]);
             smallerPtr = nextPtr+1;               /* new "candidate" => larger than match, which was smaller than target */
             matchIndex = nextPtr[1];              /* new matchIndex, larger than previous and closer to current */
         } else {
@@ -115,184 +128,205 @@ static U32 ZSTD_insertBt1(ZSTD_CCtx* zc,
             *largerPtr = matchIndex;
             commonLengthLarger = matchLength;
             if (matchIndex <= btLow) { largerPtr=&dummy32; break; }   /* beyond tree size, stop searching */
+            DEBUGLOG(8, "ZSTD_insertDUBT1: %u (>btLow=%u) is larger => %u",
+                        matchIndex, btLow, nextPtr[0]);
             largerPtr = nextPtr;
             matchIndex = nextPtr[0];
     }   }
 
     *smallerPtr = *largerPtr = 0;
-    if (bestLength > 384) return MIN(192, (U32)(bestLength - 384));   /* speed optimization */
-    assert(matchEndIdx > current + 8);
-    return matchEndIdx - (current + 8);
 }
 
-FORCE_INLINE_TEMPLATE
-void ZSTD_updateTree_internal(ZSTD_CCtx* zc,
-                const BYTE* const ip, const BYTE* const iend,
-                const U32 nbCompares, const U32 mls, const U32 extDict)
-{
-    const BYTE* const base = zc->base;
-    U32 const target = (U32)(ip - base);
-    U32 idx = zc->nextToUpdate;
-    DEBUGLOG(7, "ZSTD_updateTree_internal, from %u to %u  (extDict:%u)",
-                idx, target, extDict);
-
-    while(idx < target)
-        idx += ZSTD_insertBt1(zc, base+idx, iend, nbCompares, mls, extDict);
-    zc->nextToUpdate = target;
-}
 
-void ZSTD_updateTree(ZSTD_CCtx* zc,
-                const BYTE* const ip, const BYTE* const iend,
-                const U32 nbCompares, const U32 mls)
+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)
 {
-    ZSTD_updateTree_internal(zc, ip, iend, nbCompares, mls, 0 /*extDict*/);
-}
-
-void ZSTD_updateTree_extDict(ZSTD_CCtx* zc,
-                const BYTE* const ip, const BYTE* const iend,
-                const U32 nbCompares, const U32 mls)
-{
-    ZSTD_updateTree_internal(zc, ip, iend, nbCompares, mls, 1 /*extDict*/);
-}
+    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;
 
-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 bt = ms->chainTable;
+    U32    const btLog  = 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+1;
-    U32 dummy32;   /* to be nullified at the end */
-    size_t bestLength = 0;
+    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 */
-    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;
+    /* 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;
+        previousCandidate = matchIndex;
+        matchIndex = *nextCandidate;
+        nextCandidate = bt + 2*(matchIndex&btMask);
+        unsortedMark = bt + 2*(matchIndex&btMask) + 1;
+        nbCandidates --;
+    }
 
-        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);
-            if (matchIndex+matchLength >= dictLimit)
-                match = base + matchIndex;   /* to prepare for next usage of match[matchLength] */
-        }
+    if ( (matchIndex > unsortLimit)
+      && (*unsortedMark==ZSTD_DUBT_UNSORTED_MARK) ) {
+        DEBUGLOG(7, "ZSTD_DUBT_findBestMatch: nullify last unsorted candidate %u",
+                    matchIndex);
+        *nextCandidate = *unsortedMark = 0;   /* nullify next candidate if it's still unsorted (note : simplification, detrimental to compression ratio, beneficial for speed) */
+    }
+
+    /* 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++;
+    }
 
-        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) */
+    /* 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 ((!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 (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];
-    }   }
+            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) */
+                }
+            }
 
-    *smallerPtr = *largerPtr = 0;
+            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;
 
-    assert(matchEndIdx > current+8);
-    zc->nextToUpdate = matchEndIdx - 8;   /* skip repetitive patterns */
-    return bestLength;
+        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;
+    }
 }
 
 
 /** ZSTD_BtFindBestMatch() : Tree updater, providing best match */
 static size_t ZSTD_BtFindBestMatch (
-                        ZSTD_CCtx* zc,
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* const ip, const BYTE* const iLimit,
                         size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 mls)
+                        const U32 mls /* template */)
 {
-    if (ip < zc->base + zc->nextToUpdate) return 0;   /* skipped area */
-    ZSTD_updateTree(zc, ip, iLimit, maxNbAttempts, mls);
-    return ZSTD_insertBtAndFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, mls, 0);
+    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_CCtx* zc,   /* Index table will be updated */
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* ip, const BYTE* const iLimit,
-                        size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 matchLengthSearch)
+                        size_t* offsetPtr)
 {
-    switch(matchLengthSearch)
+    switch(cParams->searchLength)
     {
     default : /* includes case 3 */
-    case 4 : return ZSTD_BtFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4);
-    case 5 : return ZSTD_BtFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5);
+    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(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6);
+    case 6 : return ZSTD_BtFindBestMatch(ms, cParams, ip, iLimit, offsetPtr, 6);
     }
 }
 
 
 /** Tree updater, providing best match */
 static size_t ZSTD_BtFindBestMatch_extDict (
-                        ZSTD_CCtx* zc,
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* const ip, const BYTE* const iLimit,
                         size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 mls)
+                        const U32 mls)
 {
-    if (ip < zc->base + zc->nextToUpdate) return 0;   /* skipped area */
-    ZSTD_updateTree_extDict(zc, ip, iLimit, maxNbAttempts, mls);
-    return ZSTD_insertBtAndFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, mls, 1);
+    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_CCtx* zc,   /* Index table will be updated */
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* ip, const BYTE* const iLimit,
-                        size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 matchLengthSearch)
+                        size_t* offsetPtr)
 {
-    switch(matchLengthSearch)
+    switch(cParams->searchLength)
     {
     default : /* includes case 3 */
-    case 4 : return ZSTD_BtFindBestMatch_extDict(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4);
-    case 5 : return ZSTD_BtFindBestMatch_extDict(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5);
+    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(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6);
+    case 6 : return ZSTD_BtFindBestMatch_extDict(ms, cParams, ip, iLimit, offsetPtr, 6);
     }
 }
 
@@ -305,15 +339,17 @@ static size_t ZSTD_BtFindBestMatch_selectMLS_extDict (
 
 /* Update chains up to ip (excluded)
    Assumption : always within prefix (i.e. not within extDict) */
-U32 ZSTD_insertAndFindFirstIndex (ZSTD_CCtx* zc, const BYTE* ip, U32 mls)
+static U32 ZSTD_insertAndFindFirstIndex_internal(
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
+                        const BYTE* ip, U32 const mls)
 {
-    U32* const hashTable  = zc->hashTable;
-    const U32 hashLog = zc->appliedParams.cParams.hashLog;
-    U32* const chainTable = zc->chainTable;
-    const U32 chainMask = (1 << zc->appliedParams.cParams.chainLog) - 1;
-    const BYTE* const base = zc->base;
+    U32* const hashTable  = ms->hashTable;
+    const U32 hashLog = cParams->hashLog;
+    U32* const chainTable = ms->chainTable;
+    const U32 chainMask = (1 << cParams->chainLog) - 1;
+    const BYTE* const base = ms->window.base;
     const U32 target = (U32)(ip - base);
-    U32 idx = zc->nextToUpdate;
+    U32 idx = ms->nextToUpdate;
 
     while(idx < target) { /* catch up */
         size_t const h = ZSTD_hashPtr(base+idx, hashLog, mls);
@@ -322,35 +358,42 @@ U32 ZSTD_insertAndFindFirstIndex (ZSTD_CCtx* zc, const BYTE* ip, U32 mls)
         idx++;
     }
 
-    zc->nextToUpdate = target;
+    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_CCtx* zc,   /* Index table will be updated */
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* const ip, const BYTE* const iLimit,
                         size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 mls, const U32 extDict)
+                        const U32 mls, const U32 extDict)
 {
-    U32* const chainTable = zc->chainTable;
-    const U32 chainSize = (1 << zc->appliedParams.cParams.chainLog);
+    U32* const chainTable = ms->chainTable;
+    const U32 chainSize = (1 << cParams->chainLog);
     const U32 chainMask = chainSize-1;
-    const BYTE* const base = zc->base;
-    const BYTE* const dictBase = zc->dictBase;
-    const U32 dictLimit = zc->dictLimit;
+    const BYTE* const base = ms->window.base;
+    const BYTE* const dictBase = ms->window.dictBase;
+    const U32 dictLimit = ms->window.dictLimit;
     const BYTE* const prefixStart = base + dictLimit;
     const BYTE* const dictEnd = dictBase + dictLimit;
-    const U32 lowLimit = zc->lowLimit;
+    const U32 lowLimit = ms->window.lowLimit;
     const U32 current = (U32)(ip-base);
     const U32 minChain = current > chainSize ? current - chainSize : 0;
-    int nbAttempts=maxNbAttempts;
+    U32 nbAttempts = 1U << cParams->searchLog;
     size_t ml=4-1;
 
     /* HC4 match finder */
-    U32 matchIndex = ZSTD_insertAndFindFirstIndex (zc, ip, mls);
+    U32 matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);
 
     for ( ; (matchIndex>lowLimit) & (nbAttempts>0) ; nbAttempts--) {
         size_t currentMl=0;
@@ -381,35 +424,33 @@ size_t ZSTD_HcFindBestMatch_generic (
 
 
 FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS (
-                        ZSTD_CCtx* zc,
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* ip, const BYTE* const iLimit,
-                        size_t* offsetPtr,
-                        const U32 maxNbAttempts, const U32 matchLengthSearch)
+                        size_t* offsetPtr)
 {
-    switch(matchLengthSearch)
+    switch(cParams->searchLength)
     {
     default : /* includes case 3 */
-    case 4 : return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4, 0);
-    case 5 : return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5, 0);
+    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(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6, 0);
+    case 6 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 6, 0);
     }
 }
 
 
 FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
-                        ZSTD_CCtx* const zc,
+                        ZSTD_matchState_t* ms, ZSTD_compressionParameters const* cParams,
                         const BYTE* ip, const BYTE* const iLimit,
-                        size_t* const offsetPtr,
-                        U32 const maxNbAttempts, U32 const matchLengthSearch)
+                        size_t* const offsetPtr)
 {
-    switch(matchLengthSearch)
+    switch(cParams->searchLength)
     {
     default : /* includes case 3 */
-    case 4 : return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4, 1);
-    case 5 : return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5, 1);
+    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(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6, 1);
+    case 6 : return ZSTD_HcFindBestMatch_generic(ms, cParams, ip, iLimit, offsetPtr, 6, 1);
     }
 }
 
@@ -418,30 +459,29 @@ FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
 *  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)
+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)
 {
-    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;
+    const BYTE* const base = ms->window.base + ms->window.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);
+    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 = seqStorePtr->rep[0], offset_2 = seqStorePtr->rep[1], savedOffset=0;
+    U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0;
 
     /* init */
     ip += (ip==base);
-    ctx->nextToUpdate3 = ctx->nextToUpdate;
+    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;
@@ -462,13 +502,13 @@ size_t ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx,
 
         /* first search (depth 0) */
         {   size_t offsetFound = 99999999;
-            size_t const ml2 = searchMax(ctx, ip, iend, &offsetFound, maxSearches, mls);
+            size_t const ml2 = searchMax(ms, cParams, ip, iend, &offsetFound);
             if (ml2 > matchLength)
                 matchLength = ml2, start = ip, offset=offsetFound;
         }
 
         if (matchLength < 4) {
-            ip += ((ip-anchor) >> g_searchStrength) + 1;   /* jump faster over incompressible sections */
+            ip += ((ip-anchor) >> kSearchStrength) + 1;   /* jump faster over incompressible sections */
             continue;
         }
 
@@ -484,7 +524,7 @@ size_t ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx,
                     matchLength = mlRep, offset = 0, start = ip;
             }
             {   size_t offset2=99999999;
-                size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
+                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)) {
@@ -503,7 +543,7 @@ size_t ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx,
                         matchLength = ml2, offset = 0, start = ip;
                 }
                 {   size_t offset2=99999999;
-                    size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
+                    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)) {
@@ -528,7 +568,7 @@ size_t ZSTD_compressBlock_lazy_generic(ZSTD_CCtx* ctx,
         /* store sequence */
 _storeSequence:
         {   size_t const litLength = start - anchor;
-            ZSTD_storeSeq(seqStorePtr, litLength, anchor, (U32)offset, matchLength-MINMATCH);
+            ZSTD_storeSeq(seqStore, litLength, anchor, (U32)offset, matchLength-MINMATCH);
             anchor = ip = start + matchLength;
         }
 
@@ -538,73 +578,80 @@ _storeSequence:
             /* 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);
+            ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
             ip += matchLength;
             anchor = ip;
             continue;   /* faster when present ... (?) */
     }   }
 
     /* Save reps for next block */
-    seqStorePtr->repToConfirm[0] = offset_1 ? offset_1 : savedOffset;
-    seqStorePtr->repToConfirm[1] = offset_2 ? offset_2 : savedOffset;
+    rep[0] = offset_1 ? offset_1 : savedOffset;
+    rep[1] = offset_2 ? offset_2 : savedOffset;
 
     /* Return the last literals size */
     return iend - anchor;
 }
 
 
-size_t ZSTD_compressBlock_btlazy2(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 1, 2);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 1, 2);
 }
 
-size_t ZSTD_compressBlock_lazy2(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 2);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 2);
 }
 
-size_t ZSTD_compressBlock_lazy(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 1);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 1);
 }
 
-size_t ZSTD_compressBlock_greedy(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 0);
+    return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 0);
 }
 
 
 FORCE_INLINE_TEMPLATE
-size_t ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx,
-                                     const void* src, size_t srcSize,
-                                     const U32 searchMethod, const U32 depth)
+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)
 {
-    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;
-    const U32 dictLimit = ctx->dictLimit;
-    const U32 lowestIndex = ctx->lowLimit;
+    const BYTE* const base = ms->window.base;
+    const U32 dictLimit = ms->window.dictLimit;
+    const U32 lowestIndex = ms->window.lowLimit;
     const BYTE* const prefixStart = base + dictLimit;
-    const BYTE* const dictBase = ctx->dictBase;
+    const BYTE* const dictBase = ms->window.dictBase;
     const BYTE* const dictEnd  = dictBase + dictLimit;
-    const BYTE* const dictStart  = dictBase + ctx->lowLimit;
+    const BYTE* const dictStart  = dictBase + lowestIndex;
 
-    const U32 maxSearches = 1 << ctx->appliedParams.cParams.searchLog;
-    const U32 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);
+    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 = seqStorePtr->rep[0], offset_2 = seqStorePtr->rep[1];
+    U32 offset_1 = rep[0], offset_2 = rep[1];
 
     /* init */
-    ctx->nextToUpdate3 = ctx->nextToUpdate;
+    ms->nextToUpdate3 = ms->nextToUpdate;
     ip += (ip == prefixStart);
 
     /* Match Loop */
@@ -628,13 +675,13 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx,
 
         /* first search (depth 0) */
         {   size_t offsetFound = 99999999;
-            size_t const ml2 = searchMax(ctx, ip, iend, &offsetFound, maxSearches, mls);
+            size_t const ml2 = searchMax(ms, cParams, ip, iend, &offsetFound);
             if (ml2 > matchLength)
                 matchLength = ml2, start = ip, offset=offsetFound;
         }
 
          if (matchLength < 4) {
-            ip += ((ip-anchor) >> g_searchStrength) + 1;   /* jump faster over incompressible sections */
+            ip += ((ip-anchor) >> kSearchStrength) + 1;   /* jump faster over incompressible sections */
             continue;
         }
 
@@ -661,7 +708,7 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx,
 
             /* search match, depth 1 */
             {   size_t offset2=99999999;
-                size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
+                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)) {
@@ -691,7 +738,7 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx,
 
                 /* search match, depth 2 */
                 {   size_t offset2=99999999;
-                    size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls);
+                    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)) {
@@ -713,7 +760,7 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx* ctx,
         /* store sequence */
 _storeSequence:
         {   size_t const litLength = start - anchor;
-            ZSTD_storeSeq(seqStorePtr, litLength, anchor, (U32)offset, matchLength-MINMATCH);
+            ZSTD_storeSeq(seqStore, litLength, anchor, (U32)offset, matchLength-MINMATCH);
             anchor = ip = start + matchLength;
         }
 
@@ -728,7 +775,7 @@ _storeSequence:
                 const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
                 matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd, prefixStart) + 4;
                 offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset;   /* swap offset history */
-                ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, matchLength-MINMATCH);
+                ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
                 ip += matchLength;
                 anchor = ip;
                 continue;   /* faster when present ... (?) */
@@ -737,29 +784,41 @@ _storeSequence:
     }   }
 
     /* Save reps for next block */
-    seqStorePtr->repToConfirm[0] = offset_1; seqStorePtr->repToConfirm[1] = offset_2;
+    rep[0] = offset_1;
+    rep[1] = offset_2;
 
     /* Return the last literals size */
     return iend - anchor;
 }
 
 
-size_t ZSTD_compressBlock_greedy_extDict(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 0);
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 0);
 }
 
-size_t ZSTD_compressBlock_lazy_extDict(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 1);
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 1);
 }
 
-size_t ZSTD_compressBlock_lazy2_extDict(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 0, 2);
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 0, 2);
 }
 
-size_t ZSTD_compressBlock_btlazy2_extDict(ZSTD_CCtx* ctx, const void* src, size_t srcSize)
+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(ctx, src, srcSize, 1, 2);
+    return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, cParams, src, srcSize, 1, 2);
 }