diff --git a/vendor/github.com/DataDog/zstd/zstd_compress.c b/vendor/github.com/DataDog/zstd/zstd_compress.c
index 2aa26da..1476512 100644
--- a/vendor/github.com/DataDog/zstd/zstd_compress.c
+++ b/vendor/github.com/DataDog/zstd/zstd_compress.c
@@ -8,21 +8,14 @@
  * You may select, at your option, one of the above-listed licenses.
  */
 
-
-/*-*************************************
-*  Tuning parameters
-***************************************/
-#ifndef ZSTD_CLEVEL_DEFAULT
-#  define ZSTD_CLEVEL_DEFAULT 3
-#endif
-
-
 /*-*************************************
 *  Dependencies
 ***************************************/
+#include <limits.h>         /* INT_MAX */
 #include <string.h>         /* memset */
 #include "cpu.h"
 #include "mem.h"
+#include "hist.h"           /* HIST_countFast_wksp */
 #define FSE_STATIC_LINKING_ONLY   /* FSE_encodeSymbol */
 #include "fse.h"
 #define HUF_STATIC_LINKING_ONLY
@@ -54,7 +47,6 @@
     size_t workspaceSize;
     ZSTD_matchState_t matchState;
     ZSTD_compressedBlockState_t cBlockState;
-    ZSTD_compressionParameters cParams;
     ZSTD_customMem customMem;
     U32 dictID;
 };  /* typedef'd to ZSTD_CDict within "zstd.h" */
@@ -64,17 +56,26 @@
     return ZSTD_createCCtx_advanced(ZSTD_defaultCMem);
 }
 
+static void ZSTD_initCCtx(ZSTD_CCtx* cctx, ZSTD_customMem memManager)
+{
+    assert(cctx != NULL);
+    memset(cctx, 0, sizeof(*cctx));
+    cctx->customMem = memManager;
+    cctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid());
+    {   size_t const err = ZSTD_CCtx_reset(cctx, ZSTD_reset_parameters);
+        assert(!ZSTD_isError(err));
+        (void)err;
+    }
+}
+
 ZSTD_CCtx* ZSTD_createCCtx_advanced(ZSTD_customMem customMem)
 {
     ZSTD_STATIC_ASSERT(zcss_init==0);
     ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN==(0ULL - 1));
     if (!customMem.customAlloc ^ !customMem.customFree) return NULL;
-    {   ZSTD_CCtx* const cctx = (ZSTD_CCtx*)ZSTD_calloc(sizeof(ZSTD_CCtx), customMem);
+    {   ZSTD_CCtx* const cctx = (ZSTD_CCtx*)ZSTD_malloc(sizeof(ZSTD_CCtx), customMem);
         if (!cctx) return NULL;
-        cctx->customMem = customMem;
-        cctx->requestedParams.compressionLevel = ZSTD_CLEVEL_DEFAULT;
-        cctx->requestedParams.fParams.contentSizeFlag = 1;
-        cctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid());
+        ZSTD_initCCtx(cctx, customMem);
         return cctx;
     }
 }
@@ -102,17 +103,44 @@
     return cctx;
 }
 
-size_t ZSTD_freeCCtx(ZSTD_CCtx* cctx)
+/**
+ * Clears and frees all of the dictionaries in the CCtx.
+ */
+static void ZSTD_clearAllDicts(ZSTD_CCtx* cctx)
 {
-    if (cctx==NULL) return 0;   /* support free on NULL */
-    if (cctx->staticSize) return ERROR(memory_allocation);   /* not compatible with static CCtx */
+    ZSTD_free(cctx->localDict.dictBuffer, cctx->customMem);
+    ZSTD_freeCDict(cctx->localDict.cdict);
+    memset(&cctx->localDict, 0, sizeof(cctx->localDict));
+    memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict));
+    cctx->cdict = NULL;
+}
+
+static size_t ZSTD_sizeof_localDict(ZSTD_localDict dict)
+{
+    size_t const bufferSize = dict.dictBuffer != NULL ? dict.dictSize : 0;
+    size_t const cdictSize = ZSTD_sizeof_CDict(dict.cdict);
+    return bufferSize + cdictSize;
+}
+
+static void ZSTD_freeCCtxContent(ZSTD_CCtx* cctx)
+{
+    assert(cctx != NULL);
+    assert(cctx->staticSize == 0);
     ZSTD_free(cctx->workSpace, cctx->customMem); cctx->workSpace = NULL;
-    ZSTD_freeCDict(cctx->cdictLocal); cctx->cdictLocal = NULL;
+    ZSTD_clearAllDicts(cctx);
 #ifdef ZSTD_MULTITHREAD
     ZSTDMT_freeCCtx(cctx->mtctx); cctx->mtctx = NULL;
 #endif
+}
+
+size_t ZSTD_freeCCtx(ZSTD_CCtx* cctx)
+{
+    if (cctx==NULL) return 0;   /* support free on NULL */
+    RETURN_ERROR_IF(cctx->staticSize, memory_allocation,
+                    "not compatible with static CCtx");
+    ZSTD_freeCCtxContent(cctx);
     ZSTD_free(cctx, cctx->customMem);
-    return 0;   /* reserved as a potential error code in the future */
+    return 0;
 }
 
 
@@ -121,7 +149,7 @@
 #ifdef ZSTD_MULTITHREAD
     return ZSTDMT_sizeof_CCtx(cctx->mtctx);
 #else
-    (void) cctx;
+    (void)cctx;
     return 0;
 #endif
 }
@@ -131,7 +159,7 @@
 {
     if (cctx==NULL) return 0;   /* support sizeof on NULL */
     return sizeof(*cctx) + cctx->workSpaceSize
-           + ZSTD_sizeof_CDict(cctx->cdictLocal)
+           + ZSTD_sizeof_localDict(cctx->localDict)
            + ZSTD_sizeof_mtctx(cctx);
 }
 
@@ -143,21 +171,6 @@
 /* private API call, for dictBuilder only */
 const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx) { return &(ctx->seqStore); }
 
-ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams(
-        const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize)
-{
-    ZSTD_compressionParameters cParams = ZSTD_getCParams(CCtxParams->compressionLevel, srcSizeHint, dictSize);
-    if (CCtxParams->ldmParams.enableLdm) cParams.windowLog = ZSTD_LDM_DEFAULT_WINDOW_LOG;
-    if (CCtxParams->cParams.windowLog) cParams.windowLog = CCtxParams->cParams.windowLog;
-    if (CCtxParams->cParams.hashLog) cParams.hashLog = CCtxParams->cParams.hashLog;
-    if (CCtxParams->cParams.chainLog) cParams.chainLog = CCtxParams->cParams.chainLog;
-    if (CCtxParams->cParams.searchLog) cParams.searchLog = CCtxParams->cParams.searchLog;
-    if (CCtxParams->cParams.searchLength) cParams.searchLength = CCtxParams->cParams.searchLength;
-    if (CCtxParams->cParams.targetLength) cParams.targetLength = CCtxParams->cParams.targetLength;
-    if (CCtxParams->cParams.strategy) cParams.strategy = CCtxParams->cParams.strategy;
-    return cParams;
-}
-
 static ZSTD_CCtx_params ZSTD_makeCCtxParamsFromCParams(
         ZSTD_compressionParameters cParams)
 {
@@ -202,7 +215,7 @@
 }
 
 size_t ZSTD_CCtxParams_init(ZSTD_CCtx_params* cctxParams, int compressionLevel) {
-    if (!cctxParams) { return ERROR(GENERIC); }
+    RETURN_ERROR_IF(!cctxParams, GENERIC);
     memset(cctxParams, 0, sizeof(*cctxParams));
     cctxParams->compressionLevel = compressionLevel;
     cctxParams->fParams.contentSizeFlag = 1;
@@ -211,8 +224,8 @@
 
 size_t ZSTD_CCtxParams_init_advanced(ZSTD_CCtx_params* cctxParams, ZSTD_parameters params)
 {
-    if (!cctxParams) { return ERROR(GENERIC); }
-    CHECK_F( ZSTD_checkCParams(params.cParams) );
+    RETURN_ERROR_IF(!cctxParams, GENERIC);
+    FORWARD_IF_ERROR( ZSTD_checkCParams(params.cParams) );
     memset(cctxParams, 0, sizeof(*cctxParams));
     cctxParams->cParams = params.cParams;
     cctxParams->fParams = params.fParams;
@@ -234,249 +247,553 @@
     return ret;
 }
 
-#define CLAMPCHECK(val,min,max) {            \
-    if (((val)<(min)) | ((val)>(max))) {     \
-        return ERROR(parameter_outOfBound);  \
-}   }
+ZSTD_bounds ZSTD_cParam_getBounds(ZSTD_cParameter param)
+{
+    ZSTD_bounds bounds = { 0, 0, 0 };
+
+    switch(param)
+    {
+    case ZSTD_c_compressionLevel:
+        bounds.lowerBound = ZSTD_minCLevel();
+        bounds.upperBound = ZSTD_maxCLevel();
+        return bounds;
+
+    case ZSTD_c_windowLog:
+        bounds.lowerBound = ZSTD_WINDOWLOG_MIN;
+        bounds.upperBound = ZSTD_WINDOWLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_hashLog:
+        bounds.lowerBound = ZSTD_HASHLOG_MIN;
+        bounds.upperBound = ZSTD_HASHLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_chainLog:
+        bounds.lowerBound = ZSTD_CHAINLOG_MIN;
+        bounds.upperBound = ZSTD_CHAINLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_searchLog:
+        bounds.lowerBound = ZSTD_SEARCHLOG_MIN;
+        bounds.upperBound = ZSTD_SEARCHLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_minMatch:
+        bounds.lowerBound = ZSTD_MINMATCH_MIN;
+        bounds.upperBound = ZSTD_MINMATCH_MAX;
+        return bounds;
+
+    case ZSTD_c_targetLength:
+        bounds.lowerBound = ZSTD_TARGETLENGTH_MIN;
+        bounds.upperBound = ZSTD_TARGETLENGTH_MAX;
+        return bounds;
+
+    case ZSTD_c_strategy:
+        bounds.lowerBound = ZSTD_STRATEGY_MIN;
+        bounds.upperBound = ZSTD_STRATEGY_MAX;
+        return bounds;
+
+    case ZSTD_c_contentSizeFlag:
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_checksumFlag:
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_dictIDFlag:
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_nbWorkers:
+        bounds.lowerBound = 0;
+#ifdef ZSTD_MULTITHREAD
+        bounds.upperBound = ZSTDMT_NBWORKERS_MAX;
+#else
+        bounds.upperBound = 0;
+#endif
+        return bounds;
+
+    case ZSTD_c_jobSize:
+        bounds.lowerBound = 0;
+#ifdef ZSTD_MULTITHREAD
+        bounds.upperBound = ZSTDMT_JOBSIZE_MAX;
+#else
+        bounds.upperBound = 0;
+#endif
+        return bounds;
+
+    case ZSTD_c_overlapLog:
+        bounds.lowerBound = ZSTD_OVERLAPLOG_MIN;
+        bounds.upperBound = ZSTD_OVERLAPLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_enableLongDistanceMatching:
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_ldmHashLog:
+        bounds.lowerBound = ZSTD_LDM_HASHLOG_MIN;
+        bounds.upperBound = ZSTD_LDM_HASHLOG_MAX;
+        return bounds;
+
+    case ZSTD_c_ldmMinMatch:
+        bounds.lowerBound = ZSTD_LDM_MINMATCH_MIN;
+        bounds.upperBound = ZSTD_LDM_MINMATCH_MAX;
+        return bounds;
+
+    case ZSTD_c_ldmBucketSizeLog:
+        bounds.lowerBound = ZSTD_LDM_BUCKETSIZELOG_MIN;
+        bounds.upperBound = ZSTD_LDM_BUCKETSIZELOG_MAX;
+        return bounds;
+
+    case ZSTD_c_ldmHashRateLog:
+        bounds.lowerBound = ZSTD_LDM_HASHRATELOG_MIN;
+        bounds.upperBound = ZSTD_LDM_HASHRATELOG_MAX;
+        return bounds;
+
+    /* experimental parameters */
+    case ZSTD_c_rsyncable:
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_forceMaxWindow :
+        bounds.lowerBound = 0;
+        bounds.upperBound = 1;
+        return bounds;
+
+    case ZSTD_c_format:
+        ZSTD_STATIC_ASSERT(ZSTD_f_zstd1 < ZSTD_f_zstd1_magicless);
+        bounds.lowerBound = ZSTD_f_zstd1;
+        bounds.upperBound = ZSTD_f_zstd1_magicless;   /* note : how to ensure at compile time that this is the highest value enum ? */
+        return bounds;
+
+    case ZSTD_c_forceAttachDict:
+        ZSTD_STATIC_ASSERT(ZSTD_dictDefaultAttach < ZSTD_dictForceCopy);
+        bounds.lowerBound = ZSTD_dictDefaultAttach;
+        bounds.upperBound = ZSTD_dictForceCopy;       /* note : how to ensure at compile time that this is the highest value enum ? */
+        return bounds;
+
+    case ZSTD_c_literalCompressionMode:
+        ZSTD_STATIC_ASSERT(ZSTD_lcm_auto < ZSTD_lcm_huffman && ZSTD_lcm_huffman < ZSTD_lcm_uncompressed);
+        bounds.lowerBound = ZSTD_lcm_auto;
+        bounds.upperBound = ZSTD_lcm_uncompressed;
+        return bounds;
+
+    case ZSTD_c_targetCBlockSize:
+        bounds.lowerBound = ZSTD_TARGETCBLOCKSIZE_MIN;
+        bounds.upperBound = ZSTD_TARGETCBLOCKSIZE_MAX;
+        return bounds;
+
+    default:
+        {   ZSTD_bounds const boundError = { ERROR(parameter_unsupported), 0, 0 };
+            return boundError;
+        }
+    }
+}
+
+/* ZSTD_cParam_withinBounds:
+ * @return 1 if value is within cParam bounds,
+ * 0 otherwise */
+static int ZSTD_cParam_withinBounds(ZSTD_cParameter cParam, int value)
+{
+    ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam);
+    if (ZSTD_isError(bounds.error)) return 0;
+    if (value < bounds.lowerBound) return 0;
+    if (value > bounds.upperBound) return 0;
+    return 1;
+}
+
+/* ZSTD_cParam_clampBounds:
+ * Clamps the value into the bounded range.
+ */
+static size_t ZSTD_cParam_clampBounds(ZSTD_cParameter cParam, int* value)
+{
+    ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam);
+    if (ZSTD_isError(bounds.error)) return bounds.error;
+    if (*value < bounds.lowerBound) *value = bounds.lowerBound;
+    if (*value > bounds.upperBound) *value = bounds.upperBound;
+    return 0;
+}
+
+#define BOUNDCHECK(cParam, val) { \
+    RETURN_ERROR_IF(!ZSTD_cParam_withinBounds(cParam,val), \
+                    parameter_outOfBound); \
+}
 
 
 static int ZSTD_isUpdateAuthorized(ZSTD_cParameter param)
 {
     switch(param)
     {
-    case ZSTD_p_compressionLevel:
-    case ZSTD_p_hashLog:
-    case ZSTD_p_chainLog:
-    case ZSTD_p_searchLog:
-    case ZSTD_p_minMatch:
-    case ZSTD_p_targetLength:
-    case ZSTD_p_compressionStrategy:
-    case ZSTD_p_compressLiterals:
+    case ZSTD_c_compressionLevel:
+    case ZSTD_c_hashLog:
+    case ZSTD_c_chainLog:
+    case ZSTD_c_searchLog:
+    case ZSTD_c_minMatch:
+    case ZSTD_c_targetLength:
+    case ZSTD_c_strategy:
         return 1;
 
-    case ZSTD_p_format:
-    case ZSTD_p_windowLog:
-    case ZSTD_p_contentSizeFlag:
-    case ZSTD_p_checksumFlag:
-    case ZSTD_p_dictIDFlag:
-    case ZSTD_p_forceMaxWindow :
-    case ZSTD_p_nbWorkers:
-    case ZSTD_p_jobSize:
-    case ZSTD_p_overlapSizeLog:
-    case ZSTD_p_enableLongDistanceMatching:
-    case ZSTD_p_ldmHashLog:
-    case ZSTD_p_ldmMinMatch:
-    case ZSTD_p_ldmBucketSizeLog:
-    case ZSTD_p_ldmHashEveryLog:
+    case ZSTD_c_format:
+    case ZSTD_c_windowLog:
+    case ZSTD_c_contentSizeFlag:
+    case ZSTD_c_checksumFlag:
+    case ZSTD_c_dictIDFlag:
+    case ZSTD_c_forceMaxWindow :
+    case ZSTD_c_nbWorkers:
+    case ZSTD_c_jobSize:
+    case ZSTD_c_overlapLog:
+    case ZSTD_c_rsyncable:
+    case ZSTD_c_enableLongDistanceMatching:
+    case ZSTD_c_ldmHashLog:
+    case ZSTD_c_ldmMinMatch:
+    case ZSTD_c_ldmBucketSizeLog:
+    case ZSTD_c_ldmHashRateLog:
+    case ZSTD_c_forceAttachDict:
+    case ZSTD_c_literalCompressionMode:
+    case ZSTD_c_targetCBlockSize:
     default:
         return 0;
     }
 }
 
-size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned value)
+size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, int value)
 {
-    DEBUGLOG(4, "ZSTD_CCtx_setParameter (%u, %u)", (U32)param, value);
+    DEBUGLOG(4, "ZSTD_CCtx_setParameter (%i, %i)", (int)param, value);
     if (cctx->streamStage != zcss_init) {
         if (ZSTD_isUpdateAuthorized(param)) {
             cctx->cParamsChanged = 1;
         } else {
-            return ERROR(stage_wrong);
+            RETURN_ERROR(stage_wrong);
     }   }
 
     switch(param)
     {
-    case ZSTD_p_format :
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
+    case ZSTD_c_nbWorkers:
+        RETURN_ERROR_IF((value!=0) && cctx->staticSize, parameter_unsupported,
+                        "MT not compatible with static alloc");
+        break;
 
-    case ZSTD_p_compressionLevel:
-        if (cctx->cdict) return ERROR(stage_wrong);
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
+    case ZSTD_c_compressionLevel:
+    case ZSTD_c_windowLog:
+    case ZSTD_c_hashLog:
+    case ZSTD_c_chainLog:
+    case ZSTD_c_searchLog:
+    case ZSTD_c_minMatch:
+    case ZSTD_c_targetLength:
+    case ZSTD_c_strategy:
+    case ZSTD_c_ldmHashRateLog:
+    case ZSTD_c_format:
+    case ZSTD_c_contentSizeFlag:
+    case ZSTD_c_checksumFlag:
+    case ZSTD_c_dictIDFlag:
+    case ZSTD_c_forceMaxWindow:
+    case ZSTD_c_forceAttachDict:
+    case ZSTD_c_literalCompressionMode:
+    case ZSTD_c_jobSize:
+    case ZSTD_c_overlapLog:
+    case ZSTD_c_rsyncable:
+    case ZSTD_c_enableLongDistanceMatching:
+    case ZSTD_c_ldmHashLog:
+    case ZSTD_c_ldmMinMatch:
+    case ZSTD_c_ldmBucketSizeLog:
+    case ZSTD_c_targetCBlockSize:
+        break;
 
-    case ZSTD_p_windowLog:
-    case ZSTD_p_hashLog:
-    case ZSTD_p_chainLog:
-    case ZSTD_p_searchLog:
-    case ZSTD_p_minMatch:
-    case ZSTD_p_targetLength:
-    case ZSTD_p_compressionStrategy:
-        if (cctx->cdict) return ERROR(stage_wrong);
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    case ZSTD_p_compressLiterals:
-    case ZSTD_p_contentSizeFlag:
-    case ZSTD_p_checksumFlag:
-    case ZSTD_p_dictIDFlag:
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    case ZSTD_p_forceMaxWindow :  /* Force back-references to remain < windowSize,
-                                   * even when referencing into Dictionary content.
-                                   * default : 0 when using a CDict, 1 when using a Prefix */
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    case ZSTD_p_nbWorkers:
-        if ((value>0) && cctx->staticSize) {
-            return ERROR(parameter_unsupported);  /* MT not compatible with static alloc */
-        }
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    case ZSTD_p_jobSize:
-    case ZSTD_p_overlapSizeLog:
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    case ZSTD_p_enableLongDistanceMatching:
-    case ZSTD_p_ldmHashLog:
-    case ZSTD_p_ldmMinMatch:
-    case ZSTD_p_ldmBucketSizeLog:
-    case ZSTD_p_ldmHashEveryLog:
-        if (cctx->cdict) return ERROR(stage_wrong);
-        return ZSTD_CCtxParam_setParameter(&cctx->requestedParams, param, value);
-
-    default: return ERROR(parameter_unsupported);
+    default: RETURN_ERROR(parameter_unsupported);
     }
+    return ZSTD_CCtxParams_setParameter(&cctx->requestedParams, param, value);
 }
 
-size_t ZSTD_CCtxParam_setParameter(
-        ZSTD_CCtx_params* CCtxParams, ZSTD_cParameter param, unsigned value)
+size_t ZSTD_CCtxParams_setParameter(ZSTD_CCtx_params* CCtxParams,
+                                    ZSTD_cParameter param, int value)
 {
-    DEBUGLOG(4, "ZSTD_CCtxParam_setParameter (%u, %u)", (U32)param, value);
+    DEBUGLOG(4, "ZSTD_CCtxParams_setParameter (%i, %i)", (int)param, value);
     switch(param)
     {
-    case ZSTD_p_format :
-        if (value > (unsigned)ZSTD_f_zstd1_magicless)
-            return ERROR(parameter_unsupported);
+    case ZSTD_c_format :
+        BOUNDCHECK(ZSTD_c_format, value);
         CCtxParams->format = (ZSTD_format_e)value;
         return (size_t)CCtxParams->format;
 
-    case ZSTD_p_compressionLevel : {
-        int cLevel = (int)value;  /* cast expected to restore negative sign */
-        if (cLevel > ZSTD_maxCLevel()) cLevel = ZSTD_maxCLevel();
-        if (cLevel) {  /* 0 : does not change current level */
-            CCtxParams->disableLiteralCompression = (cLevel<0);  /* negative levels disable huffman */
-            CCtxParams->compressionLevel = cLevel;
+    case ZSTD_c_compressionLevel : {
+        FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(param, &value));
+        if (value) {  /* 0 : does not change current level */
+            CCtxParams->compressionLevel = value;
         }
         if (CCtxParams->compressionLevel >= 0) return CCtxParams->compressionLevel;
         return 0;  /* return type (size_t) cannot represent negative values */
     }
 
-    case ZSTD_p_windowLog :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, ZSTD_WINDOWLOG_MIN, ZSTD_WINDOWLOG_MAX);
+    case ZSTD_c_windowLog :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_windowLog, value);
         CCtxParams->cParams.windowLog = value;
         return CCtxParams->cParams.windowLog;
 
-    case ZSTD_p_hashLog :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, ZSTD_HASHLOG_MIN, ZSTD_HASHLOG_MAX);
+    case ZSTD_c_hashLog :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_hashLog, value);
         CCtxParams->cParams.hashLog = value;
         return CCtxParams->cParams.hashLog;
 
-    case ZSTD_p_chainLog :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, ZSTD_CHAINLOG_MIN, ZSTD_CHAINLOG_MAX);
+    case ZSTD_c_chainLog :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_chainLog, value);
         CCtxParams->cParams.chainLog = value;
         return CCtxParams->cParams.chainLog;
 
-    case ZSTD_p_searchLog :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, ZSTD_SEARCHLOG_MIN, ZSTD_SEARCHLOG_MAX);
+    case ZSTD_c_searchLog :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_searchLog, value);
         CCtxParams->cParams.searchLog = value;
         return value;
 
-    case ZSTD_p_minMatch :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, ZSTD_SEARCHLENGTH_MIN, ZSTD_SEARCHLENGTH_MAX);
-        CCtxParams->cParams.searchLength = value;
-        return CCtxParams->cParams.searchLength;
+    case ZSTD_c_minMatch :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_minMatch, value);
+        CCtxParams->cParams.minMatch = value;
+        return CCtxParams->cParams.minMatch;
 
-    case ZSTD_p_targetLength :
-        /* all values are valid. 0 => use default */
+    case ZSTD_c_targetLength :
+        BOUNDCHECK(ZSTD_c_targetLength, value);
         CCtxParams->cParams.targetLength = value;
         return CCtxParams->cParams.targetLength;
 
-    case ZSTD_p_compressionStrategy :
-        if (value>0)   /* 0 => use default */
-            CLAMPCHECK(value, (unsigned)ZSTD_fast, (unsigned)ZSTD_btultra);
+    case ZSTD_c_strategy :
+        if (value!=0)   /* 0 => use default */
+            BOUNDCHECK(ZSTD_c_strategy, value);
         CCtxParams->cParams.strategy = (ZSTD_strategy)value;
         return (size_t)CCtxParams->cParams.strategy;
 
-    case ZSTD_p_compressLiterals:
-        CCtxParams->disableLiteralCompression = !value;
-        return !CCtxParams->disableLiteralCompression;
-
-    case ZSTD_p_contentSizeFlag :
+    case ZSTD_c_contentSizeFlag :
         /* Content size written in frame header _when known_ (default:1) */
-        DEBUGLOG(4, "set content size flag = %u", (value>0));
-        CCtxParams->fParams.contentSizeFlag = value > 0;
+        DEBUGLOG(4, "set content size flag = %u", (value!=0));
+        CCtxParams->fParams.contentSizeFlag = value != 0;
         return CCtxParams->fParams.contentSizeFlag;
 
-    case ZSTD_p_checksumFlag :
+    case ZSTD_c_checksumFlag :
         /* A 32-bits content checksum will be calculated and written at end of frame (default:0) */
-        CCtxParams->fParams.checksumFlag = value > 0;
+        CCtxParams->fParams.checksumFlag = value != 0;
         return CCtxParams->fParams.checksumFlag;
 
-    case ZSTD_p_dictIDFlag : /* When applicable, dictionary's dictID is provided in frame header (default:1) */
-        DEBUGLOG(4, "set dictIDFlag = %u", (value>0));
+    case ZSTD_c_dictIDFlag : /* When applicable, dictionary's dictID is provided in frame header (default:1) */
+        DEBUGLOG(4, "set dictIDFlag = %u", (value!=0));
         CCtxParams->fParams.noDictIDFlag = !value;
         return !CCtxParams->fParams.noDictIDFlag;
 
-    case ZSTD_p_forceMaxWindow :
-        CCtxParams->forceWindow = (value > 0);
+    case ZSTD_c_forceMaxWindow :
+        CCtxParams->forceWindow = (value != 0);
         return CCtxParams->forceWindow;
 
-    case ZSTD_p_nbWorkers :
+    case ZSTD_c_forceAttachDict : {
+        const ZSTD_dictAttachPref_e pref = (ZSTD_dictAttachPref_e)value;
+        BOUNDCHECK(ZSTD_c_forceAttachDict, pref);
+        CCtxParams->attachDictPref = pref;
+        return CCtxParams->attachDictPref;
+    }
+
+    case ZSTD_c_literalCompressionMode : {
+        const ZSTD_literalCompressionMode_e lcm = (ZSTD_literalCompressionMode_e)value;
+        BOUNDCHECK(ZSTD_c_literalCompressionMode, lcm);
+        CCtxParams->literalCompressionMode = lcm;
+        return CCtxParams->literalCompressionMode;
+    }
+
+    case ZSTD_c_nbWorkers :
 #ifndef ZSTD_MULTITHREAD
-        if (value>0) return ERROR(parameter_unsupported);
+        RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading");
         return 0;
 #else
-        return ZSTDMT_CCtxParam_setNbWorkers(CCtxParams, value);
+        FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(param, &value));
+        CCtxParams->nbWorkers = value;
+        return CCtxParams->nbWorkers;
 #endif
 
-    case ZSTD_p_jobSize :
+    case ZSTD_c_jobSize :
 #ifndef ZSTD_MULTITHREAD
-        return ERROR(parameter_unsupported);
+        RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading");
+        return 0;
 #else
-        return ZSTDMT_CCtxParam_setMTCtxParameter(CCtxParams, ZSTDMT_p_jobSize, value);
+        /* Adjust to the minimum non-default value. */
+        if (value != 0 && value < ZSTDMT_JOBSIZE_MIN)
+            value = ZSTDMT_JOBSIZE_MIN;
+        FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(param, &value));
+        assert(value >= 0);
+        CCtxParams->jobSize = value;
+        return CCtxParams->jobSize;
 #endif
 
-    case ZSTD_p_overlapSizeLog :
+    case ZSTD_c_overlapLog :
 #ifndef ZSTD_MULTITHREAD
-        return ERROR(parameter_unsupported);
+        RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading");
+        return 0;
 #else
-        return ZSTDMT_CCtxParam_setMTCtxParameter(CCtxParams, ZSTDMT_p_overlapSectionLog, value);
+        FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(ZSTD_c_overlapLog, &value));
+        CCtxParams->overlapLog = value;
+        return CCtxParams->overlapLog;
 #endif
 
-    case ZSTD_p_enableLongDistanceMatching :
-        CCtxParams->ldmParams.enableLdm = (value>0);
+    case ZSTD_c_rsyncable :
+#ifndef ZSTD_MULTITHREAD
+        RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading");
+        return 0;
+#else
+        FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(ZSTD_c_overlapLog, &value));
+        CCtxParams->rsyncable = value;
+        return CCtxParams->rsyncable;
+#endif
+
+    case ZSTD_c_enableLongDistanceMatching :
+        CCtxParams->ldmParams.enableLdm = (value!=0);
         return CCtxParams->ldmParams.enableLdm;
 
-    case ZSTD_p_ldmHashLog :
-        if (value>0)   /* 0 ==> auto */
-            CLAMPCHECK(value, ZSTD_HASHLOG_MIN, ZSTD_HASHLOG_MAX);
+    case ZSTD_c_ldmHashLog :
+        if (value!=0)   /* 0 ==> auto */
+            BOUNDCHECK(ZSTD_c_ldmHashLog, value);
         CCtxParams->ldmParams.hashLog = value;
         return CCtxParams->ldmParams.hashLog;
 
-    case ZSTD_p_ldmMinMatch :
-        if (value>0)   /* 0 ==> default */
-            CLAMPCHECK(value, ZSTD_LDM_MINMATCH_MIN, ZSTD_LDM_MINMATCH_MAX);
+    case ZSTD_c_ldmMinMatch :
+        if (value!=0)   /* 0 ==> default */
+            BOUNDCHECK(ZSTD_c_ldmMinMatch, value);
         CCtxParams->ldmParams.minMatchLength = value;
         return CCtxParams->ldmParams.minMatchLength;
 
-    case ZSTD_p_ldmBucketSizeLog :
-        if (value > ZSTD_LDM_BUCKETSIZELOG_MAX)
-            return ERROR(parameter_outOfBound);
+    case ZSTD_c_ldmBucketSizeLog :
+        if (value!=0)   /* 0 ==> default */
+            BOUNDCHECK(ZSTD_c_ldmBucketSizeLog, value);
         CCtxParams->ldmParams.bucketSizeLog = value;
         return CCtxParams->ldmParams.bucketSizeLog;
 
-    case ZSTD_p_ldmHashEveryLog :
-        if (value > ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN)
-            return ERROR(parameter_outOfBound);
-        CCtxParams->ldmParams.hashEveryLog = value;
-        return CCtxParams->ldmParams.hashEveryLog;
+    case ZSTD_c_ldmHashRateLog :
+        RETURN_ERROR_IF(value > ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN,
+                        parameter_outOfBound);
+        CCtxParams->ldmParams.hashRateLog = value;
+        return CCtxParams->ldmParams.hashRateLog;
 
-    default: return ERROR(parameter_unsupported);
+    case ZSTD_c_targetCBlockSize :
+        if (value!=0)   /* 0 ==> default */
+            BOUNDCHECK(ZSTD_c_targetCBlockSize, value);
+        CCtxParams->targetCBlockSize = value;
+        return CCtxParams->targetCBlockSize;
+
+    default: RETURN_ERROR(parameter_unsupported, "unknown parameter");
     }
 }
 
+size_t ZSTD_CCtx_getParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, int* value)
+{
+    return ZSTD_CCtxParams_getParameter(&cctx->requestedParams, param, value);
+}
+
+size_t ZSTD_CCtxParams_getParameter(
+        ZSTD_CCtx_params* CCtxParams, ZSTD_cParameter param, int* value)
+{
+    switch(param)
+    {
+    case ZSTD_c_format :
+        *value = CCtxParams->format;
+        break;
+    case ZSTD_c_compressionLevel :
+        *value = CCtxParams->compressionLevel;
+        break;
+    case ZSTD_c_windowLog :
+        *value = (int)CCtxParams->cParams.windowLog;
+        break;
+    case ZSTD_c_hashLog :
+        *value = (int)CCtxParams->cParams.hashLog;
+        break;
+    case ZSTD_c_chainLog :
+        *value = (int)CCtxParams->cParams.chainLog;
+        break;
+    case ZSTD_c_searchLog :
+        *value = CCtxParams->cParams.searchLog;
+        break;
+    case ZSTD_c_minMatch :
+        *value = CCtxParams->cParams.minMatch;
+        break;
+    case ZSTD_c_targetLength :
+        *value = CCtxParams->cParams.targetLength;
+        break;
+    case ZSTD_c_strategy :
+        *value = (unsigned)CCtxParams->cParams.strategy;
+        break;
+    case ZSTD_c_contentSizeFlag :
+        *value = CCtxParams->fParams.contentSizeFlag;
+        break;
+    case ZSTD_c_checksumFlag :
+        *value = CCtxParams->fParams.checksumFlag;
+        break;
+    case ZSTD_c_dictIDFlag :
+        *value = !CCtxParams->fParams.noDictIDFlag;
+        break;
+    case ZSTD_c_forceMaxWindow :
+        *value = CCtxParams->forceWindow;
+        break;
+    case ZSTD_c_forceAttachDict :
+        *value = CCtxParams->attachDictPref;
+        break;
+    case ZSTD_c_literalCompressionMode :
+        *value = CCtxParams->literalCompressionMode;
+        break;
+    case ZSTD_c_nbWorkers :
+#ifndef ZSTD_MULTITHREAD
+        assert(CCtxParams->nbWorkers == 0);
+#endif
+        *value = CCtxParams->nbWorkers;
+        break;
+    case ZSTD_c_jobSize :
+#ifndef ZSTD_MULTITHREAD
+        RETURN_ERROR(parameter_unsupported, "not compiled with multithreading");
+#else
+        assert(CCtxParams->jobSize <= INT_MAX);
+        *value = (int)CCtxParams->jobSize;
+        break;
+#endif
+    case ZSTD_c_overlapLog :
+#ifndef ZSTD_MULTITHREAD
+        RETURN_ERROR(parameter_unsupported, "not compiled with multithreading");
+#else
+        *value = CCtxParams->overlapLog;
+        break;
+#endif
+    case ZSTD_c_rsyncable :
+#ifndef ZSTD_MULTITHREAD
+        RETURN_ERROR(parameter_unsupported, "not compiled with multithreading");
+#else
+        *value = CCtxParams->rsyncable;
+        break;
+#endif
+    case ZSTD_c_enableLongDistanceMatching :
+        *value = CCtxParams->ldmParams.enableLdm;
+        break;
+    case ZSTD_c_ldmHashLog :
+        *value = CCtxParams->ldmParams.hashLog;
+        break;
+    case ZSTD_c_ldmMinMatch :
+        *value = CCtxParams->ldmParams.minMatchLength;
+        break;
+    case ZSTD_c_ldmBucketSizeLog :
+        *value = CCtxParams->ldmParams.bucketSizeLog;
+        break;
+    case ZSTD_c_ldmHashRateLog :
+        *value = CCtxParams->ldmParams.hashRateLog;
+        break;
+    case ZSTD_c_targetCBlockSize :
+        *value = (int)CCtxParams->targetCBlockSize;
+        break;
+    default: RETURN_ERROR(parameter_unsupported, "unknown parameter");
+    }
+    return 0;
+}
+
 /** ZSTD_CCtx_setParametersUsingCCtxParams() :
  *  just applies `params` into `cctx`
  *  no action is performed, parameters are merely stored.
@@ -487,8 +804,9 @@
 size_t ZSTD_CCtx_setParametersUsingCCtxParams(
         ZSTD_CCtx* cctx, const ZSTD_CCtx_params* params)
 {
-    if (cctx->streamStage != zcss_init) return ERROR(stage_wrong);
-    if (cctx->cdict) return ERROR(stage_wrong);
+    DEBUGLOG(4, "ZSTD_CCtx_setParametersUsingCCtxParams");
+    RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
+    RETURN_ERROR_IF(cctx->cdict, stage_wrong);
 
     cctx->requestedParams = *params;
     return 0;
@@ -497,33 +815,71 @@
 ZSTDLIB_API size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_CCtx_setPledgedSrcSize to %u bytes", (U32)pledgedSrcSize);
-    if (cctx->streamStage != zcss_init) return ERROR(stage_wrong);
+    RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
     cctx->pledgedSrcSizePlusOne = pledgedSrcSize+1;
     return 0;
 }
 
+/**
+ * Initializes the local dict using the requested parameters.
+ * NOTE: This does not use the pledged src size, because it may be used for more
+ * than one compression.
+ */
+static size_t ZSTD_initLocalDict(ZSTD_CCtx* cctx)
+{
+    ZSTD_localDict* const dl = &cctx->localDict;
+    ZSTD_compressionParameters const cParams = ZSTD_getCParamsFromCCtxParams(
+            &cctx->requestedParams, 0, dl->dictSize);
+    if (dl->dict == NULL) {
+        /* No local dictionary. */
+        assert(dl->dictBuffer == NULL);
+        assert(dl->cdict == NULL);
+        assert(dl->dictSize == 0);
+        return 0;
+    }
+    if (dl->cdict != NULL) {
+        assert(cctx->cdict == dl->cdict);
+        /* Local dictionary already initialized. */
+        return 0;
+    }
+    assert(dl->dictSize > 0);
+    assert(cctx->cdict == NULL);
+    assert(cctx->prefixDict.dict == NULL);
+
+    dl->cdict = ZSTD_createCDict_advanced(
+            dl->dict,
+            dl->dictSize,
+            ZSTD_dlm_byRef,
+            dl->dictContentType,
+            cParams,
+            cctx->customMem);
+    RETURN_ERROR_IF(!dl->cdict, memory_allocation);
+    cctx->cdict = dl->cdict;
+    return 0;
+}
+
 size_t ZSTD_CCtx_loadDictionary_advanced(
         ZSTD_CCtx* cctx, const void* dict, size_t dictSize,
         ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType)
 {
-    if (cctx->streamStage != zcss_init) return ERROR(stage_wrong);
-    if (cctx->staticSize) return ERROR(memory_allocation);  /* no malloc for static CCtx */
+    RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
+    RETURN_ERROR_IF(cctx->staticSize, memory_allocation,
+                    "no malloc for static CCtx");
     DEBUGLOG(4, "ZSTD_CCtx_loadDictionary_advanced (size: %u)", (U32)dictSize);
-    ZSTD_freeCDict(cctx->cdictLocal);  /* in case one already exists */
-    if (dict==NULL || dictSize==0) {   /* no dictionary mode */
-        cctx->cdictLocal = NULL;
-        cctx->cdict = NULL;
+    ZSTD_clearAllDicts(cctx);  /* in case one already exists */
+    if (dict == NULL || dictSize == 0)  /* no dictionary mode */
+        return 0;
+    if (dictLoadMethod == ZSTD_dlm_byRef) {
+        cctx->localDict.dict = dict;
     } else {
-        ZSTD_compressionParameters const cParams =
-                ZSTD_getCParamsFromCCtxParams(&cctx->requestedParams, cctx->pledgedSrcSizePlusOne-1, dictSize);
-        cctx->cdictLocal = ZSTD_createCDict_advanced(
-                                dict, dictSize,
-                                dictLoadMethod, dictContentType,
-                                cParams, cctx->customMem);
-        cctx->cdict = cctx->cdictLocal;
-        if (cctx->cdictLocal == NULL)
-            return ERROR(memory_allocation);
+        void* dictBuffer = ZSTD_malloc(dictSize, cctx->customMem);
+        RETURN_ERROR_IF(!dictBuffer, memory_allocation);
+        memcpy(dictBuffer, dict, dictSize);
+        cctx->localDict.dictBuffer = dictBuffer;
+        cctx->localDict.dict = dictBuffer;
     }
+    cctx->localDict.dictSize = dictSize;
+    cctx->localDict.dictContentType = dictContentType;
     return 0;
 }
 
@@ -543,9 +899,10 @@
 
 size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict)
 {
-    if (cctx->streamStage != zcss_init) return ERROR(stage_wrong);
+    RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
+    /* Free the existing local cdict (if any) to save memory. */
+    ZSTD_clearAllDicts(cctx);
     cctx->cdict = cdict;
-    memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict));  /* exclusive */
     return 0;
 }
 
@@ -557,61 +914,67 @@
 size_t ZSTD_CCtx_refPrefix_advanced(
         ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType)
 {
-    if (cctx->streamStage != zcss_init) return ERROR(stage_wrong);
-    cctx->cdict = NULL;   /* prefix discards any prior cdict */
+    RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
+    ZSTD_clearAllDicts(cctx);
     cctx->prefixDict.dict = prefix;
     cctx->prefixDict.dictSize = prefixSize;
     cctx->prefixDict.dictContentType = dictContentType;
     return 0;
 }
 
-static void ZSTD_startNewCompression(ZSTD_CCtx* cctx)
-{
-    cctx->streamStage = zcss_init;
-    cctx->pledgedSrcSizePlusOne = 0;
-}
-
 /*! ZSTD_CCtx_reset() :
  *  Also dumps dictionary */
-void ZSTD_CCtx_reset(ZSTD_CCtx* cctx)
+size_t ZSTD_CCtx_reset(ZSTD_CCtx* cctx, ZSTD_ResetDirective reset)
 {
-    ZSTD_startNewCompression(cctx);
-    cctx->cdict = NULL;
+    if ( (reset == ZSTD_reset_session_only)
+      || (reset == ZSTD_reset_session_and_parameters) ) {
+        cctx->streamStage = zcss_init;
+        cctx->pledgedSrcSizePlusOne = 0;
+    }
+    if ( (reset == ZSTD_reset_parameters)
+      || (reset == ZSTD_reset_session_and_parameters) ) {
+        RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong);
+        ZSTD_clearAllDicts(cctx);
+        return ZSTD_CCtxParams_reset(&cctx->requestedParams);
+    }
+    return 0;
 }
 
+
 /** ZSTD_checkCParams() :
     control CParam values remain within authorized range.
     @return : 0, or an error code if one value is beyond authorized range */
 size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams)
 {
-    CLAMPCHECK(cParams.windowLog, ZSTD_WINDOWLOG_MIN, ZSTD_WINDOWLOG_MAX);
-    CLAMPCHECK(cParams.chainLog, ZSTD_CHAINLOG_MIN, ZSTD_CHAINLOG_MAX);
-    CLAMPCHECK(cParams.hashLog, ZSTD_HASHLOG_MIN, ZSTD_HASHLOG_MAX);
-    CLAMPCHECK(cParams.searchLog, ZSTD_SEARCHLOG_MIN, ZSTD_SEARCHLOG_MAX);
-    CLAMPCHECK(cParams.searchLength, ZSTD_SEARCHLENGTH_MIN, ZSTD_SEARCHLENGTH_MAX);
-    if ((U32)(cParams.targetLength) < ZSTD_TARGETLENGTH_MIN)
-        return ERROR(parameter_unsupported);
-    if ((U32)(cParams.strategy) > (U32)ZSTD_btultra)
-        return ERROR(parameter_unsupported);
+    BOUNDCHECK(ZSTD_c_windowLog, (int)cParams.windowLog);
+    BOUNDCHECK(ZSTD_c_chainLog,  (int)cParams.chainLog);
+    BOUNDCHECK(ZSTD_c_hashLog,   (int)cParams.hashLog);
+    BOUNDCHECK(ZSTD_c_searchLog, (int)cParams.searchLog);
+    BOUNDCHECK(ZSTD_c_minMatch,  (int)cParams.minMatch);
+    BOUNDCHECK(ZSTD_c_targetLength,(int)cParams.targetLength);
+    BOUNDCHECK(ZSTD_c_strategy,  cParams.strategy);
     return 0;
 }
 
 /** ZSTD_clampCParams() :
  *  make CParam values within valid range.
  *  @return : valid CParams */
-static ZSTD_compressionParameters ZSTD_clampCParams(ZSTD_compressionParameters cParams)
+static ZSTD_compressionParameters
+ZSTD_clampCParams(ZSTD_compressionParameters cParams)
 {
-#   define CLAMP(val,min,max) {      \
-        if (val<min) val=min;        \
-        else if (val>max) val=max;   \
+#   define CLAMP_TYPE(cParam, val, type) {                                \
+        ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam);         \
+        if ((int)val<bounds.lowerBound) val=(type)bounds.lowerBound;      \
+        else if ((int)val>bounds.upperBound) val=(type)bounds.upperBound; \
     }
-    CLAMP(cParams.windowLog, ZSTD_WINDOWLOG_MIN, ZSTD_WINDOWLOG_MAX);
-    CLAMP(cParams.chainLog, ZSTD_CHAINLOG_MIN, ZSTD_CHAINLOG_MAX);
-    CLAMP(cParams.hashLog, ZSTD_HASHLOG_MIN, ZSTD_HASHLOG_MAX);
-    CLAMP(cParams.searchLog, ZSTD_SEARCHLOG_MIN, ZSTD_SEARCHLOG_MAX);
-    CLAMP(cParams.searchLength, ZSTD_SEARCHLENGTH_MIN, ZSTD_SEARCHLENGTH_MAX);
-    if ((U32)(cParams.targetLength) < ZSTD_TARGETLENGTH_MIN) cParams.targetLength = ZSTD_TARGETLENGTH_MIN;
-    if ((U32)(cParams.strategy) > (U32)ZSTD_btultra) cParams.strategy = ZSTD_btultra;
+#   define CLAMP(cParam, val) CLAMP_TYPE(cParam, val, unsigned)
+    CLAMP(ZSTD_c_windowLog, cParams.windowLog);
+    CLAMP(ZSTD_c_chainLog,  cParams.chainLog);
+    CLAMP(ZSTD_c_hashLog,   cParams.hashLog);
+    CLAMP(ZSTD_c_searchLog, cParams.searchLog);
+    CLAMP(ZSTD_c_minMatch,  cParams.minMatch);
+    CLAMP(ZSTD_c_targetLength,cParams.targetLength);
+    CLAMP_TYPE(ZSTD_c_strategy,cParams.strategy, ZSTD_strategy);
     return cParams;
 }
 
@@ -624,17 +987,21 @@
 }
 
 /** ZSTD_adjustCParams_internal() :
-    optimize `cPar` for a given input (`srcSize` and `dictSize`).
-    mostly downsizing to reduce memory consumption and initialization latency.
-    Both `srcSize` and `dictSize` are optional (use 0 if unknown).
-    Note : cPar is considered validated at this stage. Use ZSTD_checkCParams() to ensure that condition. */
-ZSTD_compressionParameters ZSTD_adjustCParams_internal(ZSTD_compressionParameters cPar, unsigned long long srcSize, size_t dictSize)
+ *  optimize `cPar` for a specified input (`srcSize` and `dictSize`).
+ *  mostly downsize to reduce memory consumption and initialization latency.
+ * `srcSize` can be ZSTD_CONTENTSIZE_UNKNOWN when not known.
+ *  note : for the time being, `srcSize==0` means "unknown" too, for compatibility with older convention.
+ *  condition : cPar is presumed validated (can be checked using ZSTD_checkCParams()). */
+static ZSTD_compressionParameters
+ZSTD_adjustCParams_internal(ZSTD_compressionParameters cPar,
+                            unsigned long long srcSize,
+                            size_t dictSize)
 {
     static const U64 minSrcSize = 513; /* (1<<9) + 1 */
     static const U64 maxWindowResize = 1ULL << (ZSTD_WINDOWLOG_MAX-1);
     assert(ZSTD_checkCParams(cPar)==0);
 
-    if (dictSize && (srcSize+1<2) /* srcSize unknown */ )
+    if (dictSize && (srcSize+1<2) /* ZSTD_CONTENTSIZE_UNKNOWN and 0 mean "unknown" */ )
         srcSize = minSrcSize;  /* presumed small when there is a dictionary */
     else if (srcSize == 0)
         srcSize = ZSTD_CONTENTSIZE_UNKNOWN;  /* 0 == unknown : presumed large */
@@ -648,35 +1015,55 @@
                             ZSTD_highbit32(tSize-1) + 1;
         if (cPar.windowLog > srcLog) cPar.windowLog = srcLog;
     }
-    if (cPar.hashLog > cPar.windowLog) cPar.hashLog = cPar.windowLog;
+    if (cPar.hashLog > cPar.windowLog+1) cPar.hashLog = cPar.windowLog+1;
     {   U32 const cycleLog = ZSTD_cycleLog(cPar.chainLog, cPar.strategy);
         if (cycleLog > cPar.windowLog)
             cPar.chainLog -= (cycleLog - cPar.windowLog);
     }
 
     if (cPar.windowLog < ZSTD_WINDOWLOG_ABSOLUTEMIN)
-        cPar.windowLog = ZSTD_WINDOWLOG_ABSOLUTEMIN;  /* required for frame header */
+        cPar.windowLog = ZSTD_WINDOWLOG_ABSOLUTEMIN;  /* minimum wlog required for valid frame header */
 
     return cPar;
 }
 
-ZSTD_compressionParameters ZSTD_adjustCParams(ZSTD_compressionParameters cPar, unsigned long long srcSize, size_t dictSize)
+ZSTD_compressionParameters
+ZSTD_adjustCParams(ZSTD_compressionParameters cPar,
+                   unsigned long long srcSize,
+                   size_t dictSize)
 {
-    cPar = ZSTD_clampCParams(cPar);
+    cPar = ZSTD_clampCParams(cPar);   /* resulting cPar is necessarily valid (all parameters within range) */
     return ZSTD_adjustCParams_internal(cPar, srcSize, dictSize);
 }
 
-static size_t ZSTD_sizeof_matchState(ZSTD_compressionParameters const* cParams, const U32 forCCtx)
+ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams(
+        const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize)
+{
+    ZSTD_compressionParameters cParams = ZSTD_getCParams(CCtxParams->compressionLevel, srcSizeHint, dictSize);
+    if (CCtxParams->ldmParams.enableLdm) cParams.windowLog = ZSTD_LDM_DEFAULT_WINDOW_LOG;
+    if (CCtxParams->cParams.windowLog) cParams.windowLog = CCtxParams->cParams.windowLog;
+    if (CCtxParams->cParams.hashLog) cParams.hashLog = CCtxParams->cParams.hashLog;
+    if (CCtxParams->cParams.chainLog) cParams.chainLog = CCtxParams->cParams.chainLog;
+    if (CCtxParams->cParams.searchLog) cParams.searchLog = CCtxParams->cParams.searchLog;
+    if (CCtxParams->cParams.minMatch) cParams.minMatch = CCtxParams->cParams.minMatch;
+    if (CCtxParams->cParams.targetLength) cParams.targetLength = CCtxParams->cParams.targetLength;
+    if (CCtxParams->cParams.strategy) cParams.strategy = CCtxParams->cParams.strategy;
+    assert(!ZSTD_checkCParams(cParams));
+    return ZSTD_adjustCParams_internal(cParams, srcSizeHint, dictSize);
+}
+
+static size_t
+ZSTD_sizeof_matchState(const ZSTD_compressionParameters* const cParams,
+                       const U32 forCCtx)
 {
     size_t const chainSize = (cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cParams->chainLog);
     size_t const hSize = ((size_t)1) << cParams->hashLog;
-    U32    const hashLog3 = (forCCtx && cParams->searchLength==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0;
+    U32    const hashLog3 = (forCCtx && cParams->minMatch==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0;
     size_t const h3Size = ((size_t)1) << hashLog3;
     size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32);
     size_t const optPotentialSpace = ((MaxML+1) + (MaxLL+1) + (MaxOff+1) + (1<<Litbits)) * sizeof(U32)
                           + (ZSTD_OPT_NUM+1) * (sizeof(ZSTD_match_t)+sizeof(ZSTD_optimal_t));
-    size_t const optSpace = (forCCtx && ((cParams->strategy == ZSTD_btopt) ||
-                                         (cParams->strategy == ZSTD_btultra)))
+    size_t const optSpace = (forCCtx && (cParams->strategy >= ZSTD_btopt))
                                 ? optPotentialSpace
                                 : 0;
     DEBUGLOG(4, "chainSize: %u - hSize: %u - h3Size: %u",
@@ -686,14 +1073,13 @@
 
 size_t ZSTD_estimateCCtxSize_usingCCtxParams(const ZSTD_CCtx_params* params)
 {
-    /* Estimate CCtx size is supported for single-threaded compression only. */
-    if (params->nbWorkers > 0) { return ERROR(GENERIC); }
+    RETURN_ERROR_IF(params->nbWorkers > 0, GENERIC, "Estimate CCtx size is supported for single-threaded compression only.");
     {   ZSTD_compressionParameters const cParams =
                 ZSTD_getCParamsFromCCtxParams(params, 0, 0);
         size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << cParams.windowLog);
-        U32    const divider = (cParams.searchLength==3) ? 3 : 4;
+        U32    const divider = (cParams.minMatch==3) ? 3 : 4;
         size_t const maxNbSeq = blockSize / divider;
-        size_t const tokenSpace = blockSize + 11*maxNbSeq;
+        size_t const tokenSpace = WILDCOPY_OVERLENGTH + blockSize + 11*maxNbSeq;
         size_t const entropySpace = HUF_WORKSPACE_SIZE;
         size_t const blockStateSpace = 2 * sizeof(ZSTD_compressedBlockState_t);
         size_t const matchStateSize = ZSTD_sizeof_matchState(&cParams, /* forCCtx */ 1);
@@ -726,7 +1112,7 @@
 {
     int level;
     size_t memBudget = 0;
-    for (level=1; level<=compressionLevel; level++) {
+    for (level=MIN(compressionLevel, 1); level<=compressionLevel; level++) {
         size_t const newMB = ZSTD_estimateCCtxSize_internal(level);
         if (newMB > memBudget) memBudget = newMB;
     }
@@ -735,10 +1121,12 @@
 
 size_t ZSTD_estimateCStreamSize_usingCCtxParams(const ZSTD_CCtx_params* params)
 {
-    if (params->nbWorkers > 0) { return ERROR(GENERIC); }
-    {   size_t const CCtxSize = ZSTD_estimateCCtxSize_usingCCtxParams(params);
-        size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << params->cParams.windowLog);
-        size_t const inBuffSize = ((size_t)1 << params->cParams.windowLog) + blockSize;
+    RETURN_ERROR_IF(params->nbWorkers > 0, GENERIC, "Estimate CCtx size is supported for single-threaded compression only.");
+    {   ZSTD_compressionParameters const cParams =
+                ZSTD_getCParamsFromCCtxParams(params, 0, 0);
+        size_t const CCtxSize = ZSTD_estimateCCtxSize_usingCCtxParams(params);
+        size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << cParams.windowLog);
+        size_t const inBuffSize = ((size_t)1 << cParams.windowLog) + blockSize;
         size_t const outBuffSize = ZSTD_compressBound(blockSize) + 1;
         size_t const streamingSize = inBuffSize + outBuffSize;
 
@@ -752,15 +1140,17 @@
     return ZSTD_estimateCStreamSize_usingCCtxParams(&params);
 }
 
-static size_t ZSTD_estimateCStreamSize_internal(int compressionLevel) {
+static size_t ZSTD_estimateCStreamSize_internal(int compressionLevel)
+{
     ZSTD_compressionParameters const cParams = ZSTD_getCParams(compressionLevel, 0, 0);
     return ZSTD_estimateCStreamSize_usingCParams(cParams);
 }
 
-size_t ZSTD_estimateCStreamSize(int compressionLevel) {
+size_t ZSTD_estimateCStreamSize(int compressionLevel)
+{
     int level;
     size_t memBudget = 0;
-    for (level=1; level<=compressionLevel; level++) {
+    for (level=MIN(compressionLevel, 1); level<=compressionLevel; level++) {
         size_t const newMB = ZSTD_estimateCStreamSize_internal(level);
         if (newMB > memBudget) memBudget = newMB;
     }
@@ -786,9 +1176,27 @@
         fp.ingested = cctx->consumedSrcSize + buffered;
         fp.consumed = cctx->consumedSrcSize;
         fp.produced = cctx->producedCSize;
+        fp.flushed  = cctx->producedCSize;   /* simplified; some data might still be left within streaming output buffer */
+        fp.currentJobID = 0;
+        fp.nbActiveWorkers = 0;
         return fp;
 }   }
 
+/*! ZSTD_toFlushNow()
+ *  Only useful for multithreading scenarios currently (nbWorkers >= 1).
+ */
+size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx)
+{
+#ifdef ZSTD_MULTITHREAD
+    if (cctx->appliedParams.nbWorkers > 0) {
+        return ZSTDMT_toFlushNow(cctx->mtctx);
+    }
+#endif
+    (void)cctx;
+    return 0;   /* over-simplification; could also check if context is currently running in streaming mode, and in which case, report how many bytes are left to be flushed within output buffer */
+}
+
+
 
 static U32 ZSTD_equivalentCParams(ZSTD_compressionParameters cParams1,
                                   ZSTD_compressionParameters cParams2)
@@ -796,7 +1204,21 @@
     return (cParams1.hashLog  == cParams2.hashLog)
          & (cParams1.chainLog == cParams2.chainLog)
          & (cParams1.strategy == cParams2.strategy)   /* opt parser space */
-         & ((cParams1.searchLength==3) == (cParams2.searchLength==3));  /* hashlog3 space */
+         & ((cParams1.minMatch==3) == (cParams2.minMatch==3));  /* hashlog3 space */
+}
+
+static void ZSTD_assertEqualCParams(ZSTD_compressionParameters cParams1,
+                                    ZSTD_compressionParameters cParams2)
+{
+    (void)cParams1;
+    (void)cParams2;
+    assert(cParams1.windowLog    == cParams2.windowLog);
+    assert(cParams1.chainLog     == cParams2.chainLog);
+    assert(cParams1.hashLog      == cParams2.hashLog);
+    assert(cParams1.searchLog    == cParams2.searchLog);
+    assert(cParams1.minMatch     == cParams2.minMatch);
+    assert(cParams1.targetLength == cParams2.targetLength);
+    assert(cParams1.strategy     == cParams2.strategy);
 }
 
 /** The parameters are equivalent if ldm is not enabled in both sets or
@@ -809,7 +1231,7 @@
             ldmParams1.hashLog == ldmParams2.hashLog &&
             ldmParams1.bucketSizeLog == ldmParams2.bucketSizeLog &&
             ldmParams1.minMatchLength == ldmParams2.minMatchLength &&
-            ldmParams1.hashEveryLog == ldmParams2.hashEveryLog);
+            ldmParams1.hashRateLog == ldmParams2.hashRateLog);
 }
 
 typedef enum { ZSTDb_not_buffered, ZSTDb_buffered } ZSTD_buffered_policy_e;
@@ -817,33 +1239,51 @@
 /* ZSTD_sufficientBuff() :
  * check internal buffers exist for streaming if buffPol == ZSTDb_buffered .
  * Note : they are assumed to be correctly sized if ZSTD_equivalentCParams()==1 */
-static U32 ZSTD_sufficientBuff(size_t bufferSize1, size_t blockSize1,
+static U32 ZSTD_sufficientBuff(size_t bufferSize1, size_t maxNbSeq1,
+                            size_t maxNbLit1,
                             ZSTD_buffered_policy_e buffPol2,
                             ZSTD_compressionParameters cParams2,
                             U64 pledgedSrcSize)
 {
     size_t const windowSize2 = MAX(1, (size_t)MIN(((U64)1 << cParams2.windowLog), pledgedSrcSize));
     size_t const blockSize2 = MIN(ZSTD_BLOCKSIZE_MAX, windowSize2);
+    size_t const maxNbSeq2 = blockSize2 / ((cParams2.minMatch == 3) ? 3 : 4);
+    size_t const maxNbLit2 = blockSize2;
     size_t const neededBufferSize2 = (buffPol2==ZSTDb_buffered) ? windowSize2 + blockSize2 : 0;
-    DEBUGLOG(4, "ZSTD_sufficientBuff: is windowSize2=%u <= wlog1=%u",
-                (U32)windowSize2, cParams2.windowLog);
-    DEBUGLOG(4, "ZSTD_sufficientBuff: is blockSize2=%u <= blockSize1=%u",
-                (U32)blockSize2, (U32)blockSize1);
-    return (blockSize2 <= blockSize1) /* seqStore space depends on blockSize */
+    DEBUGLOG(4, "ZSTD_sufficientBuff: is neededBufferSize2=%u <= bufferSize1=%u",
+                (U32)neededBufferSize2, (U32)bufferSize1);
+    DEBUGLOG(4, "ZSTD_sufficientBuff: is maxNbSeq2=%u <= maxNbSeq1=%u",
+                (U32)maxNbSeq2, (U32)maxNbSeq1);
+    DEBUGLOG(4, "ZSTD_sufficientBuff: is maxNbLit2=%u <= maxNbLit1=%u",
+                (U32)maxNbLit2, (U32)maxNbLit1);
+    return (maxNbLit2 <= maxNbLit1)
+         & (maxNbSeq2 <= maxNbSeq1)
          & (neededBufferSize2 <= bufferSize1);
 }
 
 /** Equivalence for resetCCtx purposes */
 static U32 ZSTD_equivalentParams(ZSTD_CCtx_params params1,
                                  ZSTD_CCtx_params params2,
-                                 size_t buffSize1, size_t blockSize1,
+                                 size_t buffSize1,
+                                 size_t maxNbSeq1, size_t maxNbLit1,
                                  ZSTD_buffered_policy_e buffPol2,
                                  U64 pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_equivalentParams: pledgedSrcSize=%u", (U32)pledgedSrcSize);
-    return ZSTD_equivalentCParams(params1.cParams, params2.cParams) &&
-           ZSTD_equivalentLdmParams(params1.ldmParams, params2.ldmParams) &&
-           ZSTD_sufficientBuff(buffSize1, blockSize1, buffPol2, params2.cParams, pledgedSrcSize);
+    if (!ZSTD_equivalentCParams(params1.cParams, params2.cParams)) {
+      DEBUGLOG(4, "ZSTD_equivalentCParams() == 0");
+      return 0;
+    }
+    if (!ZSTD_equivalentLdmParams(params1.ldmParams, params2.ldmParams)) {
+      DEBUGLOG(4, "ZSTD_equivalentLdmParams() == 0");
+      return 0;
+    }
+    if (!ZSTD_sufficientBuff(buffSize1, maxNbSeq1, maxNbLit1, buffPol2,
+                             params2.cParams, pledgedSrcSize)) {
+      DEBUGLOG(4, "ZSTD_sufficientBuff() == 0");
+      return 0;
+    }
+    return 1;
 }
 
 static void ZSTD_reset_compressedBlockState(ZSTD_compressedBlockState_t* bs)
@@ -851,23 +1291,24 @@
     int i;
     for (i = 0; i < ZSTD_REP_NUM; ++i)
         bs->rep[i] = repStartValue[i];
-    bs->entropy.hufCTable_repeatMode = HUF_repeat_none;
-    bs->entropy.offcode_repeatMode = FSE_repeat_none;
-    bs->entropy.matchlength_repeatMode = FSE_repeat_none;
-    bs->entropy.litlength_repeatMode = FSE_repeat_none;
+    bs->entropy.huf.repeatMode = HUF_repeat_none;
+    bs->entropy.fse.offcode_repeatMode = FSE_repeat_none;
+    bs->entropy.fse.matchlength_repeatMode = FSE_repeat_none;
+    bs->entropy.fse.litlength_repeatMode = FSE_repeat_none;
 }
 
 /*! ZSTD_invalidateMatchState()
- * Invalidate all the matches in the match finder tables.
- * Requires nextSrc and base to be set (can be NULL).
+ *  Invalidate all the matches in the match finder tables.
+ *  Requires nextSrc and base to be set (can be NULL).
  */
 static void ZSTD_invalidateMatchState(ZSTD_matchState_t* ms)
 {
     ZSTD_window_clear(&ms->window);
 
-    ms->nextToUpdate = ms->window.dictLimit + 1;
+    ms->nextToUpdate = ms->window.dictLimit;
     ms->loadedDictEnd = 0;
     ms->opt.litLengthSum = 0;  /* force reset of btopt stats */
+    ms->dictMatchState = NULL;
 }
 
 /*! ZSTD_continueCCtx() :
@@ -880,6 +1321,7 @@
 
     cctx->blockSize = blockSize;   /* previous block size could be different even for same windowLog, due to pledgedSrcSize */
     cctx->appliedParams = params;
+    cctx->blockState.matchState.cParams = params.cParams;
     cctx->pledgedSrcSizePlusOne = pledgedSrcSize+1;
     cctx->consumedSrcSize = 0;
     cctx->producedCSize = 0;
@@ -900,11 +1342,17 @@
 
 typedef enum { ZSTDcrp_continue, ZSTDcrp_noMemset } ZSTD_compResetPolicy_e;
 
-static void* ZSTD_reset_matchState(ZSTD_matchState_t* ms, void* ptr, ZSTD_compressionParameters const* cParams, ZSTD_compResetPolicy_e const crp, U32 const forCCtx)
+typedef enum { ZSTD_resetTarget_CDict, ZSTD_resetTarget_CCtx } ZSTD_resetTarget_e;
+
+static void*
+ZSTD_reset_matchState(ZSTD_matchState_t* ms,
+                      void* ptr,
+                const ZSTD_compressionParameters* cParams,
+                      ZSTD_compResetPolicy_e const crp, ZSTD_resetTarget_e const forWho)
 {
     size_t const chainSize = (cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cParams->chainLog);
     size_t const hSize = ((size_t)1) << cParams->hashLog;
-    U32    const hashLog3 = (forCCtx && cParams->searchLength==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0;
+    U32    const hashLog3 = ((forWho == ZSTD_resetTarget_CCtx) && cParams->minMatch==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0;
     size_t const h3Size = ((size_t)1) << hashLog3;
     size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32);
 
@@ -912,12 +1360,15 @@
 
     ms->hashLog3 = hashLog3;
     memset(&ms->window, 0, sizeof(ms->window));
+    ms->window.dictLimit = 1;    /* start from 1, so that 1st position is valid */
+    ms->window.lowLimit = 1;     /* it ensures first and later CCtx usages compress the same */
+    ms->window.nextSrc = ms->window.base + 1;   /* see issue #1241 */
     ZSTD_invalidateMatchState(ms);
 
     /* opt parser space */
-    if (forCCtx && ((cParams->strategy == ZSTD_btopt) | (cParams->strategy == ZSTD_btultra))) {
+    if ((forWho == ZSTD_resetTarget_CCtx) && (cParams->strategy >= ZSTD_btopt)) {
         DEBUGLOG(4, "reserving optimal parser space");
-        ms->opt.litFreq = (U32*)ptr;
+        ms->opt.litFreq = (unsigned*)ptr;
         ms->opt.litLengthFreq = ms->opt.litFreq + (1<<Litbits);
         ms->opt.matchLengthFreq = ms->opt.litLengthFreq + (MaxLL+1);
         ms->opt.offCodeFreq = ms->opt.matchLengthFreq + (MaxML+1);
@@ -937,14 +1388,37 @@
     ms->hashTable3 = ms->chainTable + chainSize;
     ptr = ms->hashTable3 + h3Size;
 
+    ms->cParams = *cParams;
+
     assert(((size_t)ptr & 3) == 0);
     return ptr;
 }
 
+/* ZSTD_indexTooCloseToMax() :
+ * minor optimization : prefer memset() rather than reduceIndex()
+ * which is measurably slow in some circumstances (reported for Visual Studio).
+ * Works when re-using a context for a lot of smallish inputs :
+ * if all inputs are smaller than ZSTD_INDEXOVERFLOW_MARGIN,
+ * memset() will be triggered before reduceIndex().
+ */
+#define ZSTD_INDEXOVERFLOW_MARGIN (16 MB)
+static int ZSTD_indexTooCloseToMax(ZSTD_window_t w)
+{
+    return (size_t)(w.nextSrc - w.base) > (ZSTD_CURRENT_MAX - ZSTD_INDEXOVERFLOW_MARGIN);
+}
+
+#define ZSTD_WORKSPACETOOLARGE_FACTOR 3 /* define "workspace is too large" as this number of times larger than needed */
+#define ZSTD_WORKSPACETOOLARGE_MAXDURATION 128  /* when workspace is continuously too large
+                                         * during at least this number of times,
+                                         * context's memory usage is considered wasteful,
+                                         * because it's sized to handle a worst case scenario which rarely happens.
+                                         * In which case, resize it down to free some memory */
+
 /*! ZSTD_resetCCtx_internal() :
     note : `params` are assumed fully validated at this stage */
 static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc,
-                                      ZSTD_CCtx_params params, U64 pledgedSrcSize,
+                                      ZSTD_CCtx_params params,
+                                      U64 const pledgedSrcSize,
                                       ZSTD_compResetPolicy_e const crp,
                                       ZSTD_buffered_policy_e const zbuff)
 {
@@ -954,34 +1428,43 @@
 
     if (crp == ZSTDcrp_continue) {
         if (ZSTD_equivalentParams(zc->appliedParams, params,
-                                zc->inBuffSize, zc->blockSize,
-                                zbuff, pledgedSrcSize)) {
-            DEBUGLOG(4, "ZSTD_equivalentParams()==1 -> continue mode (wLog1=%u, blockSize1=%u)",
-                        zc->appliedParams.cParams.windowLog, (U32)zc->blockSize);
-            return ZSTD_continueCCtx(zc, params, pledgedSrcSize);
-    }   }
+                                  zc->inBuffSize,
+                                  zc->seqStore.maxNbSeq, zc->seqStore.maxNbLit,
+                                  zbuff, pledgedSrcSize) ) {
+            DEBUGLOG(4, "ZSTD_equivalentParams()==1 -> consider continue mode");
+            zc->workSpaceOversizedDuration += (zc->workSpaceOversizedDuration > 0);   /* if it was too large, it still is */
+            if (zc->workSpaceOversizedDuration <= ZSTD_WORKSPACETOOLARGE_MAXDURATION) {
+                DEBUGLOG(4, "continue mode confirmed (wLog1=%u, blockSize1=%zu)",
+                            zc->appliedParams.cParams.windowLog, zc->blockSize);
+                if (ZSTD_indexTooCloseToMax(zc->blockState.matchState.window)) {
+                    /* prefer a reset, faster than a rescale */
+                    ZSTD_reset_matchState(&zc->blockState.matchState,
+                                           zc->entropyWorkspace + HUF_WORKSPACE_SIZE_U32,
+                                          &params.cParams,
+                                           crp, ZSTD_resetTarget_CCtx);
+                }
+                return ZSTD_continueCCtx(zc, params, pledgedSrcSize);
+    }   }   }
     DEBUGLOG(4, "ZSTD_equivalentParams()==0 -> reset CCtx");
 
     if (params.ldmParams.enableLdm) {
         /* Adjust long distance matching parameters */
-        params.ldmParams.windowLog = params.cParams.windowLog;
         ZSTD_ldm_adjustParameters(&params.ldmParams, &params.cParams);
         assert(params.ldmParams.hashLog >= params.ldmParams.bucketSizeLog);
-        assert(params.ldmParams.hashEveryLog < 32);
-        zc->ldmState.hashPower =
-                ZSTD_ldm_getHashPower(params.ldmParams.minMatchLength);
+        assert(params.ldmParams.hashRateLog < 32);
+        zc->ldmState.hashPower = ZSTD_rollingHash_primePower(params.ldmParams.minMatchLength);
     }
 
     {   size_t const windowSize = MAX(1, (size_t)MIN(((U64)1 << params.cParams.windowLog), pledgedSrcSize));
         size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, windowSize);
-        U32    const divider = (params.cParams.searchLength==3) ? 3 : 4;
+        U32    const divider = (params.cParams.minMatch==3) ? 3 : 4;
         size_t const maxNbSeq = blockSize / divider;
-        size_t const tokenSpace = blockSize + 11*maxNbSeq;
+        size_t const tokenSpace = WILDCOPY_OVERLENGTH + blockSize + 11*maxNbSeq;
         size_t const buffOutSize = (zbuff==ZSTDb_buffered) ? ZSTD_compressBound(blockSize)+1 : 0;
         size_t const buffInSize = (zbuff==ZSTDb_buffered) ? windowSize + blockSize : 0;
         size_t const matchStateSize = ZSTD_sizeof_matchState(&params.cParams, /* forCCtx */ 1);
         size_t const maxNbLdmSeq = ZSTD_ldm_getMaxNbSeq(params.ldmParams, blockSize);
-        void* ptr;
+        void* ptr;   /* used to partition workSpace */
 
         /* Check if workSpace is large enough, alloc a new one if needed */
         {   size_t const entropySpace = HUF_WORKSPACE_SIZE;
@@ -993,25 +1476,33 @@
             size_t const neededSpace = entropySpace + blockStateSpace + ldmSpace +
                                        ldmSeqSpace + matchStateSize + tokenSpace +
                                        bufferSpace;
-            DEBUGLOG(4, "Need %uKB workspace, including %uKB for match state, and %uKB for buffers",
-                        (U32)(neededSpace>>10), (U32)(matchStateSize>>10), (U32)(bufferSpace>>10));
-            DEBUGLOG(4, "windowSize: %u - blockSize: %u", (U32)windowSize, (U32)blockSize);
 
-            if (zc->workSpaceSize < neededSpace) {  /* too small : resize */
-                DEBUGLOG(4, "Need to update workSpaceSize from %uK to %uK",
-                            (unsigned)(zc->workSpaceSize>>10),
-                            (unsigned)(neededSpace>>10));
-                /* static cctx : no resize, error out */
-                if (zc->staticSize) return ERROR(memory_allocation);
+            int const workSpaceTooSmall = zc->workSpaceSize < neededSpace;
+            int const workSpaceTooLarge = zc->workSpaceSize > ZSTD_WORKSPACETOOLARGE_FACTOR * neededSpace;
+            int const workSpaceWasteful = workSpaceTooLarge && (zc->workSpaceOversizedDuration > ZSTD_WORKSPACETOOLARGE_MAXDURATION);
+            zc->workSpaceOversizedDuration = workSpaceTooLarge ? zc->workSpaceOversizedDuration+1 : 0;
+
+            DEBUGLOG(4, "Need %zuKB workspace, including %zuKB for match state, and %zuKB for buffers",
+                        neededSpace>>10, matchStateSize>>10, bufferSpace>>10);
+            DEBUGLOG(4, "windowSize: %zu - blockSize: %zu", windowSize, blockSize);
+
+            if (workSpaceTooSmall || workSpaceWasteful) {
+                DEBUGLOG(4, "Resize workSpaceSize from %zuKB to %zuKB",
+                            zc->workSpaceSize >> 10,
+                            neededSpace >> 10);
+
+                RETURN_ERROR_IF(zc->staticSize, memory_allocation, "static cctx : no resize");
 
                 zc->workSpaceSize = 0;
                 ZSTD_free(zc->workSpace, zc->customMem);
                 zc->workSpace = ZSTD_malloc(neededSpace, zc->customMem);
-                if (zc->workSpace == NULL) return ERROR(memory_allocation);
+                RETURN_ERROR_IF(zc->workSpace == NULL, memory_allocation);
                 zc->workSpaceSize = neededSpace;
-                ptr = zc->workSpace;
+                zc->workSpaceOversizedDuration = 0;
 
-                /* Statically sized space. entropyWorkspace never moves (but prev/next block swap places) */
+                /* Statically sized space.
+                 * entropyWorkspace never moves,
+                 * though prev/next block swap places */
                 assert(((size_t)zc->workSpace & 3) == 0);   /* ensure correct alignment */
                 assert(zc->workSpaceSize >= 2 * sizeof(ZSTD_compressedBlockState_t));
                 zc->blockState.prevCBlock = (ZSTD_compressedBlockState_t*)zc->workSpace;
@@ -1022,13 +1513,14 @@
 
         /* init params */
         zc->appliedParams = params;
+        zc->blockState.matchState.cParams = params.cParams;
         zc->pledgedSrcSizePlusOne = pledgedSrcSize+1;
         zc->consumedSrcSize = 0;
         zc->producedCSize = 0;
         if (pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN)
             zc->appliedParams.fParams.contentSizeFlag = 0;
         DEBUGLOG(4, "pledged content size : %u ; flag : %u",
-            (U32)pledgedSrcSize, zc->appliedParams.fParams.contentSizeFlag);
+            (unsigned)pledgedSrcSize, zc->appliedParams.fParams.contentSizeFlag);
         zc->blockSize = blockSize;
 
         XXH64_reset(&zc->xxhState, 0);
@@ -1037,7 +1529,10 @@
 
         ZSTD_reset_compressedBlockState(zc->blockState.prevCBlock);
 
-        ptr = zc->entropyWorkspace + HUF_WORKSPACE_SIZE_U32;
+        ptr = ZSTD_reset_matchState(&zc->blockState.matchState,
+                                     zc->entropyWorkspace + HUF_WORKSPACE_SIZE_U32,
+                                    &params.cParams,
+                                     crp, ZSTD_resetTarget_CCtx);
 
         /* ldm hash table */
         /* initialize bucketOffsets table later for pointer alignment */
@@ -1055,16 +1550,19 @@
         }
         assert(((size_t)ptr & 3) == 0); /* ensure ptr is properly aligned */
 
-        ptr = ZSTD_reset_matchState(&zc->blockState.matchState, ptr, &params.cParams, crp, /* forCCtx */ 1);
-
         /* sequences storage */
+        zc->seqStore.maxNbSeq = maxNbSeq;
         zc->seqStore.sequencesStart = (seqDef*)ptr;
         ptr = zc->seqStore.sequencesStart + maxNbSeq;
         zc->seqStore.llCode = (BYTE*) ptr;
         zc->seqStore.mlCode = zc->seqStore.llCode + maxNbSeq;
         zc->seqStore.ofCode = zc->seqStore.mlCode + maxNbSeq;
         zc->seqStore.litStart = zc->seqStore.ofCode + maxNbSeq;
-        ptr = zc->seqStore.litStart + blockSize;
+        /* ZSTD_wildcopy() is used to copy into the literals buffer,
+         * so we have to oversize the buffer by WILDCOPY_OVERLENGTH bytes.
+         */
+        zc->seqStore.maxNbLit = blockSize;
+        ptr = zc->seqStore.litStart + blockSize + WILDCOPY_OVERLENGTH;
 
         /* ldm bucketOffsets table */
         if (params.ldmParams.enableLdm) {
@@ -1098,28 +1596,109 @@
     assert(!ZSTD_window_hasExtDict(cctx->blockState.matchState.window));
 }
 
-static size_t ZSTD_resetCCtx_usingCDict(ZSTD_CCtx* cctx,
+/* These are the approximate sizes for each strategy past which copying the
+ * dictionary tables into the working context is faster than using them
+ * in-place.
+ */
+static const size_t attachDictSizeCutoffs[ZSTD_STRATEGY_MAX+1] = {
+    8 KB,  /* unused */
+    8 KB,  /* ZSTD_fast */
+    16 KB, /* ZSTD_dfast */
+    32 KB, /* ZSTD_greedy */
+    32 KB, /* ZSTD_lazy */
+    32 KB, /* ZSTD_lazy2 */
+    32 KB, /* ZSTD_btlazy2 */
+    32 KB, /* ZSTD_btopt */
+    8 KB,  /* ZSTD_btultra */
+    8 KB   /* ZSTD_btultra2 */
+};
+
+static int ZSTD_shouldAttachDict(const ZSTD_CDict* cdict,
+                                 ZSTD_CCtx_params params,
+                                 U64 pledgedSrcSize)
+{
+    size_t cutoff = attachDictSizeCutoffs[cdict->matchState.cParams.strategy];
+    return ( pledgedSrcSize <= cutoff
+          || pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN
+          || params.attachDictPref == ZSTD_dictForceAttach )
+        && params.attachDictPref != ZSTD_dictForceCopy
+        && !params.forceWindow; /* dictMatchState isn't correctly
+                                 * handled in _enforceMaxDist */
+}
+
+static size_t
+ZSTD_resetCCtx_byAttachingCDict(ZSTD_CCtx* cctx,
+                        const ZSTD_CDict* cdict,
+                        ZSTD_CCtx_params params,
+                        U64 pledgedSrcSize,
+                        ZSTD_buffered_policy_e zbuff)
+{
+    {   const ZSTD_compressionParameters* const cdict_cParams = &cdict->matchState.cParams;
+        unsigned const windowLog = params.cParams.windowLog;
+        assert(windowLog != 0);
+        /* Resize working context table params for input only, since the dict
+         * has its own tables. */
+        params.cParams = ZSTD_adjustCParams_internal(*cdict_cParams, pledgedSrcSize, 0);
+        params.cParams.windowLog = windowLog;
+        ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize,
+                                ZSTDcrp_continue, zbuff);
+        assert(cctx->appliedParams.cParams.strategy == cdict_cParams->strategy);
+    }
+
+    {   const U32 cdictEnd = (U32)( cdict->matchState.window.nextSrc
+                                  - cdict->matchState.window.base);
+        const U32 cdictLen = cdictEnd - cdict->matchState.window.dictLimit;
+        if (cdictLen == 0) {
+            /* don't even attach dictionaries with no contents */
+            DEBUGLOG(4, "skipping attaching empty dictionary");
+        } else {
+            DEBUGLOG(4, "attaching dictionary into context");
+            cctx->blockState.matchState.dictMatchState = &cdict->matchState;
+
+            /* prep working match state so dict matches never have negative indices
+             * when they are translated to the working context's index space. */
+            if (cctx->blockState.matchState.window.dictLimit < cdictEnd) {
+                cctx->blockState.matchState.window.nextSrc =
+                    cctx->blockState.matchState.window.base + cdictEnd;
+                ZSTD_window_clear(&cctx->blockState.matchState.window);
+            }
+            /* loadedDictEnd is expressed within the referential of the active context */
+            cctx->blockState.matchState.loadedDictEnd = cctx->blockState.matchState.window.dictLimit;
+    }   }
+
+    cctx->dictID = cdict->dictID;
+
+    /* copy block state */
+    memcpy(cctx->blockState.prevCBlock, &cdict->cBlockState, sizeof(cdict->cBlockState));
+
+    return 0;
+}
+
+static size_t ZSTD_resetCCtx_byCopyingCDict(ZSTD_CCtx* cctx,
                             const ZSTD_CDict* cdict,
-                            unsigned windowLog,
-                            ZSTD_frameParameters fParams,
+                            ZSTD_CCtx_params params,
                             U64 pledgedSrcSize,
                             ZSTD_buffered_policy_e zbuff)
 {
-    {   ZSTD_CCtx_params params = cctx->requestedParams;
+    const ZSTD_compressionParameters *cdict_cParams = &cdict->matchState.cParams;
+
+    DEBUGLOG(4, "copying dictionary into context");
+
+    {   unsigned const windowLog = params.cParams.windowLog;
+        assert(windowLog != 0);
         /* Copy only compression parameters related to tables. */
-        params.cParams = cdict->cParams;
-        if (windowLog) params.cParams.windowLog = windowLog;
-        params.fParams = fParams;
+        params.cParams = *cdict_cParams;
+        params.cParams.windowLog = windowLog;
         ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize,
                                 ZSTDcrp_noMemset, zbuff);
-        assert(cctx->appliedParams.cParams.strategy == cdict->cParams.strategy);
-        assert(cctx->appliedParams.cParams.hashLog == cdict->cParams.hashLog);
-        assert(cctx->appliedParams.cParams.chainLog == cdict->cParams.chainLog);
+        assert(cctx->appliedParams.cParams.strategy == cdict_cParams->strategy);
+        assert(cctx->appliedParams.cParams.hashLog == cdict_cParams->hashLog);
+        assert(cctx->appliedParams.cParams.chainLog == cdict_cParams->chainLog);
     }
 
     /* copy tables */
-    {   size_t const chainSize = (cdict->cParams.strategy == ZSTD_fast) ? 0 : ((size_t)1 << cdict->cParams.chainLog);
-        size_t const hSize =  (size_t)1 << cdict->cParams.hashLog;
+    {   size_t const chainSize = (cdict_cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cdict_cParams->chainLog);
+        size_t const hSize =  (size_t)1 << cdict_cParams->hashLog;
         size_t const tableSpace = (chainSize + hSize) * sizeof(U32);
         assert((U32*)cctx->blockState.matchState.chainTable == (U32*)cctx->blockState.matchState.hashTable + hSize);  /* chainTable must follow hashTable */
         assert((U32*)cctx->blockState.matchState.hashTable3 == (U32*)cctx->blockState.matchState.chainTable + chainSize);
@@ -1127,6 +1706,7 @@
         assert((U32*)cdict->matchState.hashTable3 == (U32*)cdict->matchState.chainTable + chainSize);
         memcpy(cctx->blockState.matchState.hashTable, cdict->matchState.hashTable, tableSpace);   /* presumes all tables follow each other */
     }
+
     /* Zero the hashTable3, since the cdict never fills it */
     {   size_t const h3Size = (size_t)1 << cctx->blockState.matchState.hashLog3;
         assert(cdict->matchState.hashLog3 == 0);
@@ -1134,14 +1714,13 @@
     }
 
     /* copy dictionary offsets */
-    {
-        ZSTD_matchState_t const* srcMatchState = &cdict->matchState;
+    {   ZSTD_matchState_t const* srcMatchState = &cdict->matchState;
         ZSTD_matchState_t* dstMatchState = &cctx->blockState.matchState;
         dstMatchState->window       = srcMatchState->window;
         dstMatchState->nextToUpdate = srcMatchState->nextToUpdate;
-        dstMatchState->nextToUpdate3= srcMatchState->nextToUpdate3;
         dstMatchState->loadedDictEnd= srcMatchState->loadedDictEnd;
     }
+
     cctx->dictID = cdict->dictID;
 
     /* copy block state */
@@ -1150,6 +1729,28 @@
     return 0;
 }
 
+/* We have a choice between copying the dictionary context into the working
+ * context, or referencing the dictionary context from the working context
+ * in-place. We decide here which strategy to use. */
+static size_t ZSTD_resetCCtx_usingCDict(ZSTD_CCtx* cctx,
+                            const ZSTD_CDict* cdict,
+                            ZSTD_CCtx_params params,
+                            U64 pledgedSrcSize,
+                            ZSTD_buffered_policy_e zbuff)
+{
+
+    DEBUGLOG(4, "ZSTD_resetCCtx_usingCDict (pledgedSrcSize=%u)",
+                (unsigned)pledgedSrcSize);
+
+    if (ZSTD_shouldAttachDict(cdict, params, pledgedSrcSize)) {
+        return ZSTD_resetCCtx_byAttachingCDict(
+            cctx, cdict, params, pledgedSrcSize, zbuff);
+    } else {
+        return ZSTD_resetCCtx_byCopyingCDict(
+            cctx, cdict, params, pledgedSrcSize, zbuff);
+    }
+}
+
 /*! ZSTD_copyCCtx_internal() :
  *  Duplicate an existing context `srcCCtx` into another one `dstCCtx`.
  *  Only works during stage ZSTDcs_init (i.e. after creation, but before first call to ZSTD_compressContinue()).
@@ -1164,7 +1765,7 @@
                             ZSTD_buffered_policy_e zbuff)
 {
     DEBUGLOG(5, "ZSTD_copyCCtx_internal");
-    if (srcCCtx->stage!=ZSTDcs_init) return ERROR(stage_wrong);
+    RETURN_ERROR_IF(srcCCtx->stage!=ZSTDcs_init, stage_wrong);
 
     memcpy(&dstCCtx->customMem, &srcCCtx->customMem, sizeof(ZSTD_customMem));
     {   ZSTD_CCtx_params params = dstCCtx->requestedParams;
@@ -1192,11 +1793,10 @@
 
     /* copy dictionary offsets */
     {
-        ZSTD_matchState_t const* srcMatchState = &srcCCtx->blockState.matchState;
+        const ZSTD_matchState_t* srcMatchState = &srcCCtx->blockState.matchState;
         ZSTD_matchState_t* dstMatchState = &dstCCtx->blockState.matchState;
         dstMatchState->window       = srcMatchState->window;
         dstMatchState->nextToUpdate = srcMatchState->nextToUpdate;
-        dstMatchState->nextToUpdate3= srcMatchState->nextToUpdate3;
         dstMatchState->loadedDictEnd= srcMatchState->loadedDictEnd;
     }
     dstCCtx->dictID = srcCCtx->dictID;
@@ -1266,16 +1866,15 @@
 
 /*! ZSTD_reduceIndex() :
 *   rescale all indexes to avoid future overflow (indexes are U32) */
-static void ZSTD_reduceIndex (ZSTD_CCtx* zc, const U32 reducerValue)
+static void ZSTD_reduceIndex (ZSTD_matchState_t* ms, ZSTD_CCtx_params const* params, const U32 reducerValue)
 {
-    ZSTD_matchState_t* const ms = &zc->blockState.matchState;
-    {   U32 const hSize = (U32)1 << zc->appliedParams.cParams.hashLog;
+    {   U32 const hSize = (U32)1 << params->cParams.hashLog;
         ZSTD_reduceTable(ms->hashTable, hSize, reducerValue);
     }
 
-    if (zc->appliedParams.cParams.strategy != ZSTD_fast) {
-        U32 const chainSize = (U32)1 << zc->appliedParams.cParams.chainLog;
-        if (zc->appliedParams.cParams.strategy == ZSTD_btlazy2)
+    if (params->cParams.strategy != ZSTD_fast) {
+        U32 const chainSize = (U32)1 << params->cParams.chainLog;
+        if (params->cParams.strategy == ZSTD_btlazy2)
             ZSTD_reduceTable_btlazy2(ms->chainTable, chainSize, reducerValue);
         else
             ZSTD_reduceTable(ms->chainTable, chainSize, reducerValue);
@@ -1294,21 +1893,22 @@
 
 /* See doc/zstd_compression_format.md for detailed format description */
 
-size_t ZSTD_noCompressBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
+static size_t ZSTD_noCompressBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize, U32 lastBlock)
 {
-    if (srcSize + ZSTD_blockHeaderSize > dstCapacity) return ERROR(dstSize_tooSmall);
+    U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw)<<1) + (U32)(srcSize << 3);
+    RETURN_ERROR_IF(srcSize + ZSTD_blockHeaderSize > dstCapacity,
+                    dstSize_tooSmall);
+    MEM_writeLE24(dst, cBlockHeader24);
     memcpy((BYTE*)dst + ZSTD_blockHeaderSize, src, srcSize);
-    MEM_writeLE24(dst, (U32)(srcSize << 2) + (U32)bt_raw);
-    return ZSTD_blockHeaderSize+srcSize;
+    return ZSTD_blockHeaderSize + srcSize;
 }
 
-
 static size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
 {
     BYTE* const ostart = (BYTE* const)dst;
     U32   const flSize = 1 + (srcSize>31) + (srcSize>4095);
 
-    if (srcSize + flSize > dstCapacity) return ERROR(dstSize_tooSmall);
+    RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall);
 
     switch(flSize)
     {
@@ -1356,16 +1956,27 @@
 }
 
 
-static size_t ZSTD_minGain(size_t srcSize) { return (srcSize >> 6) + 2; }
+/* ZSTD_minGain() :
+ * minimum compression required
+ * to generate a compress block or a compressed literals section.
+ * note : use same formula for both situations */
+static size_t ZSTD_minGain(size_t srcSize, ZSTD_strategy strat)
+{
+    U32 const minlog = (strat>=ZSTD_btultra) ? (U32)(strat) - 1 : 6;
+    ZSTD_STATIC_ASSERT(ZSTD_btultra == 8);
+    assert(ZSTD_cParam_withinBounds(ZSTD_c_strategy, strat));
+    return (srcSize >> minlog) + 2;
+}
 
-static size_t ZSTD_compressLiterals (ZSTD_entropyCTables_t const* prevEntropy,
-                                     ZSTD_entropyCTables_t* nextEntropy,
+static size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
+                                     ZSTD_hufCTables_t* nextHuf,
                                      ZSTD_strategy strategy, int disableLiteralCompression,
                                      void* dst, size_t dstCapacity,
                                const void* src, size_t srcSize,
-                                     U32* workspace, const int bmi2)
+                                     void* workspace, size_t wkspSize,
+                               const int bmi2)
 {
-    size_t const minGain = ZSTD_minGain(srcSize);
+    size_t const minGain = ZSTD_minGain(srcSize, strategy);
     size_t const lhSize = 3 + (srcSize >= 1 KB) + (srcSize >= 16 KB);
     BYTE*  const ostart = (BYTE*)dst;
     U32 singleStream = srcSize < 256;
@@ -1376,27 +1987,25 @@
                 disableLiteralCompression);
 
     /* Prepare nextEntropy assuming reusing the existing table */
-    nextEntropy->hufCTable_repeatMode = prevEntropy->hufCTable_repeatMode;
-    memcpy(nextEntropy->hufCTable, prevEntropy->hufCTable,
-           sizeof(prevEntropy->hufCTable));
+    memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
 
     if (disableLiteralCompression)
         return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
 
     /* small ? don't even attempt compression (speed opt) */
 #   define COMPRESS_LITERALS_SIZE_MIN 63
-    {   size_t const minLitSize = (prevEntropy->hufCTable_repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
+    {   size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
         if (srcSize <= minLitSize) return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
     }
 
-    if (dstCapacity < lhSize+1) return ERROR(dstSize_tooSmall);   /* not enough space for compression */
-    {   HUF_repeat repeat = prevEntropy->hufCTable_repeatMode;
+    RETURN_ERROR_IF(dstCapacity < lhSize+1, dstSize_tooSmall, "not enough space for compression");
+    {   HUF_repeat repeat = prevHuf->repeatMode;
         int const preferRepeat = strategy < ZSTD_lazy ? srcSize <= 1024 : 0;
         if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1;
         cLitSize = singleStream ? HUF_compress1X_repeat(ostart+lhSize, dstCapacity-lhSize, src, srcSize, 255, 11,
-                                      workspace, HUF_WORKSPACE_SIZE, (HUF_CElt*)nextEntropy->hufCTable, &repeat, preferRepeat, bmi2)
+                                      workspace, wkspSize, (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2)
                                 : HUF_compress4X_repeat(ostart+lhSize, dstCapacity-lhSize, src, srcSize, 255, 11,
-                                      workspace, HUF_WORKSPACE_SIZE, (HUF_CElt*)nextEntropy->hufCTable, &repeat, preferRepeat, bmi2);
+                                      workspace, wkspSize, (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2);
         if (repeat != HUF_repeat_none) {
             /* reused the existing table */
             hType = set_repeat;
@@ -1404,17 +2013,17 @@
     }
 
     if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) {
-        memcpy(nextEntropy->hufCTable, prevEntropy->hufCTable, sizeof(prevEntropy->hufCTable));
+        memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
         return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
     }
     if (cLitSize==1) {
-        memcpy(nextEntropy->hufCTable, prevEntropy->hufCTable, sizeof(prevEntropy->hufCTable));
+        memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
         return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
     }
 
     if (hType == set_compressed) {
         /* using a newly constructed table */
-        nextEntropy->hufCTable_repeatMode = HUF_repeat_check;
+        nextHuf->repeatMode = HUF_repeat_check;
     }
 
     /* Build header */
@@ -1451,6 +2060,7 @@
     BYTE* const mlCodeTable = seqStorePtr->mlCode;
     U32 const nbSeq = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart);
     U32 u;
+    assert(nbSeq <= seqStorePtr->maxNbSeq);
     for (u=0; u<nbSeq; u++) {
         U32 const llv = sequences[u].litLength;
         U32 const mlv = sequences[u].matchLength;
@@ -1464,72 +2074,243 @@
         mlCodeTable[seqStorePtr->longLengthPos] = MaxML;
 }
 
+
+/**
+ * -log2(x / 256) lookup table for x in [0, 256).
+ * If x == 0: Return 0
+ * Else: Return floor(-log2(x / 256) * 256)
+ */
+static unsigned const kInverseProbabilityLog256[256] = {
+    0,    2048, 1792, 1642, 1536, 1453, 1386, 1329, 1280, 1236, 1197, 1162,
+    1130, 1100, 1073, 1047, 1024, 1001, 980,  960,  941,  923,  906,  889,
+    874,  859,  844,  830,  817,  804,  791,  779,  768,  756,  745,  734,
+    724,  714,  704,  694,  685,  676,  667,  658,  650,  642,  633,  626,
+    618,  610,  603,  595,  588,  581,  574,  567,  561,  554,  548,  542,
+    535,  529,  523,  517,  512,  506,  500,  495,  489,  484,  478,  473,
+    468,  463,  458,  453,  448,  443,  438,  434,  429,  424,  420,  415,
+    411,  407,  402,  398,  394,  390,  386,  382,  377,  373,  370,  366,
+    362,  358,  354,  350,  347,  343,  339,  336,  332,  329,  325,  322,
+    318,  315,  311,  308,  305,  302,  298,  295,  292,  289,  286,  282,
+    279,  276,  273,  270,  267,  264,  261,  258,  256,  253,  250,  247,
+    244,  241,  239,  236,  233,  230,  228,  225,  222,  220,  217,  215,
+    212,  209,  207,  204,  202,  199,  197,  194,  192,  190,  187,  185,
+    182,  180,  178,  175,  173,  171,  168,  166,  164,  162,  159,  157,
+    155,  153,  151,  149,  146,  144,  142,  140,  138,  136,  134,  132,
+    130,  128,  126,  123,  121,  119,  117,  115,  114,  112,  110,  108,
+    106,  104,  102,  100,  98,   96,   94,   93,   91,   89,   87,   85,
+    83,   82,   80,   78,   76,   74,   73,   71,   69,   67,   66,   64,
+    62,   61,   59,   57,   55,   54,   52,   50,   49,   47,   46,   44,
+    42,   41,   39,   37,   36,   34,   33,   31,   30,   28,   26,   25,
+    23,   22,   20,   19,   17,   16,   14,   13,   11,   10,   8,    7,
+    5,    4,    2,    1,
+};
+
+
+/**
+ * Returns the cost in bits of encoding the distribution described by count
+ * using the entropy bound.
+ */
+static size_t ZSTD_entropyCost(unsigned const* count, unsigned const max, size_t const total)
+{
+    unsigned cost = 0;
+    unsigned s;
+    for (s = 0; s <= max; ++s) {
+        unsigned norm = (unsigned)((256 * count[s]) / total);
+        if (count[s] != 0 && norm == 0)
+            norm = 1;
+        assert(count[s] < total);
+        cost += count[s] * kInverseProbabilityLog256[norm];
+    }
+    return cost >> 8;
+}
+
+
+/**
+ * Returns the cost in bits of encoding the distribution in count using the
+ * table described by norm. The max symbol support by norm is assumed >= max.
+ * norm must be valid for every symbol with non-zero probability in count.
+ */
+static size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
+                                    unsigned const* count, unsigned const max)
+{
+    unsigned const shift = 8 - accuracyLog;
+    size_t cost = 0;
+    unsigned s;
+    assert(accuracyLog <= 8);
+    for (s = 0; s <= max; ++s) {
+        unsigned const normAcc = norm[s] != -1 ? norm[s] : 1;
+        unsigned const norm256 = normAcc << shift;
+        assert(norm256 > 0);
+        assert(norm256 < 256);
+        cost += count[s] * kInverseProbabilityLog256[norm256];
+    }
+    return cost >> 8;
+}
+
+
+static unsigned ZSTD_getFSEMaxSymbolValue(FSE_CTable const* ctable) {
+  void const* ptr = ctable;
+  U16 const* u16ptr = (U16 const*)ptr;
+  U32 const maxSymbolValue = MEM_read16(u16ptr + 1);
+  return maxSymbolValue;
+}
+
+
+/**
+ * Returns the cost in bits of encoding the distribution in count using ctable.
+ * Returns an error if ctable cannot represent all the symbols in count.
+ */
+static size_t ZSTD_fseBitCost(
+    FSE_CTable const* ctable,
+    unsigned const* count,
+    unsigned const max)
+{
+    unsigned const kAccuracyLog = 8;
+    size_t cost = 0;
+    unsigned s;
+    FSE_CState_t cstate;
+    FSE_initCState(&cstate, ctable);
+    RETURN_ERROR_IF(ZSTD_getFSEMaxSymbolValue(ctable) < max, GENERIC,
+                    "Repeat FSE_CTable has maxSymbolValue %u < %u",
+                    ZSTD_getFSEMaxSymbolValue(ctable), max);
+    for (s = 0; s <= max; ++s) {
+        unsigned const tableLog = cstate.stateLog;
+        unsigned const badCost = (tableLog + 1) << kAccuracyLog;
+        unsigned const bitCost = FSE_bitCost(cstate.symbolTT, tableLog, s, kAccuracyLog);
+        if (count[s] == 0)
+            continue;
+        RETURN_ERROR_IF(bitCost >= badCost, GENERIC,
+                        "Repeat FSE_CTable has Prob[%u] == 0", s);
+        cost += count[s] * bitCost;
+    }
+    return cost >> kAccuracyLog;
+}
+
+/**
+ * Returns the cost in bytes of encoding the normalized count header.
+ * Returns an error if any of the helper functions return an error.
+ */
+static size_t ZSTD_NCountCost(unsigned const* count, unsigned const max,
+                              size_t const nbSeq, unsigned const FSELog)
+{
+    BYTE wksp[FSE_NCOUNTBOUND];
+    S16 norm[MaxSeq + 1];
+    const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max);
+    FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq, max));
+    return FSE_writeNCount(wksp, sizeof(wksp), norm, max, tableLog);
+}
+
+
 typedef enum {
     ZSTD_defaultDisallowed = 0,
     ZSTD_defaultAllowed = 1
 } ZSTD_defaultPolicy_e;
 
-MEM_STATIC
-symbolEncodingType_e ZSTD_selectEncodingType(
-        FSE_repeat* repeatMode, size_t const mostFrequent, size_t nbSeq,
-        U32 defaultNormLog, ZSTD_defaultPolicy_e const isDefaultAllowed)
+MEM_STATIC symbolEncodingType_e
+ZSTD_selectEncodingType(
+        FSE_repeat* repeatMode, unsigned const* count, unsigned const max,
+        size_t const mostFrequent, size_t nbSeq, unsigned const FSELog,
+        FSE_CTable const* prevCTable,
+        short const* defaultNorm, U32 defaultNormLog,
+        ZSTD_defaultPolicy_e const isDefaultAllowed,
+        ZSTD_strategy const strategy)
 {
-#define MIN_SEQ_FOR_DYNAMIC_FSE   64
-#define MAX_SEQ_FOR_STATIC_FSE  1000
     ZSTD_STATIC_ASSERT(ZSTD_defaultDisallowed == 0 && ZSTD_defaultAllowed != 0);
-    if ((mostFrequent == nbSeq) && (!isDefaultAllowed || nbSeq > 2)) {
+    if (mostFrequent == nbSeq) {
+        *repeatMode = FSE_repeat_none;
+        if (isDefaultAllowed && nbSeq <= 2) {
+            /* Prefer set_basic over set_rle when there are 2 or less symbols,
+             * since RLE uses 1 byte, but set_basic uses 5-6 bits per symbol.
+             * If basic encoding isn't possible, always choose RLE.
+             */
+            DEBUGLOG(5, "Selected set_basic");
+            return set_basic;
+        }
         DEBUGLOG(5, "Selected set_rle");
-        /* Prefer set_basic over set_rle when there are 2 or less symbols,
-         * since RLE uses 1 byte, but set_basic uses 5-6 bits per symbol.
-         * If basic encoding isn't possible, always choose RLE.
-         */
-        *repeatMode = FSE_repeat_check;
         return set_rle;
     }
-    if ( isDefaultAllowed
-      && (*repeatMode == FSE_repeat_valid) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) {
-        DEBUGLOG(5, "Selected set_repeat");
-        return set_repeat;
-    }
-    if ( isDefaultAllowed
-      && ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (defaultNormLog-1)))) ) {
-        DEBUGLOG(5, "Selected set_basic");
-        /* The format allows default tables to be repeated, but it isn't useful.
-         * When using simple heuristics to select encoding type, we don't want
-         * to confuse these tables with dictionaries. When running more careful
-         * analysis, we don't need to waste time checking both repeating tables
-         * and default tables.
-         */
-        *repeatMode = FSE_repeat_none;
-        return set_basic;
+    if (strategy < ZSTD_lazy) {
+        if (isDefaultAllowed) {
+            size_t const staticFse_nbSeq_max = 1000;
+            size_t const mult = 10 - strategy;
+            size_t const baseLog = 3;
+            size_t const dynamicFse_nbSeq_min = (((size_t)1 << defaultNormLog) * mult) >> baseLog;  /* 28-36 for offset, 56-72 for lengths */
+            assert(defaultNormLog >= 5 && defaultNormLog <= 6);  /* xx_DEFAULTNORMLOG */
+            assert(mult <= 9 && mult >= 7);
+            if ( (*repeatMode == FSE_repeat_valid)
+              && (nbSeq < staticFse_nbSeq_max) ) {
+                DEBUGLOG(5, "Selected set_repeat");
+                return set_repeat;
+            }
+            if ( (nbSeq < dynamicFse_nbSeq_min)
+              || (mostFrequent < (nbSeq >> (defaultNormLog-1))) ) {
+                DEBUGLOG(5, "Selected set_basic");
+                /* The format allows default tables to be repeated, but it isn't useful.
+                 * When using simple heuristics to select encoding type, we don't want
+                 * to confuse these tables with dictionaries. When running more careful
+                 * analysis, we don't need to waste time checking both repeating tables
+                 * and default tables.
+                 */
+                *repeatMode = FSE_repeat_none;
+                return set_basic;
+            }
+        }
+    } else {
+        size_t const basicCost = isDefaultAllowed ? ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, count, max) : ERROR(GENERIC);
+        size_t const repeatCost = *repeatMode != FSE_repeat_none ? ZSTD_fseBitCost(prevCTable, count, max) : ERROR(GENERIC);
+        size_t const NCountCost = ZSTD_NCountCost(count, max, nbSeq, FSELog);
+        size_t const compressedCost = (NCountCost << 3) + ZSTD_entropyCost(count, max, nbSeq);
+
+        if (isDefaultAllowed) {
+            assert(!ZSTD_isError(basicCost));
+            assert(!(*repeatMode == FSE_repeat_valid && ZSTD_isError(repeatCost)));
+        }
+        assert(!ZSTD_isError(NCountCost));
+        assert(compressedCost < ERROR(maxCode));
+        DEBUGLOG(5, "Estimated bit costs: basic=%u\trepeat=%u\tcompressed=%u",
+                    (unsigned)basicCost, (unsigned)repeatCost, (unsigned)compressedCost);
+        if (basicCost <= repeatCost && basicCost <= compressedCost) {
+            DEBUGLOG(5, "Selected set_basic");
+            assert(isDefaultAllowed);
+            *repeatMode = FSE_repeat_none;
+            return set_basic;
+        }
+        if (repeatCost <= compressedCost) {
+            DEBUGLOG(5, "Selected set_repeat");
+            assert(!ZSTD_isError(repeatCost));
+            return set_repeat;
+        }
+        assert(compressedCost < basicCost && compressedCost < repeatCost);
     }
     DEBUGLOG(5, "Selected set_compressed");
     *repeatMode = FSE_repeat_check;
     return set_compressed;
 }
 
-MEM_STATIC
-size_t ZSTD_buildCTable(void* dst, size_t dstCapacity,
-        FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type,
-        U32* count, U32 max,
-        BYTE const* codeTable, size_t nbSeq,
-        S16 const* defaultNorm, U32 defaultNormLog, U32 defaultMax,
-        FSE_CTable const* prevCTable, size_t prevCTableSize,
-        void* workspace, size_t workspaceSize)
+MEM_STATIC size_t
+ZSTD_buildCTable(void* dst, size_t dstCapacity,
+                FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type,
+                unsigned* count, U32 max,
+                const BYTE* codeTable, size_t nbSeq,
+                const S16* defaultNorm, U32 defaultNormLog, U32 defaultMax,
+                const FSE_CTable* prevCTable, size_t prevCTableSize,
+                void* workspace, size_t workspaceSize)
 {
     BYTE* op = (BYTE*)dst;
-    BYTE const* const oend = op + dstCapacity;
+    const BYTE* const oend = op + dstCapacity;
+    DEBUGLOG(6, "ZSTD_buildCTable (dstCapacity=%u)", (unsigned)dstCapacity);
 
     switch (type) {
     case set_rle:
+        FORWARD_IF_ERROR(FSE_buildCTable_rle(nextCTable, (BYTE)max));
+        RETURN_ERROR_IF(dstCapacity==0, dstSize_tooSmall);
         *op = codeTable[0];
-        CHECK_F(FSE_buildCTable_rle(nextCTable, (BYTE)max));
         return 1;
     case set_repeat:
         memcpy(nextCTable, prevCTable, prevCTableSize);
         return 0;
     case set_basic:
-        CHECK_F(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, workspace, workspaceSize));  /* note : could be pre-calculated */
+        FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, workspace, workspaceSize));  /* note : could be pre-calculated */
         return 0;
     case set_compressed: {
         S16 norm[MaxSeq + 1];
@@ -1540,14 +2321,14 @@
             nbSeq_1--;
         }
         assert(nbSeq_1 > 1);
-        CHECK_F(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max));
+        FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max));
         {   size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog);   /* overflow protected */
-            if (FSE_isError(NCountSize)) return NCountSize;
-            CHECK_F(FSE_buildCTable_wksp(nextCTable, norm, max, tableLog, workspace, workspaceSize));
+            FORWARD_IF_ERROR(NCountSize);
+            FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, norm, max, tableLog, workspace, workspaceSize));
             return NCountSize;
         }
     }
-    default: return assert(0), ERROR(GENERIC);
+    default: assert(0); RETURN_ERROR(GENERIC);
     }
 }
 
@@ -1564,7 +2345,12 @@
     FSE_CState_t  stateOffsetBits;
     FSE_CState_t  stateLitLength;
 
-    CHECK_E(BIT_initCStream(&blockStream, dst, dstCapacity), dstSize_tooSmall); /* not enough space remaining */
+    RETURN_ERROR_IF(
+        ERR_isError(BIT_initCStream(&blockStream, dst, dstCapacity)),
+        dstSize_tooSmall, "not enough space remaining");
+    DEBUGLOG(6, "available space for bitstream : %i  (dstCapacity=%u)",
+                (int)(blockStream.endPtr - blockStream.startPtr),
+                (unsigned)dstCapacity);
 
     /* first symbols */
     FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq-1]);
@@ -1597,9 +2383,9 @@
             U32  const ofBits = ofCode;
             U32  const mlBits = ML_bits[mlCode];
             DEBUGLOG(6, "encoding: litlen:%2u - matchlen:%2u - offCode:%7u",
-                        sequences[n].litLength,
-                        sequences[n].matchLength + MINMATCH,
-                        sequences[n].offset);
+                        (unsigned)sequences[n].litLength,
+                        (unsigned)sequences[n].matchLength + MINMATCH,
+                        (unsigned)sequences[n].offset);
                                                                             /* 32b*/  /* 64b*/
                                                                             /* (7)*/  /* (7)*/
             FSE_encodeSymbol(&blockStream, &stateOffsetBits, ofCode);       /* 15 */  /* 15 */
@@ -1624,6 +2410,7 @@
                 BIT_addBits(&blockStream, sequences[n].offset, ofBits);     /* 31 */
             }
             BIT_flushBits(&blockStream);                                    /* (7)*/
+            DEBUGLOG(7, "remaining space : %i", (int)(blockStream.endPtr - blockStream.ptr));
     }   }
 
     DEBUGLOG(6, "ZSTD_encodeSequences: flushing ML state with %u bits", stateMatchLength.stateLog);
@@ -1634,7 +2421,7 @@
     FSE_flushCState(&blockStream, &stateLitLength);
 
     {   size_t const streamSize = BIT_closeCStream(&blockStream);
-        if (streamSize==0) return ERROR(dstSize_tooSmall);   /* not enough space */
+        RETURN_ERROR_IF(streamSize==0, dstSize_tooSmall, "not enough space");
         return streamSize;
     }
 }
@@ -1674,13 +2461,14 @@
 
 #endif
 
-size_t ZSTD_encodeSequences(
+static size_t ZSTD_encodeSequences(
             void* dst, size_t dstCapacity,
             FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
             FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
             FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
             seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2)
 {
+    DEBUGLOG(5, "ZSTD_encodeSequences: dstCapacity = %u", (unsigned)dstCapacity);
 #if DYNAMIC_BMI2
     if (bmi2) {
         return ZSTD_encodeSequences_bmi2(dst, dstCapacity,
@@ -1698,18 +2486,38 @@
                                         sequences, nbSeq, longOffsets);
 }
 
-MEM_STATIC size_t ZSTD_compressSequences_internal(seqStore_t* seqStorePtr,
-                              ZSTD_entropyCTables_t const* prevEntropy,
-                              ZSTD_entropyCTables_t* nextEntropy,
-                              ZSTD_CCtx_params const* cctxParams,
-                              void* dst, size_t dstCapacity, U32* workspace,
-                              const int bmi2)
+static int ZSTD_disableLiteralsCompression(const ZSTD_CCtx_params* cctxParams)
+{
+    switch (cctxParams->literalCompressionMode) {
+    case ZSTD_lcm_huffman:
+        return 0;
+    case ZSTD_lcm_uncompressed:
+        return 1;
+    default:
+        assert(0 /* impossible: pre-validated */);
+        /* fall-through */
+    case ZSTD_lcm_auto:
+        return (cctxParams->cParams.strategy == ZSTD_fast) && (cctxParams->cParams.targetLength > 0);
+    }
+}
+
+/* ZSTD_compressSequences_internal():
+ * actually compresses both literals and sequences */
+MEM_STATIC size_t
+ZSTD_compressSequences_internal(seqStore_t* seqStorePtr,
+                          const ZSTD_entropyCTables_t* prevEntropy,
+                                ZSTD_entropyCTables_t* nextEntropy,
+                          const ZSTD_CCtx_params* cctxParams,
+                                void* dst, size_t dstCapacity,
+                                void* workspace, size_t wkspSize,
+                          const int bmi2)
 {
     const int longOffsets = cctxParams->cParams.windowLog > STREAM_ACCUMULATOR_MIN;
-    U32 count[MaxSeq+1];
-    FSE_CTable* CTable_LitLength = nextEntropy->litlengthCTable;
-    FSE_CTable* CTable_OffsetBits = nextEntropy->offcodeCTable;
-    FSE_CTable* CTable_MatchLength = nextEntropy->matchlengthCTable;
+    ZSTD_strategy const strategy = cctxParams->cParams.strategy;
+    unsigned count[MaxSeq+1];
+    FSE_CTable* CTable_LitLength = nextEntropy->fse.litlengthCTable;
+    FSE_CTable* CTable_OffsetBits = nextEntropy->fse.offcodeCTable;
+    FSE_CTable* CTable_MatchLength = nextEntropy->fse.matchlengthCTable;
     U32 LLtype, Offtype, MLtype;   /* compressed, raw or rle */
     const seqDef* const sequences = seqStorePtr->sequencesStart;
     const BYTE* const ofCodeTable = seqStorePtr->ofCode;
@@ -1720,87 +2528,114 @@
     BYTE* op = ostart;
     size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart;
     BYTE* seqHead;
+    BYTE* lastNCount = NULL;
 
     ZSTD_STATIC_ASSERT(HUF_WORKSPACE_SIZE >= (1<<MAX(MLFSELog,LLFSELog)));
+    DEBUGLOG(5, "ZSTD_compressSequences_internal");
 
     /* Compress literals */
     {   const BYTE* const literals = seqStorePtr->litStart;
         size_t const litSize = seqStorePtr->lit - literals;
         size_t const cSize = ZSTD_compressLiterals(
-                                    prevEntropy, nextEntropy,
-                                    cctxParams->cParams.strategy, cctxParams->disableLiteralCompression,
+                                    &prevEntropy->huf, &nextEntropy->huf,
+                                    cctxParams->cParams.strategy,
+                                    ZSTD_disableLiteralsCompression(cctxParams),
                                     op, dstCapacity,
                                     literals, litSize,
-                                    workspace, bmi2);
-        if (ZSTD_isError(cSize))
-          return cSize;
+                                    workspace, wkspSize,
+                                    bmi2);
+        FORWARD_IF_ERROR(cSize);
         assert(cSize <= dstCapacity);
         op += cSize;
     }
 
     /* Sequences Header */
-    if ((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/) return ERROR(dstSize_tooSmall);
+    RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/,
+                    dstSize_tooSmall);
     if (nbSeq < 0x7F)
         *op++ = (BYTE)nbSeq;
     else if (nbSeq < LONGNBSEQ)
         op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2;
     else
         op[0]=0xFF, MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)), op+=3;
+    assert(op <= oend);
     if (nbSeq==0) {
-      memcpy(nextEntropy->litlengthCTable, prevEntropy->litlengthCTable, sizeof(prevEntropy->litlengthCTable));
-      nextEntropy->litlength_repeatMode = prevEntropy->litlength_repeatMode;
-      memcpy(nextEntropy->offcodeCTable, prevEntropy->offcodeCTable, sizeof(prevEntropy->offcodeCTable));
-      nextEntropy->offcode_repeatMode = prevEntropy->offcode_repeatMode;
-      memcpy(nextEntropy->matchlengthCTable, prevEntropy->matchlengthCTable, sizeof(prevEntropy->matchlengthCTable));
-      nextEntropy->matchlength_repeatMode = prevEntropy->matchlength_repeatMode;
-      return op - ostart;
+        /* Copy the old tables over as if we repeated them */
+        memcpy(&nextEntropy->fse, &prevEntropy->fse, sizeof(prevEntropy->fse));
+        return op - ostart;
     }
 
     /* seqHead : flags for FSE encoding type */
     seqHead = op++;
+    assert(op <= oend);
 
     /* convert length/distances into codes */
     ZSTD_seqToCodes(seqStorePtr);
     /* build CTable for Literal Lengths */
-    {   U32 max = MaxLL;
-        size_t const mostFrequent = FSE_countFast_wksp(count, &max, llCodeTable, nbSeq, workspace);
+    {   unsigned max = MaxLL;
+        size_t const mostFrequent = HIST_countFast_wksp(count, &max, llCodeTable, nbSeq, workspace, wkspSize);   /* can't fail */
         DEBUGLOG(5, "Building LL table");
-        nextEntropy->litlength_repeatMode = prevEntropy->litlength_repeatMode;
-        LLtype = ZSTD_selectEncodingType(&nextEntropy->litlength_repeatMode, mostFrequent, nbSeq, LL_defaultNormLog, ZSTD_defaultAllowed);
+        nextEntropy->fse.litlength_repeatMode = prevEntropy->fse.litlength_repeatMode;
+        LLtype = ZSTD_selectEncodingType(&nextEntropy->fse.litlength_repeatMode,
+                                        count, max, mostFrequent, nbSeq,
+                                        LLFSELog, prevEntropy->fse.litlengthCTable,
+                                        LL_defaultNorm, LL_defaultNormLog,
+                                        ZSTD_defaultAllowed, strategy);
+        assert(set_basic < set_compressed && set_rle < set_compressed);
+        assert(!(LLtype < set_compressed && nextEntropy->fse.litlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
         {   size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype,
-                    count, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL,
-                    prevEntropy->litlengthCTable, sizeof(prevEntropy->litlengthCTable),
-                    workspace, HUF_WORKSPACE_SIZE);
-            if (ZSTD_isError(countSize)) return countSize;
+                                                    count, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL,
+                                                    prevEntropy->fse.litlengthCTable, sizeof(prevEntropy->fse.litlengthCTable),
+                                                    workspace, wkspSize);
+            FORWARD_IF_ERROR(countSize);
+            if (LLtype == set_compressed)
+                lastNCount = op;
             op += countSize;
+            assert(op <= oend);
     }   }
     /* build CTable for Offsets */
-    {   U32 max = MaxOff;
-        size_t const mostFrequent = FSE_countFast_wksp(count, &max, ofCodeTable, nbSeq, workspace);
+    {   unsigned max = MaxOff;
+        size_t const mostFrequent = HIST_countFast_wksp(count, &max, ofCodeTable, nbSeq, workspace, wkspSize);  /* can't fail */
         /* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */
         ZSTD_defaultPolicy_e const defaultPolicy = (max <= DefaultMaxOff) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed;
         DEBUGLOG(5, "Building OF table");
-        nextEntropy->offcode_repeatMode = prevEntropy->offcode_repeatMode;
-        Offtype = ZSTD_selectEncodingType(&nextEntropy->offcode_repeatMode, mostFrequent, nbSeq, OF_defaultNormLog, defaultPolicy);
+        nextEntropy->fse.offcode_repeatMode = prevEntropy->fse.offcode_repeatMode;
+        Offtype = ZSTD_selectEncodingType(&nextEntropy->fse.offcode_repeatMode,
+                                        count, max, mostFrequent, nbSeq,
+                                        OffFSELog, prevEntropy->fse.offcodeCTable,
+                                        OF_defaultNorm, OF_defaultNormLog,
+                                        defaultPolicy, strategy);
+        assert(!(Offtype < set_compressed && nextEntropy->fse.offcode_repeatMode != FSE_repeat_none)); /* We don't copy tables */
         {   size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype,
-                    count, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff,
-                    prevEntropy->offcodeCTable, sizeof(prevEntropy->offcodeCTable),
-                    workspace, HUF_WORKSPACE_SIZE);
-            if (ZSTD_isError(countSize)) return countSize;
+                                                    count, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff,
+                                                    prevEntropy->fse.offcodeCTable, sizeof(prevEntropy->fse.offcodeCTable),
+                                                    workspace, wkspSize);
+            FORWARD_IF_ERROR(countSize);
+            if (Offtype == set_compressed)
+                lastNCount = op;
             op += countSize;
+            assert(op <= oend);
     }   }
     /* build CTable for MatchLengths */
-    {   U32 max = MaxML;
-        size_t const mostFrequent = FSE_countFast_wksp(count, &max, mlCodeTable, nbSeq, workspace);
-        DEBUGLOG(5, "Building ML table");
-        nextEntropy->matchlength_repeatMode = prevEntropy->matchlength_repeatMode;
-        MLtype = ZSTD_selectEncodingType(&nextEntropy->matchlength_repeatMode, mostFrequent, nbSeq, ML_defaultNormLog, ZSTD_defaultAllowed);
+    {   unsigned max = MaxML;
+        size_t const mostFrequent = HIST_countFast_wksp(count, &max, mlCodeTable, nbSeq, workspace, wkspSize);   /* can't fail */
+        DEBUGLOG(5, "Building ML table (remaining space : %i)", (int)(oend-op));
+        nextEntropy->fse.matchlength_repeatMode = prevEntropy->fse.matchlength_repeatMode;
+        MLtype = ZSTD_selectEncodingType(&nextEntropy->fse.matchlength_repeatMode,
+                                        count, max, mostFrequent, nbSeq,
+                                        MLFSELog, prevEntropy->fse.matchlengthCTable,
+                                        ML_defaultNorm, ML_defaultNormLog,
+                                        ZSTD_defaultAllowed, strategy);
+        assert(!(MLtype < set_compressed && nextEntropy->fse.matchlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
         {   size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype,
-                    count, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML,
-                    prevEntropy->matchlengthCTable, sizeof(prevEntropy->matchlengthCTable),
-                    workspace, HUF_WORKSPACE_SIZE);
-            if (ZSTD_isError(countSize)) return countSize;
+                                                    count, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML,
+                                                    prevEntropy->fse.matchlengthCTable, sizeof(prevEntropy->fse.matchlengthCTable),
+                                                    workspace, wkspSize);
+            FORWARD_IF_ERROR(countSize);
+            if (MLtype == set_compressed)
+                lastNCount = op;
             op += countSize;
+            assert(op <= oend);
     }   }
 
     *seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2));
@@ -1812,65 +2647,104 @@
                                         CTable_LitLength, llCodeTable,
                                         sequences, nbSeq,
                                         longOffsets, bmi2);
-        if (ZSTD_isError(bitstreamSize)) return bitstreamSize;
+        FORWARD_IF_ERROR(bitstreamSize);
         op += bitstreamSize;
+        assert(op <= oend);
+        /* zstd versions <= 1.3.4 mistakenly report corruption when
+         * FSE_readNCount() receives a buffer < 4 bytes.
+         * Fixed by https://github.com/facebook/zstd/pull/1146.
+         * This can happen when the last set_compressed table present is 2
+         * bytes and the bitstream is only one byte.
+         * In this exceedingly rare case, we will simply emit an uncompressed
+         * block, since it isn't worth optimizing.
+         */
+        if (lastNCount && (op - lastNCount) < 4) {
+            /* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
+            assert(op - lastNCount == 3);
+            DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by "
+                        "emitting an uncompressed block.");
+            return 0;
+        }
     }
 
+    DEBUGLOG(5, "compressed block size : %u", (unsigned)(op - ostart));
     return op - ostart;
 }
 
-MEM_STATIC size_t ZSTD_compressSequences(seqStore_t* seqStorePtr,
-                              ZSTD_entropyCTables_t const* prevEntropy,
-                              ZSTD_entropyCTables_t* nextEntropy,
-                              ZSTD_CCtx_params const* cctxParams,
-                              void* dst, size_t dstCapacity,
-                              size_t srcSize, U32* workspace, int bmi2)
+MEM_STATIC size_t
+ZSTD_compressSequences(seqStore_t* seqStorePtr,
+                       const ZSTD_entropyCTables_t* prevEntropy,
+                             ZSTD_entropyCTables_t* nextEntropy,
+                       const ZSTD_CCtx_params* cctxParams,
+                             void* dst, size_t dstCapacity,
+                             size_t srcSize,
+                             void* workspace, size_t wkspSize,
+                             int bmi2)
 {
     size_t const cSize = ZSTD_compressSequences_internal(
-            seqStorePtr, prevEntropy, nextEntropy, cctxParams, dst, dstCapacity,
-            workspace, bmi2);
+                            seqStorePtr, prevEntropy, nextEntropy, cctxParams,
+                            dst, dstCapacity,
+                            workspace, wkspSize, bmi2);
+    if (cSize == 0) return 0;
     /* When srcSize <= dstCapacity, there is enough space to write a raw uncompressed block.
      * Since we ran out of space, block must be not compressible, so fall back to raw uncompressed block.
      */
     if ((cSize == ERROR(dstSize_tooSmall)) & (srcSize <= dstCapacity))
         return 0;  /* block not compressed */
-    if (ZSTD_isError(cSize)) return cSize;
+    FORWARD_IF_ERROR(cSize);
 
     /* Check compressibility */
-    {   size_t const maxCSize = srcSize - ZSTD_minGain(srcSize);  /* note : fixed formula, maybe should depend on compression level, or strategy */
+    {   size_t const maxCSize = srcSize - ZSTD_minGain(srcSize, cctxParams->cParams.strategy);
         if (cSize >= maxCSize) return 0;  /* block not compressed */
     }
 
-    /* We check that dictionaries have offset codes available for the first
-     * block. After the first block, the offcode table might not have large
-     * enough codes to represent the offsets in the data.
-     */
-    if (nextEntropy->offcode_repeatMode == FSE_repeat_valid)
-        nextEntropy->offcode_repeatMode = FSE_repeat_check;
-
     return cSize;
 }
 
 /* ZSTD_selectBlockCompressor() :
  * Not static, but internal use only (used by long distance matcher)
  * assumption : strat is a valid strategy */
-ZSTD_blockCompressor ZSTD_selectBlockCompressor(ZSTD_strategy strat, int extDict)
+ZSTD_blockCompressor ZSTD_selectBlockCompressor(ZSTD_strategy strat, ZSTD_dictMode_e dictMode)
 {
-    static const ZSTD_blockCompressor blockCompressor[2][(unsigned)ZSTD_btultra+1] = {
+    static const ZSTD_blockCompressor blockCompressor[3][ZSTD_STRATEGY_MAX+1] = {
         { ZSTD_compressBlock_fast  /* default for 0 */,
-          ZSTD_compressBlock_fast, ZSTD_compressBlock_doubleFast, ZSTD_compressBlock_greedy,
-          ZSTD_compressBlock_lazy, ZSTD_compressBlock_lazy2, ZSTD_compressBlock_btlazy2,
-          ZSTD_compressBlock_btopt, ZSTD_compressBlock_btultra },
+          ZSTD_compressBlock_fast,
+          ZSTD_compressBlock_doubleFast,
+          ZSTD_compressBlock_greedy,
+          ZSTD_compressBlock_lazy,
+          ZSTD_compressBlock_lazy2,
+          ZSTD_compressBlock_btlazy2,
+          ZSTD_compressBlock_btopt,
+          ZSTD_compressBlock_btultra,
+          ZSTD_compressBlock_btultra2 },
         { ZSTD_compressBlock_fast_extDict  /* default for 0 */,
-          ZSTD_compressBlock_fast_extDict, ZSTD_compressBlock_doubleFast_extDict, ZSTD_compressBlock_greedy_extDict,
-          ZSTD_compressBlock_lazy_extDict,ZSTD_compressBlock_lazy2_extDict, ZSTD_compressBlock_btlazy2_extDict,
-          ZSTD_compressBlock_btopt_extDict, ZSTD_compressBlock_btultra_extDict }
+          ZSTD_compressBlock_fast_extDict,
+          ZSTD_compressBlock_doubleFast_extDict,
+          ZSTD_compressBlock_greedy_extDict,
+          ZSTD_compressBlock_lazy_extDict,
+          ZSTD_compressBlock_lazy2_extDict,
+          ZSTD_compressBlock_btlazy2_extDict,
+          ZSTD_compressBlock_btopt_extDict,
+          ZSTD_compressBlock_btultra_extDict,
+          ZSTD_compressBlock_btultra_extDict },
+        { ZSTD_compressBlock_fast_dictMatchState  /* default for 0 */,
+          ZSTD_compressBlock_fast_dictMatchState,
+          ZSTD_compressBlock_doubleFast_dictMatchState,
+          ZSTD_compressBlock_greedy_dictMatchState,
+          ZSTD_compressBlock_lazy_dictMatchState,
+          ZSTD_compressBlock_lazy2_dictMatchState,
+          ZSTD_compressBlock_btlazy2_dictMatchState,
+          ZSTD_compressBlock_btopt_dictMatchState,
+          ZSTD_compressBlock_btultra_dictMatchState,
+          ZSTD_compressBlock_btultra_dictMatchState }
     };
+    ZSTD_blockCompressor selectedCompressor;
     ZSTD_STATIC_ASSERT((unsigned)ZSTD_fast == 1);
 
-    assert((U32)strat >= (U32)ZSTD_fast);
-    assert((U32)strat <= (U32)ZSTD_btultra);
-    return blockCompressor[extDict!=0][(U32)strat];
+    assert(ZSTD_cParam_withinBounds(ZSTD_c_strategy, strat));
+    selectedCompressor = blockCompressor[(int)dictMode][(int)strat];
+    assert(selectedCompressor != NULL);
+    return selectedCompressor;
 }
 
 static void ZSTD_storeLastLiterals(seqStore_t* seqStorePtr,
@@ -1880,36 +2754,47 @@
     seqStorePtr->lit += lastLLSize;
 }
 
-static void ZSTD_resetSeqStore(seqStore_t* ssPtr)
+void ZSTD_resetSeqStore(seqStore_t* ssPtr)
 {
     ssPtr->lit = ssPtr->litStart;
     ssPtr->sequences = ssPtr->sequencesStart;
     ssPtr->longLengthID = 0;
 }
 
-static size_t ZSTD_compressBlock_internal(ZSTD_CCtx* zc,
-                                        void* dst, size_t dstCapacity,
-                                        const void* src, size_t srcSize)
+typedef enum { ZSTDbss_compress, ZSTDbss_noCompress } ZSTD_buildSeqStore_e;
+
+static size_t ZSTD_buildSeqStore(ZSTD_CCtx* zc, const void* src, size_t srcSize)
 {
     ZSTD_matchState_t* const ms = &zc->blockState.matchState;
-    DEBUGLOG(5, "ZSTD_compressBlock_internal (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u)",
-                (U32)dstCapacity, ms->window.dictLimit, ms->nextToUpdate);
+    DEBUGLOG(5, "ZSTD_buildSeqStore (srcSize=%zu)", srcSize);
+    assert(srcSize <= ZSTD_BLOCKSIZE_MAX);
+    /* Assert that we have correctly flushed the ctx params into the ms's copy */
+    ZSTD_assertEqualCParams(zc->appliedParams.cParams, ms->cParams);
     if (srcSize < MIN_CBLOCK_SIZE+ZSTD_blockHeaderSize+1) {
-        ZSTD_ldm_skipSequences(&zc->externSeqStore, srcSize, zc->appliedParams.cParams.searchLength);
-        return 0;   /* don't even attempt compression below a certain srcSize */
+        ZSTD_ldm_skipSequences(&zc->externSeqStore, srcSize, zc->appliedParams.cParams.minMatch);
+        return ZSTDbss_noCompress; /* don't even attempt compression below a certain srcSize */
     }
     ZSTD_resetSeqStore(&(zc->seqStore));
+    /* required for optimal parser to read stats from dictionary */
+    ms->opt.symbolCosts = &zc->blockState.prevCBlock->entropy;
+    /* tell the optimal parser how we expect to compress literals */
+    ms->opt.literalCompressionMode = zc->appliedParams.literalCompressionMode;
+    /* a gap between an attached dict and the current window is not safe,
+     * they must remain adjacent,
+     * and when that stops being the case, the dict must be unset */
+    assert(ms->dictMatchState == NULL || ms->loadedDictEnd == ms->window.dictLimit);
 
     /* limited update after a very long match */
     {   const BYTE* const base = ms->window.base;
         const BYTE* const istart = (const BYTE*)src;
         const U32 current = (U32)(istart-base);
+        if (sizeof(ptrdiff_t)==8) assert(istart - base < (ptrdiff_t)(U32)(-1));   /* ensure no overflow */
         if (current > ms->nextToUpdate + 384)
             ms->nextToUpdate = current - MIN(192, (U32)(current - ms->nextToUpdate - 384));
     }
 
     /* select and store sequences */
-    {   U32 const extDict = ZSTD_window_hasExtDict(ms->window);
+    {   ZSTD_dictMode_e const dictMode = ZSTD_matchState_dictMode(ms);
         size_t lastLLSize;
         {   int i;
             for (i = 0; i < ZSTD_REP_NUM; ++i)
@@ -1922,8 +2807,7 @@
                 ZSTD_ldm_blockCompress(&zc->externSeqStore,
                                        ms, &zc->seqStore,
                                        zc->blockState.nextCBlock->rep,
-                                       &zc->appliedParams.cParams,
-                                       src, srcSize, extDict);
+                                       src, srcSize);
             assert(zc->externSeqStore.pos <= zc->externSeqStore.size);
         } else if (zc->appliedParams.ldmParams.enableLdm) {
             rawSeqStore_t ldmSeqStore = {NULL, 0, 0, 0};
@@ -1931,7 +2815,7 @@
             ldmSeqStore.seq = zc->ldmSequences;
             ldmSeqStore.capacity = zc->maxNbLdmSequences;
             /* Updates ldmSeqStore.size */
-            CHECK_F(ZSTD_ldm_generateSequences(&zc->ldmState, &ldmSeqStore,
+            FORWARD_IF_ERROR(ZSTD_ldm_generateSequences(&zc->ldmState, &ldmSeqStore,
                                                &zc->appliedParams.ldmParams,
                                                src, srcSize));
             /* Updates ldmSeqStore.pos */
@@ -1939,30 +2823,73 @@
                 ZSTD_ldm_blockCompress(&ldmSeqStore,
                                        ms, &zc->seqStore,
                                        zc->blockState.nextCBlock->rep,
-                                       &zc->appliedParams.cParams,
-                                       src, srcSize, extDict);
+                                       src, srcSize);
             assert(ldmSeqStore.pos == ldmSeqStore.size);
         } else {   /* not long range mode */
-            ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor(zc->appliedParams.cParams.strategy, extDict);
-            lastLLSize = blockCompressor(ms, &zc->seqStore, zc->blockState.nextCBlock->rep, &zc->appliedParams.cParams, src, srcSize);
+            ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor(zc->appliedParams.cParams.strategy, dictMode);
+            lastLLSize = blockCompressor(ms, &zc->seqStore, zc->blockState.nextCBlock->rep, src, srcSize);
         }
         {   const BYTE* const lastLiterals = (const BYTE*)src + srcSize - lastLLSize;
             ZSTD_storeLastLiterals(&zc->seqStore, lastLiterals, lastLLSize);
     }   }
+    return ZSTDbss_compress;
+}
+
+static size_t ZSTD_compressBlock_internal(ZSTD_CCtx* zc,
+                                        void* dst, size_t dstCapacity,
+                                        const void* src, size_t srcSize)
+{
+    size_t cSize;
+    DEBUGLOG(5, "ZSTD_compressBlock_internal (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u)",
+                (unsigned)dstCapacity, (unsigned)zc->blockState.matchState.window.dictLimit, (unsigned)zc->blockState.matchState.nextToUpdate);
+
+    {   const size_t bss = ZSTD_buildSeqStore(zc, src, srcSize);
+        FORWARD_IF_ERROR(bss);
+        if (bss == ZSTDbss_noCompress) { cSize = 0; goto out; }
+    }
 
     /* encode sequences and literals */
-    {   size_t const cSize = ZSTD_compressSequences(&zc->seqStore,
-                                &zc->blockState.prevCBlock->entropy, &zc->blockState.nextCBlock->entropy,
-                                &zc->appliedParams,
-                                dst, dstCapacity,
-                                srcSize, zc->entropyWorkspace, zc->bmi2);
-        if (ZSTD_isError(cSize) || cSize == 0) return cSize;
-        /* confirm repcodes and entropy tables */
-        {   ZSTD_compressedBlockState_t* const tmp = zc->blockState.prevCBlock;
-            zc->blockState.prevCBlock = zc->blockState.nextCBlock;
-            zc->blockState.nextCBlock = tmp;
-        }
-        return cSize;
+    cSize = ZSTD_compressSequences(&zc->seqStore,
+            &zc->blockState.prevCBlock->entropy, &zc->blockState.nextCBlock->entropy,
+            &zc->appliedParams,
+            dst, dstCapacity,
+            srcSize,
+            zc->entropyWorkspace, HUF_WORKSPACE_SIZE /* statically allocated in resetCCtx */,
+            zc->bmi2);
+
+out:
+    if (!ZSTD_isError(cSize) && cSize != 0) {
+        /* confirm repcodes and entropy tables when emitting a compressed block */
+        ZSTD_compressedBlockState_t* const tmp = zc->blockState.prevCBlock;
+        zc->blockState.prevCBlock = zc->blockState.nextCBlock;
+        zc->blockState.nextCBlock = tmp;
+    }
+    /* We check that dictionaries have offset codes available for the first
+     * block. After the first block, the offcode table might not have large
+     * enough codes to represent the offsets in the data.
+     */
+    if (zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid)
+        zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check;
+
+    return cSize;
+}
+
+
+static void ZSTD_overflowCorrectIfNeeded(ZSTD_matchState_t* ms, ZSTD_CCtx_params const* params, void const* ip, void const* iend)
+{
+    if (ZSTD_window_needOverflowCorrection(ms->window, iend)) {
+        U32 const maxDist = (U32)1 << params->cParams.windowLog;
+        U32 const cycleLog = ZSTD_cycleLog(params->cParams.chainLog, params->cParams.strategy);
+        U32 const correction = ZSTD_window_correctOverflow(&ms->window, cycleLog, maxDist, ip);
+        ZSTD_STATIC_ASSERT(ZSTD_CHAINLOG_MAX <= 30);
+        ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX_32 <= 30);
+        ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX <= 31);
+        ZSTD_reduceIndex(ms, params, correction);
+        if (ms->nextToUpdate < correction) ms->nextToUpdate = 0;
+        else ms->nextToUpdate -= correction;
+        /* invalidate dictionaries on overflow correction */
+        ms->loadedDictEnd = 0;
+        ms->dictMatchState = NULL;
     }
 }
 
@@ -1985,9 +2912,9 @@
     BYTE* const ostart = (BYTE*)dst;
     BYTE* op = ostart;
     U32 const maxDist = (U32)1 << cctx->appliedParams.cParams.windowLog;
-    assert(cctx->appliedParams.cParams.windowLog <= 31);
+    assert(cctx->appliedParams.cParams.windowLog <= ZSTD_WINDOWLOG_MAX);
 
-    DEBUGLOG(5, "ZSTD_compress_frameChunk (blockSize=%u)", (U32)blockSize);
+    DEBUGLOG(5, "ZSTD_compress_frameChunk (blockSize=%u)", (unsigned)blockSize);
     if (cctx->appliedParams.fParams.checksumFlag && srcSize)
         XXH64_update(&cctx->xxhState, src, srcSize);
 
@@ -1995,36 +2922,25 @@
         ZSTD_matchState_t* const ms = &cctx->blockState.matchState;
         U32 const lastBlock = lastFrameChunk & (blockSize >= remaining);
 
-        if (dstCapacity < ZSTD_blockHeaderSize + MIN_CBLOCK_SIZE)
-            return ERROR(dstSize_tooSmall);   /* not enough space to store compressed block */
+        RETURN_ERROR_IF(dstCapacity < ZSTD_blockHeaderSize + MIN_CBLOCK_SIZE,
+                        dstSize_tooSmall,
+                        "not enough space to store compressed block");
         if (remaining < blockSize) blockSize = remaining;
 
-        if (ZSTD_window_needOverflowCorrection(ms->window, ip + blockSize)) {
-            U32 const cycleLog = ZSTD_cycleLog(cctx->appliedParams.cParams.chainLog, cctx->appliedParams.cParams.strategy);
-            U32 const correction = ZSTD_window_correctOverflow(&ms->window, cycleLog, maxDist, ip);
-            ZSTD_STATIC_ASSERT(ZSTD_CHAINLOG_MAX <= 30);
-            ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX_32 <= 30);
-            ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX <= 31);
+        ZSTD_overflowCorrectIfNeeded(ms, &cctx->appliedParams, ip, ip + blockSize);
+        ZSTD_checkDictValidity(&ms->window, ip + blockSize, maxDist, &ms->loadedDictEnd, &ms->dictMatchState);
 
-            ZSTD_reduceIndex(cctx, correction);
-            if (ms->nextToUpdate < correction) ms->nextToUpdate = 0;
-            else ms->nextToUpdate -= correction;
-            ms->loadedDictEnd = 0;
-        }
-        ZSTD_window_enforceMaxDist(&ms->window, ip + blockSize, maxDist, &ms->loadedDictEnd);
+        /* Ensure hash/chain table insertion resumes no sooner than lowlimit */
         if (ms->nextToUpdate < ms->window.lowLimit) ms->nextToUpdate = ms->window.lowLimit;
 
         {   size_t cSize = ZSTD_compressBlock_internal(cctx,
                                 op+ZSTD_blockHeaderSize, dstCapacity-ZSTD_blockHeaderSize,
                                 ip, blockSize);
-            if (ZSTD_isError(cSize)) return cSize;
+            FORWARD_IF_ERROR(cSize);
 
             if (cSize == 0) {  /* block is not compressible */
-                U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw)<<1) + (U32)(blockSize << 3);
-                if (blockSize + ZSTD_blockHeaderSize > dstCapacity) return ERROR(dstSize_tooSmall);
-                MEM_writeLE32(op, cBlockHeader24);   /* 4th byte will be overwritten */
-                memcpy(op + ZSTD_blockHeaderSize, ip, blockSize);
-                cSize = ZSTD_blockHeaderSize + blockSize;
+                cSize = ZSTD_noCompressBlock(op, dstCapacity, ip, blockSize, lastBlock);
+                FORWARD_IF_ERROR(cSize);
             } else {
                 U32 const cBlockHeader24 = lastBlock + (((U32)bt_compressed)<<1) + (U32)(cSize << 3);
                 MEM_writeLE24(op, cBlockHeader24);
@@ -2038,11 +2954,11 @@
             assert(dstCapacity >= cSize);
             dstCapacity -= cSize;
             DEBUGLOG(5, "ZSTD_compress_frameChunk: adding a block of size %u",
-                        (U32)cSize);
+                        (unsigned)cSize);
     }   }
 
     if (lastFrameChunk && (op>ostart)) cctx->stage = ZSTDcs_ending;
-    return op-ostart;
+    return (size_t)(op-ostart);
 }
 
 
@@ -2057,18 +2973,19 @@
     BYTE  const windowLogByte = (BYTE)((params.cParams.windowLog - ZSTD_WINDOWLOG_ABSOLUTEMIN) << 3);
     U32   const fcsCode = params.fParams.contentSizeFlag ?
                      (pledgedSrcSize>=256) + (pledgedSrcSize>=65536+256) + (pledgedSrcSize>=0xFFFFFFFFU) : 0;  /* 0-3 */
-    BYTE  const frameHeaderDecriptionByte = (BYTE)(dictIDSizeCode + (checksumFlag<<2) + (singleSegment<<5) + (fcsCode<<6) );
+    BYTE  const frameHeaderDescriptionByte = (BYTE)(dictIDSizeCode + (checksumFlag<<2) + (singleSegment<<5) + (fcsCode<<6) );
     size_t pos=0;
 
-    if (dstCapacity < ZSTD_frameHeaderSize_max) return ERROR(dstSize_tooSmall);
+    assert(!(params.fParams.contentSizeFlag && pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN));
+    RETURN_ERROR_IF(dstCapacity < ZSTD_FRAMEHEADERSIZE_MAX, dstSize_tooSmall);
     DEBUGLOG(4, "ZSTD_writeFrameHeader : dictIDFlag : %u ; dictID : %u ; dictIDSizeCode : %u",
-                !params.fParams.noDictIDFlag, dictID,  dictIDSizeCode);
+                !params.fParams.noDictIDFlag, (unsigned)dictID, (unsigned)dictIDSizeCode);
 
     if (params.format == ZSTD_f_zstd1) {
         MEM_writeLE32(dst, ZSTD_MAGICNUMBER);
         pos = 4;
     }
-    op[pos++] = frameHeaderDecriptionByte;
+    op[pos++] = frameHeaderDescriptionByte;
     if (!singleSegment) op[pos++] = windowLogByte;
     switch(dictIDSizeCode)
     {
@@ -2092,11 +3009,11 @@
 /* ZSTD_writeLastEmptyBlock() :
  * output an empty Block with end-of-frame mark to complete a frame
  * @return : size of data written into `dst` (== ZSTD_blockHeaderSize (defined in zstd_internal.h))
- *           or an error code if `dstCapcity` is too small (<ZSTD_blockHeaderSize)
+ *           or an error code if `dstCapacity` is too small (<ZSTD_blockHeaderSize)
  */
 size_t ZSTD_writeLastEmptyBlock(void* dst, size_t dstCapacity)
 {
-    if (dstCapacity < ZSTD_blockHeaderSize) return ERROR(dstSize_tooSmall);
+    RETURN_ERROR_IF(dstCapacity < ZSTD_blockHeaderSize, dstSize_tooSmall);
     {   U32 const cBlockHeader24 = 1 /*lastBlock*/ + (((U32)bt_raw)<<1);  /* 0 size */
         MEM_writeLE24(dst, cBlockHeader24);
         return ZSTD_blockHeaderSize;
@@ -2105,10 +3022,9 @@
 
 size_t ZSTD_referenceExternalSequences(ZSTD_CCtx* cctx, rawSeq* seq, size_t nbSeq)
 {
-    if (cctx->stage != ZSTDcs_init)
-        return ERROR(stage_wrong);
-    if (cctx->appliedParams.ldmParams.enableLdm)
-        return ERROR(parameter_unsupported);
+    RETURN_ERROR_IF(cctx->stage != ZSTDcs_init, stage_wrong);
+    RETURN_ERROR_IF(cctx->appliedParams.ldmParams.enableLdm,
+                    parameter_unsupported);
     cctx->externSeqStore.seq = seq;
     cctx->externSeqStore.size = nbSeq;
     cctx->externSeqStore.capacity = nbSeq;
@@ -2122,17 +3038,19 @@
                         const void* src, size_t srcSize,
                                U32 frame, U32 lastFrameChunk)
 {
-    ZSTD_matchState_t* ms = &cctx->blockState.matchState;
+    ZSTD_matchState_t* const ms = &cctx->blockState.matchState;
     size_t fhSize = 0;
 
     DEBUGLOG(5, "ZSTD_compressContinue_internal, stage: %u, srcSize: %u",
-                cctx->stage, (U32)srcSize);
-    if (cctx->stage==ZSTDcs_created) return ERROR(stage_wrong);   /* missing init (ZSTD_compressBegin) */
+                cctx->stage, (unsigned)srcSize);
+    RETURN_ERROR_IF(cctx->stage==ZSTDcs_created, stage_wrong,
+                    "missing init (ZSTD_compressBegin)");
 
     if (frame && (cctx->stage==ZSTDcs_init)) {
         fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, cctx->appliedParams,
                                        cctx->pledgedSrcSizePlusOne-1, cctx->dictID);
-        if (ZSTD_isError(fhSize)) return fhSize;
+        FORWARD_IF_ERROR(fhSize);
+        assert(fhSize <= dstCapacity);
         dstCapacity -= fhSize;
         dst = (char*)dst + fhSize;
         cctx->stage = ZSTDcs_ongoing;
@@ -2143,22 +3061,31 @@
     if (!ZSTD_window_update(&ms->window, src, srcSize)) {
         ms->nextToUpdate = ms->window.dictLimit;
     }
-    if (cctx->appliedParams.ldmParams.enableLdm)
+    if (cctx->appliedParams.ldmParams.enableLdm) {
         ZSTD_window_update(&cctx->ldmState.window, src, srcSize);
+    }
 
-    DEBUGLOG(5, "ZSTD_compressContinue_internal (blockSize=%u)", (U32)cctx->blockSize);
+    if (!frame) {
+        /* overflow check and correction for block mode */
+        ZSTD_overflowCorrectIfNeeded(ms, &cctx->appliedParams, src, (BYTE const*)src + srcSize);
+    }
+
+    DEBUGLOG(5, "ZSTD_compressContinue_internal (blockSize=%u)", (unsigned)cctx->blockSize);
     {   size_t const cSize = frame ?
                              ZSTD_compress_frameChunk (cctx, dst, dstCapacity, src, srcSize, lastFrameChunk) :
                              ZSTD_compressBlock_internal (cctx, dst, dstCapacity, src, srcSize);
-        if (ZSTD_isError(cSize)) return cSize;
+        FORWARD_IF_ERROR(cSize);
         cctx->consumedSrcSize += srcSize;
         cctx->producedCSize += (cSize + fhSize);
-        if (cctx->appliedParams.fParams.contentSizeFlag) {  /* control src size */
-            if (cctx->consumedSrcSize+1 > cctx->pledgedSrcSizePlusOne) {
-                DEBUGLOG(4, "error : pledgedSrcSize = %u, while realSrcSize >= %u",
-                    (U32)cctx->pledgedSrcSizePlusOne-1, (U32)cctx->consumedSrcSize);
-                return ERROR(srcSize_wrong);
-            }
+        assert(!(cctx->appliedParams.fParams.contentSizeFlag && cctx->pledgedSrcSizePlusOne == 0));
+        if (cctx->pledgedSrcSizePlusOne != 0) {  /* control src size */
+            ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN == (unsigned long long)-1);
+            RETURN_ERROR_IF(
+                cctx->consumedSrcSize+1 > cctx->pledgedSrcSizePlusOne,
+                srcSize_wrong,
+                "error : pledgedSrcSize = %u, while realSrcSize >= %u",
+                (unsigned)cctx->pledgedSrcSizePlusOne-1,
+                (unsigned)cctx->consumedSrcSize);
         }
         return cSize + fhSize;
     }
@@ -2168,7 +3095,7 @@
                               void* dst, size_t dstCapacity,
                         const void* src, size_t srcSize)
 {
-    DEBUGLOG(5, "ZSTD_compressContinue (srcSize=%u)", (U32)srcSize);
+    DEBUGLOG(5, "ZSTD_compressContinue (srcSize=%u)", (unsigned)srcSize);
     return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 1 /* frame mode */, 0 /* last chunk */);
 }
 
@@ -2183,49 +3110,66 @@
 size_t ZSTD_compressBlock(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize)
 {
     size_t const blockSizeMax = ZSTD_getBlockSize(cctx);
-    if (srcSize > blockSizeMax) return ERROR(srcSize_wrong);
+    RETURN_ERROR_IF(srcSize > blockSizeMax, srcSize_wrong);
+
     return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 0 /* frame mode */, 0 /* last chunk */);
 }
 
 /*! ZSTD_loadDictionaryContent() :
  *  @return : 0, or an error code
  */
-static size_t ZSTD_loadDictionaryContent(ZSTD_matchState_t* ms, ZSTD_CCtx_params const* params, const void* src, size_t srcSize)
+static size_t ZSTD_loadDictionaryContent(ZSTD_matchState_t* ms,
+                                         ZSTD_CCtx_params const* params,
+                                         const void* src, size_t srcSize,
+                                         ZSTD_dictTableLoadMethod_e dtlm)
 {
-    const BYTE* const ip = (const BYTE*) src;
+    const BYTE* ip = (const BYTE*) src;
     const BYTE* const iend = ip + srcSize;
-    ZSTD_compressionParameters const* cParams = &params->cParams;
 
     ZSTD_window_update(&ms->window, src, srcSize);
     ms->loadedDictEnd = params->forceWindow ? 0 : (U32)(iend - ms->window.base);
 
+    /* Assert that we the ms params match the params we're being given */
+    ZSTD_assertEqualCParams(params->cParams, ms->cParams);
+
     if (srcSize <= HASH_READ_SIZE) return 0;
 
-    switch(params->cParams.strategy)
-    {
-    case ZSTD_fast:
-        ZSTD_fillHashTable(ms, cParams, iend);
-        break;
-    case ZSTD_dfast:
-        ZSTD_fillDoubleHashTable(ms, cParams, iend);
-        break;
+    while (iend - ip > HASH_READ_SIZE) {
+        size_t const remaining = iend - ip;
+        size_t const chunk = MIN(remaining, ZSTD_CHUNKSIZE_MAX);
+        const BYTE* const ichunk = ip + chunk;
 
-    case ZSTD_greedy:
-    case ZSTD_lazy:
-    case ZSTD_lazy2:
-        if (srcSize >= HASH_READ_SIZE)
-            ZSTD_insertAndFindFirstIndex(ms, cParams, iend-HASH_READ_SIZE);
-        break;
+        ZSTD_overflowCorrectIfNeeded(ms, params, ip, ichunk);
 
-    case ZSTD_btlazy2:   /* we want the dictionary table fully sorted */
-    case ZSTD_btopt:
-    case ZSTD_btultra:
-        if (srcSize >= HASH_READ_SIZE)
-            ZSTD_updateTree(ms, cParams, iend-HASH_READ_SIZE, iend);
-        break;
+        switch(params->cParams.strategy)
+        {
+        case ZSTD_fast:
+            ZSTD_fillHashTable(ms, ichunk, dtlm);
+            break;
+        case ZSTD_dfast:
+            ZSTD_fillDoubleHashTable(ms, ichunk, dtlm);
+            break;
 
-    default:
-        assert(0);  /* not possible : not a valid strategy id */
+        case ZSTD_greedy:
+        case ZSTD_lazy:
+        case ZSTD_lazy2:
+            if (chunk >= HASH_READ_SIZE)
+                ZSTD_insertAndFindFirstIndex(ms, ichunk-HASH_READ_SIZE);
+            break;
+
+        case ZSTD_btlazy2:   /* we want the dictionary table fully sorted */
+        case ZSTD_btopt:
+        case ZSTD_btultra:
+        case ZSTD_btultra2:
+            if (chunk >= HASH_READ_SIZE)
+                ZSTD_updateTree(ms, ichunk-HASH_READ_SIZE, ichunk);
+            break;
+
+        default:
+            assert(0);  /* not possible : not a valid strategy id */
+        }
+
+        ip = ichunk;
     }
 
     ms->nextToUpdate = (U32)(iend - ms->window.base);
@@ -2239,9 +3183,9 @@
    NOTE: This behavior is not standard and could be improved in the future. */
 static size_t ZSTD_checkDictNCount(short* normalizedCounter, unsigned dictMaxSymbolValue, unsigned maxSymbolValue) {
     U32 s;
-    if (dictMaxSymbolValue < maxSymbolValue) return ERROR(dictionary_corrupted);
+    RETURN_ERROR_IF(dictMaxSymbolValue < maxSymbolValue, dictionary_corrupted);
     for (s = 0; s <= maxSymbolValue; ++s) {
-        if (normalizedCounter[s] == 0) return ERROR(dictionary_corrupted);
+        RETURN_ERROR_IF(normalizedCounter[s] == 0, dictionary_corrupted);
     }
     return 0;
 }
@@ -2256,7 +3200,12 @@
  *  assumptions : magic number supposed already checked
  *                dictSize supposed > 8
  */
-static size_t ZSTD_loadZstdDictionary(ZSTD_compressedBlockState_t* bs, ZSTD_matchState_t* ms, ZSTD_CCtx_params const* params, const void* dict, size_t dictSize, void* workspace)
+static size_t ZSTD_loadZstdDictionary(ZSTD_compressedBlockState_t* bs,
+                                      ZSTD_matchState_t* ms,
+                                      ZSTD_CCtx_params const* params,
+                                      const void* dict, size_t dictSize,
+                                      ZSTD_dictTableLoadMethod_e dtlm,
+                                      void* workspace)
 {
     const BYTE* dictPtr = (const BYTE*)dict;
     const BYTE* const dictEnd = dictPtr + dictSize;
@@ -2265,53 +3214,65 @@
     size_t dictID;
 
     ZSTD_STATIC_ASSERT(HUF_WORKSPACE_SIZE >= (1<<MAX(MLFSELog,LLFSELog)));
+    assert(dictSize > 8);
+    assert(MEM_readLE32(dictPtr) == ZSTD_MAGIC_DICTIONARY);
 
     dictPtr += 4;   /* skip magic number */
     dictID = params->fParams.noDictIDFlag ? 0 :  MEM_readLE32(dictPtr);
     dictPtr += 4;
 
     {   unsigned maxSymbolValue = 255;
-        size_t const hufHeaderSize = HUF_readCTable((HUF_CElt*)bs->entropy.hufCTable, &maxSymbolValue, dictPtr, dictEnd-dictPtr);
-        if (HUF_isError(hufHeaderSize)) return ERROR(dictionary_corrupted);
-        if (maxSymbolValue < 255) return ERROR(dictionary_corrupted);
+        size_t const hufHeaderSize = HUF_readCTable((HUF_CElt*)bs->entropy.huf.CTable, &maxSymbolValue, dictPtr, dictEnd-dictPtr);
+        RETURN_ERROR_IF(HUF_isError(hufHeaderSize), dictionary_corrupted);
+        RETURN_ERROR_IF(maxSymbolValue < 255, dictionary_corrupted);
         dictPtr += hufHeaderSize;
     }
 
     {   unsigned offcodeLog;
         size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd-dictPtr);
-        if (FSE_isError(offcodeHeaderSize)) return ERROR(dictionary_corrupted);
-        if (offcodeLog > OffFSELog) return ERROR(dictionary_corrupted);
+        RETURN_ERROR_IF(FSE_isError(offcodeHeaderSize), dictionary_corrupted);
+        RETURN_ERROR_IF(offcodeLog > OffFSELog, dictionary_corrupted);
         /* Defer checking offcodeMaxValue because we need to know the size of the dictionary content */
-        CHECK_E( FSE_buildCTable_wksp(bs->entropy.offcodeCTable, offcodeNCount, offcodeMaxValue, offcodeLog, workspace, HUF_WORKSPACE_SIZE),
-                 dictionary_corrupted);
+        /* fill all offset symbols to avoid garbage at end of table */
+        RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp(
+                bs->entropy.fse.offcodeCTable,
+                offcodeNCount, MaxOff, offcodeLog,
+                workspace, HUF_WORKSPACE_SIZE)),
+            dictionary_corrupted);
         dictPtr += offcodeHeaderSize;
     }
 
     {   short matchlengthNCount[MaxML+1];
         unsigned matchlengthMaxValue = MaxML, matchlengthLog;
         size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd-dictPtr);
-        if (FSE_isError(matchlengthHeaderSize)) return ERROR(dictionary_corrupted);
-        if (matchlengthLog > MLFSELog) return ERROR(dictionary_corrupted);
+        RETURN_ERROR_IF(FSE_isError(matchlengthHeaderSize), dictionary_corrupted);
+        RETURN_ERROR_IF(matchlengthLog > MLFSELog, dictionary_corrupted);
         /* Every match length code must have non-zero probability */
-        CHECK_F( ZSTD_checkDictNCount(matchlengthNCount, matchlengthMaxValue, MaxML));
-        CHECK_E( FSE_buildCTable_wksp(bs->entropy.matchlengthCTable, matchlengthNCount, matchlengthMaxValue, matchlengthLog, workspace, HUF_WORKSPACE_SIZE),
-                 dictionary_corrupted);
+        FORWARD_IF_ERROR( ZSTD_checkDictNCount(matchlengthNCount, matchlengthMaxValue, MaxML));
+        RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp(
+                bs->entropy.fse.matchlengthCTable,
+                matchlengthNCount, matchlengthMaxValue, matchlengthLog,
+                workspace, HUF_WORKSPACE_SIZE)),
+            dictionary_corrupted);
         dictPtr += matchlengthHeaderSize;
     }
 
     {   short litlengthNCount[MaxLL+1];
         unsigned litlengthMaxValue = MaxLL, litlengthLog;
         size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd-dictPtr);
-        if (FSE_isError(litlengthHeaderSize)) return ERROR(dictionary_corrupted);
-        if (litlengthLog > LLFSELog) return ERROR(dictionary_corrupted);
+        RETURN_ERROR_IF(FSE_isError(litlengthHeaderSize), dictionary_corrupted);
+        RETURN_ERROR_IF(litlengthLog > LLFSELog, dictionary_corrupted);
         /* Every literal length code must have non-zero probability */
-        CHECK_F( ZSTD_checkDictNCount(litlengthNCount, litlengthMaxValue, MaxLL));
-        CHECK_E( FSE_buildCTable_wksp(bs->entropy.litlengthCTable, litlengthNCount, litlengthMaxValue, litlengthLog, workspace, HUF_WORKSPACE_SIZE),
-                 dictionary_corrupted);
+        FORWARD_IF_ERROR( ZSTD_checkDictNCount(litlengthNCount, litlengthMaxValue, MaxLL));
+        RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp(
+                bs->entropy.fse.litlengthCTable,
+                litlengthNCount, litlengthMaxValue, litlengthLog,
+                workspace, HUF_WORKSPACE_SIZE)),
+            dictionary_corrupted);
         dictPtr += litlengthHeaderSize;
     }
 
-    if (dictPtr+12 > dictEnd) return ERROR(dictionary_corrupted);
+    RETURN_ERROR_IF(dictPtr+12 > dictEnd, dictionary_corrupted);
     bs->rep[0] = MEM_readLE32(dictPtr+0);
     bs->rep[1] = MEM_readLE32(dictPtr+4);
     bs->rep[2] = MEM_readLE32(dictPtr+8);
@@ -2324,30 +3285,33 @@
             offcodeMax = ZSTD_highbit32(maxOffset); /* Calculate minimum offset code required to represent maxOffset */
         }
         /* All offset values <= dictContentSize + 128 KB must be representable */
-        CHECK_F (ZSTD_checkDictNCount(offcodeNCount, offcodeMaxValue, MIN(offcodeMax, MaxOff)));
+        FORWARD_IF_ERROR(ZSTD_checkDictNCount(offcodeNCount, offcodeMaxValue, MIN(offcodeMax, MaxOff)));
         /* All repCodes must be <= dictContentSize and != 0*/
         {   U32 u;
             for (u=0; u<3; u++) {
-                if (bs->rep[u] == 0) return ERROR(dictionary_corrupted);
-                if (bs->rep[u] > dictContentSize) return ERROR(dictionary_corrupted);
+                RETURN_ERROR_IF(bs->rep[u] == 0, dictionary_corrupted);
+                RETURN_ERROR_IF(bs->rep[u] > dictContentSize, dictionary_corrupted);
         }   }
 
-        bs->entropy.hufCTable_repeatMode = HUF_repeat_valid;
-        bs->entropy.offcode_repeatMode = FSE_repeat_valid;
-        bs->entropy.matchlength_repeatMode = FSE_repeat_valid;
-        bs->entropy.litlength_repeatMode = FSE_repeat_valid;
-        CHECK_F(ZSTD_loadDictionaryContent(ms, params, dictPtr, dictContentSize));
+        bs->entropy.huf.repeatMode = HUF_repeat_valid;
+        bs->entropy.fse.offcode_repeatMode = FSE_repeat_valid;
+        bs->entropy.fse.matchlength_repeatMode = FSE_repeat_valid;
+        bs->entropy.fse.litlength_repeatMode = FSE_repeat_valid;
+        FORWARD_IF_ERROR(ZSTD_loadDictionaryContent(ms, params, dictPtr, dictContentSize, dtlm));
         return dictID;
     }
 }
 
 /** ZSTD_compress_insertDictionary() :
 *   @return : dictID, or an error code */
-static size_t ZSTD_compress_insertDictionary(ZSTD_compressedBlockState_t* bs, ZSTD_matchState_t* ms,
-                                             ZSTD_CCtx_params const* params,
-                                       const void* dict, size_t dictSize,
-                                             ZSTD_dictContentType_e dictContentType,
-                                             void* workspace)
+static size_t
+ZSTD_compress_insertDictionary(ZSTD_compressedBlockState_t* bs,
+                               ZSTD_matchState_t* ms,
+                         const ZSTD_CCtx_params* params,
+                         const void* dict, size_t dictSize,
+                               ZSTD_dictContentType_e dictContentType,
+                               ZSTD_dictTableLoadMethod_e dtlm,
+                               void* workspace)
 {
     DEBUGLOG(4, "ZSTD_compress_insertDictionary (dictSize=%u)", (U32)dictSize);
     if ((dict==NULL) || (dictSize<=8)) return 0;
@@ -2356,30 +3320,30 @@
 
     /* dict restricted modes */
     if (dictContentType == ZSTD_dct_rawContent)
-        return ZSTD_loadDictionaryContent(ms, params, dict, dictSize);
+        return ZSTD_loadDictionaryContent(ms, params, dict, dictSize, dtlm);
 
     if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) {
         if (dictContentType == ZSTD_dct_auto) {
             DEBUGLOG(4, "raw content dictionary detected");
-            return ZSTD_loadDictionaryContent(ms, params, dict, dictSize);
+            return ZSTD_loadDictionaryContent(ms, params, dict, dictSize, dtlm);
         }
-        if (dictContentType == ZSTD_dct_fullDict)
-            return ERROR(dictionary_wrong);
+        RETURN_ERROR_IF(dictContentType == ZSTD_dct_fullDict, dictionary_wrong);
         assert(0);   /* impossible */
     }
 
     /* dict as full zstd dictionary */
-    return ZSTD_loadZstdDictionary(bs, ms, params, dict, dictSize, workspace);
+    return ZSTD_loadZstdDictionary(bs, ms, params, dict, dictSize, dtlm, workspace);
 }
 
 /*! ZSTD_compressBegin_internal() :
  * @return : 0, or an error code */
-size_t ZSTD_compressBegin_internal(ZSTD_CCtx* cctx,
-                             const void* dict, size_t dictSize,
-                             ZSTD_dictContentType_e dictContentType,
-                             const ZSTD_CDict* cdict,
-                             ZSTD_CCtx_params params, U64 pledgedSrcSize,
-                             ZSTD_buffered_policy_e zbuff)
+static size_t ZSTD_compressBegin_internal(ZSTD_CCtx* cctx,
+                                    const void* dict, size_t dictSize,
+                                    ZSTD_dictContentType_e dictContentType,
+                                    ZSTD_dictTableLoadMethod_e dtlm,
+                                    const ZSTD_CDict* cdict,
+                                    ZSTD_CCtx_params params, U64 pledgedSrcSize,
+                                    ZSTD_buffered_policy_e zbuff)
 {
     DEBUGLOG(4, "ZSTD_compressBegin_internal: wlog=%u", params.cParams.windowLog);
     /* params are supposed to be fully validated at this point */
@@ -2387,19 +3351,16 @@
     assert(!((dict) && (cdict)));  /* either dict or cdict, not both */
 
     if (cdict && cdict->dictContentSize>0) {
-        cctx->requestedParams = params;
-        return ZSTD_resetCCtx_usingCDict(cctx, cdict, params.cParams.windowLog,
-                                         params.fParams, pledgedSrcSize, zbuff);
+        return ZSTD_resetCCtx_usingCDict(cctx, cdict, params, pledgedSrcSize, zbuff);
     }
 
-    CHECK_F( ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize,
+    FORWARD_IF_ERROR( ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize,
                                      ZSTDcrp_continue, zbuff) );
-    {
-        size_t const dictID = ZSTD_compress_insertDictionary(
+    {   size_t const dictID = ZSTD_compress_insertDictionary(
                 cctx->blockState.prevCBlock, &cctx->blockState.matchState,
-                &params, dict, dictSize, dictContentType, cctx->entropyWorkspace);
-        if (ZSTD_isError(dictID)) return dictID;
-        assert(dictID <= (size_t)(U32)-1);
+                &params, dict, dictSize, dictContentType, dtlm, cctx->entropyWorkspace);
+        FORWARD_IF_ERROR(dictID);
+        assert(dictID <= UINT_MAX);
         cctx->dictID = (U32)dictID;
     }
     return 0;
@@ -2408,15 +3369,16 @@
 size_t ZSTD_compressBegin_advanced_internal(ZSTD_CCtx* cctx,
                                     const void* dict, size_t dictSize,
                                     ZSTD_dictContentType_e dictContentType,
+                                    ZSTD_dictTableLoadMethod_e dtlm,
                                     const ZSTD_CDict* cdict,
                                     ZSTD_CCtx_params params,
                                     unsigned long long pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_compressBegin_advanced_internal: wlog=%u", params.cParams.windowLog);
     /* compression parameters verification and optimization */
-    CHECK_F( ZSTD_checkCParams(params.cParams) );
+    FORWARD_IF_ERROR( ZSTD_checkCParams(params.cParams) );
     return ZSTD_compressBegin_internal(cctx,
-                                       dict, dictSize, dictContentType,
+                                       dict, dictSize, dictContentType, dtlm,
                                        cdict,
                                        params, pledgedSrcSize,
                                        ZSTDb_not_buffered);
@@ -2431,7 +3393,7 @@
     ZSTD_CCtx_params const cctxParams =
             ZSTD_assignParamsToCCtxParams(cctx->requestedParams, params);
     return ZSTD_compressBegin_advanced_internal(cctx,
-                                            dict, dictSize, ZSTD_dct_auto,
+                                            dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast,
                                             NULL /*cdict*/,
                                             cctxParams, pledgedSrcSize);
 }
@@ -2441,8 +3403,8 @@
     ZSTD_parameters const params = ZSTD_getParams(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, dictSize);
     ZSTD_CCtx_params const cctxParams =
             ZSTD_assignParamsToCCtxParams(cctx->requestedParams, params);
-    DEBUGLOG(4, "ZSTD_compressBegin_usingDict (dictSize=%u)", (U32)dictSize);
-    return ZSTD_compressBegin_internal(cctx, dict, dictSize, ZSTD_dct_auto, NULL,
+    DEBUGLOG(4, "ZSTD_compressBegin_usingDict (dictSize=%u)", (unsigned)dictSize);
+    return ZSTD_compressBegin_internal(cctx, dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast, NULL,
                                        cctxParams, ZSTD_CONTENTSIZE_UNKNOWN, ZSTDb_not_buffered);
 }
 
@@ -2462,12 +3424,12 @@
     size_t fhSize = 0;
 
     DEBUGLOG(4, "ZSTD_writeEpilogue");
-    if (cctx->stage == ZSTDcs_created) return ERROR(stage_wrong);  /* init missing */
+    RETURN_ERROR_IF(cctx->stage == ZSTDcs_created, stage_wrong, "init missing");
 
     /* special case : empty frame */
     if (cctx->stage == ZSTDcs_init) {
         fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, cctx->appliedParams, 0, 0);
-        if (ZSTD_isError(fhSize)) return fhSize;
+        FORWARD_IF_ERROR(fhSize);
         dstCapacity -= fhSize;
         op += fhSize;
         cctx->stage = ZSTDcs_ongoing;
@@ -2476,7 +3438,7 @@
     if (cctx->stage != ZSTDcs_ending) {
         /* write one last empty block, make it the "last" block */
         U32 const cBlockHeader24 = 1 /* last block */ + (((U32)bt_raw)<<1) + 0;
-        if (dstCapacity<4) return ERROR(dstSize_tooSmall);
+        RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall);
         MEM_writeLE32(op, cBlockHeader24);
         op += ZSTD_blockHeaderSize;
         dstCapacity -= ZSTD_blockHeaderSize;
@@ -2484,8 +3446,8 @@
 
     if (cctx->appliedParams.fParams.checksumFlag) {
         U32 const checksum = (U32) XXH64_digest(&cctx->xxhState);
-        if (dstCapacity<4) return ERROR(dstSize_tooSmall);
-        DEBUGLOG(4, "ZSTD_writeEpilogue: write checksum : %08X", checksum);
+        RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall);
+        DEBUGLOG(4, "ZSTD_writeEpilogue: write checksum : %08X", (unsigned)checksum);
         MEM_writeLE32(op, checksum);
         op += 4;
     }
@@ -2502,45 +3464,53 @@
     size_t const cSize = ZSTD_compressContinue_internal(cctx,
                                 dst, dstCapacity, src, srcSize,
                                 1 /* frame mode */, 1 /* last chunk */);
-    if (ZSTD_isError(cSize)) return cSize;
+    FORWARD_IF_ERROR(cSize);
     endResult = ZSTD_writeEpilogue(cctx, (char*)dst + cSize, dstCapacity-cSize);
-    if (ZSTD_isError(endResult)) return endResult;
-    if (cctx->appliedParams.fParams.contentSizeFlag) {  /* control src size */
+    FORWARD_IF_ERROR(endResult);
+    assert(!(cctx->appliedParams.fParams.contentSizeFlag && cctx->pledgedSrcSizePlusOne == 0));
+    if (cctx->pledgedSrcSizePlusOne != 0) {  /* control src size */
+        ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN == (unsigned long long)-1);
         DEBUGLOG(4, "end of frame : controlling src size");
-        if (cctx->pledgedSrcSizePlusOne != cctx->consumedSrcSize+1) {
-            DEBUGLOG(4, "error : pledgedSrcSize = %u, while realSrcSize = %u",
-                (U32)cctx->pledgedSrcSizePlusOne-1, (U32)cctx->consumedSrcSize);
-            return ERROR(srcSize_wrong);
-    }   }
+        RETURN_ERROR_IF(
+            cctx->pledgedSrcSizePlusOne != cctx->consumedSrcSize+1,
+            srcSize_wrong,
+             "error : pledgedSrcSize = %u, while realSrcSize = %u",
+            (unsigned)cctx->pledgedSrcSizePlusOne-1,
+            (unsigned)cctx->consumedSrcSize);
+    }
     return cSize + endResult;
 }
 
 
 static size_t ZSTD_compress_internal (ZSTD_CCtx* cctx,
-                               void* dst, size_t dstCapacity,
-                         const void* src, size_t srcSize,
-                         const void* dict,size_t dictSize,
-                               ZSTD_parameters params)
+                                      void* dst, size_t dstCapacity,
+                                const void* src, size_t srcSize,
+                                const void* dict,size_t dictSize,
+                                      ZSTD_parameters params)
 {
     ZSTD_CCtx_params const cctxParams =
             ZSTD_assignParamsToCCtxParams(cctx->requestedParams, params);
     DEBUGLOG(4, "ZSTD_compress_internal");
     return ZSTD_compress_advanced_internal(cctx,
-                                          dst, dstCapacity,
-                                          src, srcSize,
-                                          dict, dictSize,
-                                          cctxParams);
+                                           dst, dstCapacity,
+                                           src, srcSize,
+                                           dict, dictSize,
+                                           cctxParams);
 }
 
-size_t ZSTD_compress_advanced (ZSTD_CCtx* ctx,
+size_t ZSTD_compress_advanced (ZSTD_CCtx* cctx,
                                void* dst, size_t dstCapacity,
                          const void* src, size_t srcSize,
                          const void* dict,size_t dictSize,
                                ZSTD_parameters params)
 {
     DEBUGLOG(4, "ZSTD_compress_advanced");
-    CHECK_F(ZSTD_checkCParams(params.cParams));
-    return ZSTD_compress_internal(ctx, dst, dstCapacity, src, srcSize, dict, dictSize, params);
+    FORWARD_IF_ERROR(ZSTD_checkCParams(params.cParams));
+    return ZSTD_compress_internal(cctx,
+                                  dst, dstCapacity,
+                                  src, srcSize,
+                                  dict, dictSize,
+                                  params);
 }
 
 /* Internal */
@@ -2551,37 +3521,44 @@
         const void* dict,size_t dictSize,
         ZSTD_CCtx_params params)
 {
-    DEBUGLOG(4, "ZSTD_compress_advanced_internal (srcSize:%u)",
-                (U32)srcSize);
-    CHECK_F( ZSTD_compressBegin_internal(cctx, dict, dictSize, ZSTD_dct_auto, NULL,
-                                         params, srcSize, ZSTDb_not_buffered) );
+    DEBUGLOG(4, "ZSTD_compress_advanced_internal (srcSize:%u)", (unsigned)srcSize);
+    FORWARD_IF_ERROR( ZSTD_compressBegin_internal(cctx,
+                         dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast, NULL,
+                         params, srcSize, ZSTDb_not_buffered) );
     return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize);
 }
 
-size_t ZSTD_compress_usingDict(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize,
-                               const void* dict, size_t dictSize, int compressionLevel)
+size_t ZSTD_compress_usingDict(ZSTD_CCtx* cctx,
+                               void* dst, size_t dstCapacity,
+                         const void* src, size_t srcSize,
+                         const void* dict, size_t dictSize,
+                               int compressionLevel)
 {
-    ZSTD_parameters const params = ZSTD_getParams(compressionLevel, srcSize ? srcSize : 1, dict ? dictSize : 0);
+    ZSTD_parameters const params = ZSTD_getParams(compressionLevel, srcSize + (!srcSize), dict ? dictSize : 0);
     ZSTD_CCtx_params cctxParams = ZSTD_assignParamsToCCtxParams(cctx->requestedParams, params);
     assert(params.fParams.contentSizeFlag == 1);
-    ZSTD_CCtxParam_setParameter(&cctxParams, ZSTD_p_compressLiterals, compressionLevel>=0);
     return ZSTD_compress_advanced_internal(cctx, dst, dstCapacity, src, srcSize, dict, dictSize, cctxParams);
 }
 
-size_t ZSTD_compressCCtx (ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize, int compressionLevel)
+size_t ZSTD_compressCCtx(ZSTD_CCtx* cctx,
+                         void* dst, size_t dstCapacity,
+                   const void* src, size_t srcSize,
+                         int compressionLevel)
 {
-    DEBUGLOG(4, "ZSTD_compressCCtx (srcSize=%u)", (U32)srcSize);
+    DEBUGLOG(4, "ZSTD_compressCCtx (srcSize=%u)", (unsigned)srcSize);
+    assert(cctx != NULL);
     return ZSTD_compress_usingDict(cctx, dst, dstCapacity, src, srcSize, NULL, 0, compressionLevel);
 }
 
-size_t ZSTD_compress(void* dst, size_t dstCapacity, const void* src, size_t srcSize, int compressionLevel)
+size_t ZSTD_compress(void* dst, size_t dstCapacity,
+               const void* src, size_t srcSize,
+                     int compressionLevel)
 {
     size_t result;
     ZSTD_CCtx ctxBody;
-    memset(&ctxBody, 0, sizeof(ctxBody));
-    ctxBody.customMem = ZSTD_defaultCMem;
+    ZSTD_initCCtx(&ctxBody, ZSTD_defaultCMem);
     result = ZSTD_compressCCtx(&ctxBody, dst, dstCapacity, src, srcSize, compressionLevel);
-    ZSTD_free(ctxBody.workSpace, ZSTD_defaultCMem);  /* can't free ctxBody itself, as it's on stack; free only heap content */
+    ZSTD_freeCCtxContent(&ctxBody);   /* can't free ctxBody itself, as it's on stack; free only heap content */
     return result;
 }
 
@@ -2594,7 +3571,7 @@
         size_t dictSize, ZSTD_compressionParameters cParams,
         ZSTD_dictLoadMethod_e dictLoadMethod)
 {
-    DEBUGLOG(5, "sizeof(ZSTD_CDict) : %u", (U32)sizeof(ZSTD_CDict));
+    DEBUGLOG(5, "sizeof(ZSTD_CDict) : %u", (unsigned)sizeof(ZSTD_CDict));
     return sizeof(ZSTD_CDict) + HUF_WORKSPACE_SIZE + ZSTD_sizeof_matchState(&cParams, /* forCCtx */ 0)
            + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize);
 }
@@ -2608,7 +3585,7 @@
 size_t ZSTD_sizeof_CDict(const ZSTD_CDict* cdict)
 {
     if (cdict==NULL) return 0;   /* support sizeof on NULL */
-    DEBUGLOG(5, "sizeof(*cdict) : %u", (U32)sizeof(*cdict));
+    DEBUGLOG(5, "sizeof(*cdict) : %u", (unsigned)sizeof(*cdict));
     return cdict->workspaceSize + (cdict->dictBuffer ? cdict->dictContentSize : 0) + sizeof(*cdict);
 }
 
@@ -2619,9 +3596,9 @@
                     ZSTD_dictContentType_e dictContentType,
                     ZSTD_compressionParameters cParams)
 {
-    DEBUGLOG(3, "ZSTD_initCDict_internal, dictContentType %u", (U32)dictContentType);
+    DEBUGLOG(3, "ZSTD_initCDict_internal (dictContentType:%u)", (unsigned)dictContentType);
     assert(!ZSTD_checkCParams(cParams));
-    cdict->cParams = cParams;
+    cdict->matchState.cParams = cParams;
     if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dictBuffer) || (!dictSize)) {
         cdict->dictBuffer = NULL;
         cdict->dictContent = dictBuffer;
@@ -2629,17 +3606,17 @@
         void* const internalBuffer = ZSTD_malloc(dictSize, cdict->customMem);
         cdict->dictBuffer = internalBuffer;
         cdict->dictContent = internalBuffer;
-        if (!internalBuffer) return ERROR(memory_allocation);
+        RETURN_ERROR_IF(!internalBuffer, memory_allocation);
         memcpy(internalBuffer, dictBuffer, dictSize);
     }
     cdict->dictContentSize = dictSize;
 
     /* Reset the state to no dictionary */
     ZSTD_reset_compressedBlockState(&cdict->cBlockState);
-    {   void* const end = ZSTD_reset_matchState(
-                &cdict->matchState,
-                (U32*)cdict->workspace + HUF_WORKSPACE_SIZE_U32,
-                &cParams, ZSTDcrp_continue, /* forCCtx */ 0);
+    {   void* const end = ZSTD_reset_matchState(&cdict->matchState,
+                            (U32*)cdict->workspace + HUF_WORKSPACE_SIZE_U32,
+                            &cParams,
+                             ZSTDcrp_continue, ZSTD_resetTarget_CDict);
         assert(end == (char*)cdict->workspace + cdict->workspaceSize);
         (void)end;
     }
@@ -2654,8 +3631,8 @@
         {   size_t const dictID = ZSTD_compress_insertDictionary(
                     &cdict->cBlockState, &cdict->matchState, &params,
                     cdict->dictContent, cdict->dictContentSize,
-                    dictContentType, cdict->workspace);
-            if (ZSTD_isError(dictID)) return dictID;
+                    dictContentType, ZSTD_dtlm_full, cdict->workspace);
+            FORWARD_IF_ERROR(dictID);
             assert(dictID <= (size_t)(U32)-1);
             cdict->dictID = (U32)dictID;
         }
@@ -2669,7 +3646,7 @@
                                       ZSTD_dictContentType_e dictContentType,
                                       ZSTD_compressionParameters cParams, ZSTD_customMem customMem)
 {
-    DEBUGLOG(3, "ZSTD_createCDict_advanced, mode %u", (U32)dictContentType);
+    DEBUGLOG(3, "ZSTD_createCDict_advanced, mode %u", (unsigned)dictContentType);
     if (!customMem.customAlloc ^ !customMem.customFree) return NULL;
 
     {   ZSTD_CDict* const cdict = (ZSTD_CDict*)ZSTD_malloc(sizeof(ZSTD_CDict), customMem);
@@ -2750,7 +3727,7 @@
     void* ptr;
     if ((size_t)workspace & 7) return NULL;  /* 8-aligned */
     DEBUGLOG(4, "(workspaceSize < neededSize) : (%u < %u) => %u",
-        (U32)workspaceSize, (U32)neededSize, (U32)(workspaceSize < neededSize));
+        (unsigned)workspaceSize, (unsigned)neededSize, (unsigned)(workspaceSize < neededSize));
     if (workspaceSize < neededSize) return NULL;
 
     if (dictLoadMethod == ZSTD_dlm_byCopy) {
@@ -2775,7 +3752,7 @@
 ZSTD_compressionParameters ZSTD_getCParamsFromCDict(const ZSTD_CDict* cdict)
 {
     assert(cdict != NULL);
-    return cdict->cParams;
+    return cdict->matchState.cParams;
 }
 
 /* ZSTD_compressBegin_usingCDict_advanced() :
@@ -2785,7 +3762,7 @@
     ZSTD_frameParameters const fParams, unsigned long long const pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_compressBegin_usingCDict_advanced");
-    if (cdict==NULL) return ERROR(dictionary_wrong);
+    RETURN_ERROR_IF(cdict==NULL, dictionary_wrong);
     {   ZSTD_CCtx_params params = cctx->requestedParams;
         params.cParams = ZSTD_getCParamsFromCDict(cdict);
         /* Increase window log to fit the entire dictionary and source if the
@@ -2799,7 +3776,7 @@
         }
         params.fParams = fParams;
         return ZSTD_compressBegin_internal(cctx,
-                                           NULL, 0, ZSTD_dct_auto,
+                                           NULL, 0, ZSTD_dct_auto, ZSTD_dtlm_fast,
                                            cdict,
                                            params, pledgedSrcSize,
                                            ZSTDb_not_buffered);
@@ -2813,7 +3790,7 @@
 {
     ZSTD_frameParameters const fParams = { 0 /*content*/, 0 /*checksum*/, 0 /*noDictID*/ };
     DEBUGLOG(4, "ZSTD_compressBegin_usingCDict : dictIDFlag == %u", !fParams.noDictIDFlag);
-    return ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, 0);
+    return ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, ZSTD_CONTENTSIZE_UNKNOWN);
 }
 
 size_t ZSTD_compress_usingCDict_advanced(ZSTD_CCtx* cctx,
@@ -2821,7 +3798,7 @@
                                 const void* src, size_t srcSize,
                                 const ZSTD_CDict* cdict, ZSTD_frameParameters fParams)
 {
-    CHECK_F (ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, srcSize));   /* will check if cdict != NULL */
+    FORWARD_IF_ERROR(ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, srcSize));   /* will check if cdict != NULL */
     return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize);
 }
 
@@ -2880,16 +3857,17 @@
 static size_t ZSTD_resetCStream_internal(ZSTD_CStream* cctx,
                     const void* const dict, size_t const dictSize, ZSTD_dictContentType_e const dictContentType,
                     const ZSTD_CDict* const cdict,
-                    ZSTD_CCtx_params const params, unsigned long long const pledgedSrcSize)
+                    ZSTD_CCtx_params params, unsigned long long const pledgedSrcSize)
 {
-    DEBUGLOG(4, "ZSTD_resetCStream_internal (disableLiteralCompression=%i)",
-                params.disableLiteralCompression);
+    DEBUGLOG(4, "ZSTD_resetCStream_internal");
+    /* Finalize the compression parameters */
+    params.cParams = ZSTD_getCParamsFromCCtxParams(&params, pledgedSrcSize, dictSize);
     /* params are supposed to be fully validated at this point */
     assert(!ZSTD_isError(ZSTD_checkCParams(params.cParams)));
     assert(!((dict) && (cdict)));  /* either dict or cdict, not both */
 
-    CHECK_F( ZSTD_compressBegin_internal(cctx,
-                                         dict, dictSize, dictContentType,
+    FORWARD_IF_ERROR( ZSTD_compressBegin_internal(cctx,
+                                         dict, dictSize, dictContentType, ZSTD_dtlm_fast,
                                          cdict,
                                          params, pledgedSrcSize,
                                          ZSTDb_buffered) );
@@ -2906,14 +3884,17 @@
 
 /* ZSTD_resetCStream():
  * pledgedSrcSize == 0 means "unknown" */
-size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize)
+size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pss)
 {
-    ZSTD_CCtx_params params = zcs->requestedParams;
-    DEBUGLOG(4, "ZSTD_resetCStream: pledgedSrcSize = %u", (U32)pledgedSrcSize);
-    if (pledgedSrcSize==0) pledgedSrcSize = ZSTD_CONTENTSIZE_UNKNOWN;
-    params.fParams.contentSizeFlag = 1;
-    params.cParams = ZSTD_getCParamsFromCCtxParams(&params, pledgedSrcSize, 0);
-    return ZSTD_resetCStream_internal(zcs, NULL, 0, ZSTD_dct_auto, zcs->cdict, params, pledgedSrcSize);
+    /* temporary : 0 interpreted as "unknown" during transition period.
+     * Users willing to specify "unknown" **must** use ZSTD_CONTENTSIZE_UNKNOWN.
+     * 0 will be interpreted as "empty" in the future.
+     */
+    U64 const pledgedSrcSize = (pss==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss;
+    DEBUGLOG(4, "ZSTD_resetCStream: pledgedSrcSize = %u", (unsigned)pledgedSrcSize);
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) );
+    return 0;
 }
 
 /*! ZSTD_initCStream_internal() :
@@ -2925,31 +3906,18 @@
                     ZSTD_CCtx_params params, unsigned long long pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_initCStream_internal");
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) );
     assert(!ZSTD_isError(ZSTD_checkCParams(params.cParams)));
+    zcs->requestedParams = params;
     assert(!((dict) && (cdict)));  /* either dict or cdict, not both */
-
-    if (dict && dictSize >= 8) {
-        DEBUGLOG(4, "loading dictionary of size %u", (U32)dictSize);
-        if (zcs->staticSize) {   /* static CCtx : never uses malloc */
-            /* incompatible with internal cdict creation */
-            return ERROR(memory_allocation);
-        }
-        ZSTD_freeCDict(zcs->cdictLocal);
-        zcs->cdictLocal = ZSTD_createCDict_advanced(dict, dictSize,
-                                            ZSTD_dlm_byCopy, ZSTD_dct_auto,
-                                            params.cParams, zcs->customMem);
-        zcs->cdict = zcs->cdictLocal;
-        if (zcs->cdictLocal == NULL) return ERROR(memory_allocation);
+    if (dict) {
+        FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) );
     } else {
-        if (cdict) {
-            params.cParams = ZSTD_getCParamsFromCDict(cdict);  /* cParams are enforced from cdict; it includes windowLog */
-        }
-        ZSTD_freeCDict(zcs->cdictLocal);
-        zcs->cdictLocal = NULL;
-        zcs->cdict = cdict;
+        /* Dictionary is cleared if !cdict */
+        FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) );
     }
-
-    return ZSTD_resetCStream_internal(zcs, NULL, 0, ZSTD_dct_auto, zcs->cdict, params, pledgedSrcSize);
+    return 0;
 }
 
 /* ZSTD_initCStream_usingCDict_advanced() :
@@ -2960,22 +3928,20 @@
                                             unsigned long long pledgedSrcSize)
 {
     DEBUGLOG(4, "ZSTD_initCStream_usingCDict_advanced");
-    if (!cdict) return ERROR(dictionary_wrong); /* cannot handle NULL cdict (does not know what to do) */
-    {   ZSTD_CCtx_params params = zcs->requestedParams;
-        params.cParams = ZSTD_getCParamsFromCDict(cdict);
-        params.fParams = fParams;
-        return ZSTD_initCStream_internal(zcs,
-                                NULL, 0, cdict,
-                                params, pledgedSrcSize);
-    }
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) );
+    zcs->requestedParams.fParams = fParams;
+    FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) );
+    return 0;
 }
 
 /* note : cdict must outlive compression session */
 size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict)
 {
-    ZSTD_frameParameters const fParams = { 0 /* contentSizeFlag */, 0 /* checksum */, 0 /* hideDictID */ };
     DEBUGLOG(4, "ZSTD_initCStream_usingCDict");
-    return ZSTD_initCStream_usingCDict_advanced(zcs, cdict, fParams, ZSTD_CONTENTSIZE_UNKNOWN);  /* note : will check that cdict != NULL */
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) );
+    return 0;
 }
 
 
@@ -2985,43 +3951,66 @@
  * dict is loaded with default parameters ZSTD_dm_auto and ZSTD_dlm_byCopy. */
 size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs,
                                  const void* dict, size_t dictSize,
-                                 ZSTD_parameters params, unsigned long long pledgedSrcSize)
+                                 ZSTD_parameters params, unsigned long long pss)
 {
-    DEBUGLOG(4, "ZSTD_initCStream_advanced: pledgedSrcSize=%u, flag=%u",
-                (U32)pledgedSrcSize, params.fParams.contentSizeFlag);
-    CHECK_F( ZSTD_checkCParams(params.cParams) );
-    if ((pledgedSrcSize==0) && (params.fParams.contentSizeFlag==0)) pledgedSrcSize = ZSTD_CONTENTSIZE_UNKNOWN;  /* for compatibility with older programs relying on this behavior. Users should now specify ZSTD_CONTENTSIZE_UNKNOWN. This line will be removed in the future. */
-    {   ZSTD_CCtx_params const cctxParams = ZSTD_assignParamsToCCtxParams(zcs->requestedParams, params);
-        return ZSTD_initCStream_internal(zcs, dict, dictSize, NULL /*cdict*/, cctxParams, pledgedSrcSize);
-    }
+    /* for compatibility with older programs relying on this behavior.
+     * Users should now specify ZSTD_CONTENTSIZE_UNKNOWN.
+     * This line will be removed in the future.
+     */
+    U64 const pledgedSrcSize = (pss==0 && params.fParams.contentSizeFlag==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss;
+    DEBUGLOG(4, "ZSTD_initCStream_advanced");
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) );
+    FORWARD_IF_ERROR( ZSTD_checkCParams(params.cParams) );
+    zcs->requestedParams = ZSTD_assignParamsToCCtxParams(zcs->requestedParams, params);
+    FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) );
+    return 0;
 }
 
 size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, const void* dict, size_t dictSize, int compressionLevel)
 {
-    ZSTD_parameters const params = ZSTD_getParams(compressionLevel, 0, dictSize);
-    ZSTD_CCtx_params const cctxParams =
-            ZSTD_assignParamsToCCtxParams(zcs->requestedParams, params);
-    return ZSTD_initCStream_internal(zcs, dict, dictSize, NULL, cctxParams, ZSTD_CONTENTSIZE_UNKNOWN);
+    DEBUGLOG(4, "ZSTD_initCStream_usingDict");
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) );
+    return 0;
 }
 
 size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, int compressionLevel, unsigned long long pss)
 {
-    U64 const pledgedSrcSize = (pss==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss;  /* temporary : 0 interpreted as "unknown" during transition period. Users willing to specify "unknown" **must** use ZSTD_CONTENTSIZE_UNKNOWN. `0` will be interpreted as "empty" in the future */
-    ZSTD_parameters const params = ZSTD_getParams(compressionLevel, pledgedSrcSize, 0);
-    ZSTD_CCtx_params const cctxParams = ZSTD_assignParamsToCCtxParams(zcs->requestedParams, params);
-    return ZSTD_initCStream_internal(zcs, NULL, 0, NULL, cctxParams, pledgedSrcSize);
+    /* temporary : 0 interpreted as "unknown" during transition period.
+     * Users willing to specify "unknown" **must** use ZSTD_CONTENTSIZE_UNKNOWN.
+     * 0 will be interpreted as "empty" in the future.
+     */
+    U64 const pledgedSrcSize = (pss==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss;
+    DEBUGLOG(4, "ZSTD_initCStream_srcSize");
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, NULL) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) );
+    return 0;
 }
 
 size_t ZSTD_initCStream(ZSTD_CStream* zcs, int compressionLevel)
 {
     DEBUGLOG(4, "ZSTD_initCStream");
-    return ZSTD_initCStream_srcSize(zcs, compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN);
+    FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, NULL) );
+    FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) );
+    return 0;
 }
 
 /*======   Compression   ======*/
 
-MEM_STATIC size_t ZSTD_limitCopy(void* dst, size_t dstCapacity,
-                           const void* src, size_t srcSize)
+static size_t ZSTD_nextInputSizeHint(const ZSTD_CCtx* cctx)
+{
+    size_t hintInSize = cctx->inBuffTarget - cctx->inBuffPos;
+    if (hintInSize==0) hintInSize = cctx->blockSize;
+    return hintInSize;
+}
+
+static size_t ZSTD_limitCopy(void* dst, size_t dstCapacity,
+                       const void* src, size_t srcSize)
 {
     size_t const length = MIN(dstCapacity, srcSize);
     if (length) memcpy(dst, src, length);
@@ -3029,13 +4018,13 @@
 }
 
 /** ZSTD_compressStream_generic():
- *  internal function for all *compressStream*() variants and *compress_generic()
+ *  internal function for all *compressStream*() variants
  *  non-static, because can be called from zstdmt_compress.c
  * @return : hint size for next input */
-size_t ZSTD_compressStream_generic(ZSTD_CStream* zcs,
-                                   ZSTD_outBuffer* output,
-                                   ZSTD_inBuffer* input,
-                                   ZSTD_EndDirective const flushMode)
+static size_t ZSTD_compressStream_generic(ZSTD_CStream* zcs,
+                                          ZSTD_outBuffer* output,
+                                          ZSTD_inBuffer* input,
+                                          ZSTD_EndDirective const flushMode)
 {
     const char* const istart = (const char*)input->src;
     const char* const iend = istart + input->size;
@@ -3046,7 +4035,7 @@
     U32 someMoreWork = 1;
 
     /* check expectations */
-    DEBUGLOG(5, "ZSTD_compressStream_generic, flush=%u", (U32)flushMode);
+    DEBUGLOG(5, "ZSTD_compressStream_generic, flush=%u", (unsigned)flushMode);
     assert(zcs->inBuff != NULL);
     assert(zcs->inBuffSize > 0);
     assert(zcs->outBuff !=  NULL);
@@ -3058,8 +4047,7 @@
         switch(zcs->streamStage)
         {
         case zcss_init:
-            /* call ZSTD_initCStream() first ! */
-            return ERROR(init_missing);
+            RETURN_ERROR(init_missing, "call ZSTD_initCStream() first!");
 
         case zcss_load:
             if ( (flushMode == ZSTD_e_end)
@@ -3068,12 +4056,12 @@
                 /* shortcut to compression pass directly into output buffer */
                 size_t const cSize = ZSTD_compressEnd(zcs,
                                                 op, oend-op, ip, iend-ip);
-                DEBUGLOG(4, "ZSTD_compressEnd : %u", (U32)cSize);
-                if (ZSTD_isError(cSize)) return cSize;
+                DEBUGLOG(4, "ZSTD_compressEnd : cSize=%u", (unsigned)cSize);
+                FORWARD_IF_ERROR(cSize);
                 ip = iend;
                 op += cSize;
                 zcs->frameEnded = 1;
-                ZSTD_startNewCompression(zcs);
+                ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
                 someMoreWork = 0; break;
             }
             /* complete loading into inBuffer */
@@ -3110,14 +4098,14 @@
                                     zcs->inBuff + zcs->inToCompress, iSize) :
                         ZSTD_compressContinue(zcs, cDst, oSize,
                                     zcs->inBuff + zcs->inToCompress, iSize);
-                if (ZSTD_isError(cSize)) return cSize;
+                FORWARD_IF_ERROR(cSize);
                 zcs->frameEnded = lastBlock;
                 /* prepare next block */
                 zcs->inBuffTarget = zcs->inBuffPos + zcs->blockSize;
                 if (zcs->inBuffTarget > zcs->inBuffSize)
                     zcs->inBuffPos = 0, zcs->inBuffTarget = zcs->blockSize;
                 DEBUGLOG(5, "inBuffTarget:%u / inBuffSize:%u",
-                         (U32)zcs->inBuffTarget, (U32)zcs->inBuffSize);
+                         (unsigned)zcs->inBuffTarget, (unsigned)zcs->inBuffSize);
                 if (!lastBlock)
                     assert(zcs->inBuffTarget <= zcs->inBuffSize);
                 zcs->inToCompress = zcs->inBuffPos;
@@ -3126,7 +4114,7 @@
                     if (zcs->frameEnded) {
                         DEBUGLOG(5, "Frame completed directly in outBuffer");
                         someMoreWork = 0;
-                        ZSTD_startNewCompression(zcs);
+                        ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
                     }
                     break;
                 }
@@ -3138,10 +4126,10 @@
         case zcss_flush:
             DEBUGLOG(5, "flush stage");
             {   size_t const toFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize;
-                size_t const flushed = ZSTD_limitCopy(op, oend-op,
+                size_t const flushed = ZSTD_limitCopy(op, (size_t)(oend-op),
                             zcs->outBuff + zcs->outBuffFlushedSize, toFlush);
                 DEBUGLOG(5, "toFlush: %u into %u ==> flushed: %u",
-                            (U32)toFlush, (U32)(oend-op), (U32)flushed);
+                            (unsigned)toFlush, (unsigned)(oend-op), (unsigned)flushed);
                 op += flushed;
                 zcs->outBuffFlushedSize += flushed;
                 if (toFlush!=flushed) {
@@ -3154,7 +4142,7 @@
                 if (zcs->frameEnded) {
                     DEBUGLOG(5, "Frame completed on flush");
                     someMoreWork = 0;
-                    ZSTD_startNewCompression(zcs);
+                    ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
                     break;
                 }
                 zcs->streamStage = zcss_load;
@@ -3169,63 +4157,67 @@
     input->pos = ip - istart;
     output->pos = op - ostart;
     if (zcs->frameEnded) return 0;
-    {   size_t hintInSize = zcs->inBuffTarget - zcs->inBuffPos;
-        if (hintInSize==0) hintInSize = zcs->blockSize;
-        return hintInSize;
+    return ZSTD_nextInputSizeHint(zcs);
+}
+
+static size_t ZSTD_nextInputSizeHint_MTorST(const ZSTD_CCtx* cctx)
+{
+#ifdef ZSTD_MULTITHREAD
+    if (cctx->appliedParams.nbWorkers >= 1) {
+        assert(cctx->mtctx != NULL);
+        return ZSTDMT_nextInputSizeHint(cctx->mtctx);
     }
+#endif
+    return ZSTD_nextInputSizeHint(cctx);
+
 }
 
 size_t ZSTD_compressStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output, ZSTD_inBuffer* input)
 {
-    /* check conditions */
-    if (output->pos > output->size) return ERROR(GENERIC);
-    if (input->pos  > input->size)  return ERROR(GENERIC);
-
-    return ZSTD_compressStream_generic(zcs, output, input, ZSTD_e_continue);
+    FORWARD_IF_ERROR( ZSTD_compressStream2(zcs, output, input, ZSTD_e_continue) );
+    return ZSTD_nextInputSizeHint_MTorST(zcs);
 }
 
 
-size_t ZSTD_compress_generic (ZSTD_CCtx* cctx,
-                              ZSTD_outBuffer* output,
-                              ZSTD_inBuffer* input,
-                              ZSTD_EndDirective endOp)
+size_t ZSTD_compressStream2( ZSTD_CCtx* cctx,
+                             ZSTD_outBuffer* output,
+                             ZSTD_inBuffer* input,
+                             ZSTD_EndDirective endOp)
 {
-    DEBUGLOG(5, "ZSTD_compress_generic, endOp=%u ", (U32)endOp);
+    DEBUGLOG(5, "ZSTD_compressStream2, endOp=%u ", (unsigned)endOp);
     /* check conditions */
-    if (output->pos > output->size) return ERROR(GENERIC);
-    if (input->pos  > input->size)  return ERROR(GENERIC);
+    RETURN_ERROR_IF(output->pos > output->size, GENERIC);
+    RETURN_ERROR_IF(input->pos  > input->size, GENERIC);
     assert(cctx!=NULL);
 
     /* transparent initialization stage */
     if (cctx->streamStage == zcss_init) {
         ZSTD_CCtx_params params = cctx->requestedParams;
         ZSTD_prefixDict const prefixDict = cctx->prefixDict;
-        memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict));  /* single usage */
-        assert(prefixDict.dict==NULL || cctx->cdict==NULL);   /* only one can be set */
-        DEBUGLOG(4, "ZSTD_compress_generic : transparent init stage");
+        FORWARD_IF_ERROR( ZSTD_initLocalDict(cctx) ); /* Init the local dict if present. */
+        memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict));   /* single usage */
+        assert(prefixDict.dict==NULL || cctx->cdict==NULL);    /* only one can be set */
+        DEBUGLOG(4, "ZSTD_compressStream2 : transparent init stage");
         if (endOp == ZSTD_e_end) cctx->pledgedSrcSizePlusOne = input->size + 1;  /* auto-fix pledgedSrcSize */
         params.cParams = ZSTD_getCParamsFromCCtxParams(
                 &cctx->requestedParams, cctx->pledgedSrcSizePlusOne-1, 0 /*dictSize*/);
 
+
 #ifdef ZSTD_MULTITHREAD
         if ((cctx->pledgedSrcSizePlusOne-1) <= ZSTDMT_JOBSIZE_MIN) {
             params.nbWorkers = 0; /* do not invoke multi-threading when src size is too small */
         }
         if (params.nbWorkers > 0) {
             /* mt context creation */
-            if (cctx->mtctx == NULL || (params.nbWorkers != ZSTDMT_getNbWorkers(cctx->mtctx))) {
-                DEBUGLOG(4, "ZSTD_compress_generic: creating new mtctx for nbWorkers=%u",
+            if (cctx->mtctx == NULL) {
+                DEBUGLOG(4, "ZSTD_compressStream2: creating new mtctx for nbWorkers=%u",
                             params.nbWorkers);
-                if (cctx->mtctx != NULL)
-                    DEBUGLOG(4, "ZSTD_compress_generic: previous nbWorkers was %u",
-                                ZSTDMT_getNbWorkers(cctx->mtctx));
-                ZSTDMT_freeCCtx(cctx->mtctx);
                 cctx->mtctx = ZSTDMT_createCCtx_advanced(params.nbWorkers, cctx->customMem);
-                if (cctx->mtctx == NULL) return ERROR(memory_allocation);
+                RETURN_ERROR_IF(cctx->mtctx == NULL, memory_allocation);
             }
             /* mt compression */
             DEBUGLOG(4, "call ZSTDMT_initCStream_internal as nbWorkers=%u", params.nbWorkers);
-            CHECK_F( ZSTDMT_initCStream_internal(
+            FORWARD_IF_ERROR( ZSTDMT_initCStream_internal(
                         cctx->mtctx,
                         prefixDict.dict, prefixDict.dictSize, ZSTD_dct_rawContent,
                         cctx->cdict, params, cctx->pledgedSrcSizePlusOne-1) );
@@ -3233,35 +4225,47 @@
             cctx->appliedParams.nbWorkers = params.nbWorkers;
         } else
 #endif
-        {   CHECK_F( ZSTD_resetCStream_internal(cctx,
+        {   FORWARD_IF_ERROR( ZSTD_resetCStream_internal(cctx,
                             prefixDict.dict, prefixDict.dictSize, prefixDict.dictContentType,
                             cctx->cdict,
                             params, cctx->pledgedSrcSizePlusOne-1) );
             assert(cctx->streamStage == zcss_load);
             assert(cctx->appliedParams.nbWorkers == 0);
     }   }
+    /* end of transparent initialization stage */
 
     /* compression stage */
 #ifdef ZSTD_MULTITHREAD
     if (cctx->appliedParams.nbWorkers > 0) {
+        int const forceMaxProgress = (endOp == ZSTD_e_flush || endOp == ZSTD_e_end);
+        size_t flushMin;
+        assert(forceMaxProgress || endOp == ZSTD_e_continue /* Protection for a new flush type */);
         if (cctx->cParamsChanged) {
             ZSTDMT_updateCParams_whileCompressing(cctx->mtctx, &cctx->requestedParams);
             cctx->cParamsChanged = 0;
         }
-        {   size_t const flushMin = ZSTDMT_compressStream_generic(cctx->mtctx, output, input, endOp);
+        do {
+            flushMin = ZSTDMT_compressStream_generic(cctx->mtctx, output, input, endOp);
             if ( ZSTD_isError(flushMin)
               || (endOp == ZSTD_e_end && flushMin == 0) ) { /* compression completed */
-                ZSTD_startNewCompression(cctx);
+                ZSTD_CCtx_reset(cctx, ZSTD_reset_session_only);
             }
-            return flushMin;
-    }   }
+            FORWARD_IF_ERROR(flushMin);
+        } while (forceMaxProgress && flushMin != 0 && output->pos < output->size);
+        DEBUGLOG(5, "completed ZSTD_compressStream2 delegating to ZSTDMT_compressStream_generic");
+        /* Either we don't require maximum forward progress, we've finished the
+         * flush, or we are out of output space.
+         */
+        assert(!forceMaxProgress || flushMin == 0 || output->pos == output->size);
+        return flushMin;
+    }
 #endif
-    CHECK_F( ZSTD_compressStream_generic(cctx, output, input, endOp) );
-    DEBUGLOG(5, "completed ZSTD_compress_generic");
+    FORWARD_IF_ERROR( ZSTD_compressStream_generic(cctx, output, input, endOp) );
+    DEBUGLOG(5, "completed ZSTD_compressStream2");
     return cctx->outBuffContentSize - cctx->outBuffFlushedSize; /* remaining to flush */
 }
 
-size_t ZSTD_compress_generic_simpleArgs (
+size_t ZSTD_compressStream2_simpleArgs (
                             ZSTD_CCtx* cctx,
                             void* dst, size_t dstCapacity, size_t* dstPos,
                       const void* src, size_t srcSize, size_t* srcPos,
@@ -3269,13 +4273,33 @@
 {
     ZSTD_outBuffer output = { dst, dstCapacity, *dstPos };
     ZSTD_inBuffer  input  = { src, srcSize, *srcPos };
-    /* ZSTD_compress_generic() will check validity of dstPos and srcPos */
-    size_t const cErr = ZSTD_compress_generic(cctx, &output, &input, endOp);
+    /* ZSTD_compressStream2() will check validity of dstPos and srcPos */
+    size_t const cErr = ZSTD_compressStream2(cctx, &output, &input, endOp);
     *dstPos = output.pos;
     *srcPos = input.pos;
     return cErr;
 }
 
+size_t ZSTD_compress2(ZSTD_CCtx* cctx,
+                      void* dst, size_t dstCapacity,
+                      const void* src, size_t srcSize)
+{
+    ZSTD_CCtx_reset(cctx, ZSTD_reset_session_only);
+    {   size_t oPos = 0;
+        size_t iPos = 0;
+        size_t const result = ZSTD_compressStream2_simpleArgs(cctx,
+                                        dst, dstCapacity, &oPos,
+                                        src, srcSize, &iPos,
+                                        ZSTD_e_end);
+        FORWARD_IF_ERROR(result);
+        if (result != 0) {  /* compression not completed, due to lack of output space */
+            assert(oPos == dstCapacity);
+            RETURN_ERROR(dstSize_tooSmall);
+        }
+        assert(iPos == srcSize);   /* all input is expected consumed */
+        return oPos;
+    }
+}
 
 /*======   Finalize   ======*/
 
@@ -3284,21 +4308,21 @@
 size_t ZSTD_flushStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output)
 {
     ZSTD_inBuffer input = { NULL, 0, 0 };
-    if (output->pos > output->size) return ERROR(GENERIC);
-    CHECK_F( ZSTD_compressStream_generic(zcs, output, &input, ZSTD_e_flush) );
-    return zcs->outBuffContentSize - zcs->outBuffFlushedSize;  /* remaining to flush */
+    return ZSTD_compressStream2(zcs, output, &input, ZSTD_e_flush);
 }
 
 
 size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output)
 {
     ZSTD_inBuffer input = { NULL, 0, 0 };
-    if (output->pos > output->size) return ERROR(GENERIC);
-    CHECK_F( ZSTD_compressStream_generic(zcs, output, &input, ZSTD_e_end) );
+    size_t const remainingToFlush = ZSTD_compressStream2(zcs, output, &input, ZSTD_e_end);
+    FORWARD_IF_ERROR( remainingToFlush );
+    if (zcs->appliedParams.nbWorkers > 0) return remainingToFlush;   /* minimal estimation */
+    /* single thread mode : attempt to calculate remaining to flush more precisely */
     {   size_t const lastBlockSize = zcs->frameEnded ? 0 : ZSTD_BLOCKHEADERSIZE;
-        size_t const checksumSize = zcs->frameEnded ? 0 : zcs->appliedParams.fParams.checksumFlag * 4;
-        size_t const toFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize + lastBlockSize + checksumSize;
-        DEBUGLOG(4, "ZSTD_endStream : remaining to flush : %u", (U32)toFlush);
+        size_t const checksumSize = (size_t)(zcs->frameEnded ? 0 : zcs->appliedParams.fParams.checksumFlag * 4);
+        size_t const toFlush = remainingToFlush + lastBlockSize + checksumSize;
+        DEBUGLOG(4, "ZSTD_endStream : remaining to flush : %u", (unsigned)toFlush);
         return toFlush;
     }
 }
@@ -3308,122 +4332,123 @@
 
 #define ZSTD_MAX_CLEVEL     22
 int ZSTD_maxCLevel(void) { return ZSTD_MAX_CLEVEL; }
+int ZSTD_minCLevel(void) { return (int)-ZSTD_TARGETLENGTH_MAX; }
 
 static const ZSTD_compressionParameters ZSTD_defaultCParameters[4][ZSTD_MAX_CLEVEL+1] = {
-{   /* "default" - guarantees a monotonically increasing memory budget */
+{   /* "default" - for any srcSize > 256 KB */
     /* W,  C,  H,  S,  L, TL, strat */
     { 19, 12, 13,  1,  6,  1, ZSTD_fast    },  /* base for negative levels */
-    { 19, 13, 14,  1,  7,  1, ZSTD_fast    },  /* level  1 */
-    { 19, 15, 16,  1,  6,  1, ZSTD_fast    },  /* level  2 */
-    { 20, 16, 17,  1,  5,  8, ZSTD_dfast   },  /* level  3 */
-    { 20, 17, 18,  1,  5,  8, ZSTD_dfast   },  /* level  4 */
-    { 20, 17, 18,  2,  5, 16, ZSTD_greedy  },  /* level  5 */
-    { 21, 17, 19,  2,  5, 16, ZSTD_lazy    },  /* level  6 */
-    { 21, 18, 19,  3,  5, 16, ZSTD_lazy    },  /* level  7 */
-    { 21, 18, 20,  3,  5, 16, ZSTD_lazy2   },  /* level  8 */
-    { 21, 19, 20,  3,  5, 16, ZSTD_lazy2   },  /* level  9 */
-    { 21, 19, 21,  4,  5, 16, ZSTD_lazy2   },  /* level 10 */
-    { 22, 20, 22,  4,  5, 16, ZSTD_lazy2   },  /* level 11 */
-    { 22, 20, 22,  5,  5, 16, ZSTD_lazy2   },  /* level 12 */
-    { 22, 21, 22,  4,  5, 32, ZSTD_btlazy2 },  /* level 13 */
-    { 22, 21, 22,  5,  5, 32, ZSTD_btlazy2 },  /* level 14 */
-    { 22, 22, 22,  6,  5, 32, ZSTD_btlazy2 },  /* level 15 */
-    { 22, 21, 22,  4,  5, 48, ZSTD_btopt   },  /* level 16 */
-    { 23, 22, 22,  4,  4, 48, ZSTD_btopt   },  /* level 17 */
-    { 23, 22, 22,  5,  3, 64, ZSTD_btopt   },  /* level 18 */
-    { 23, 23, 22,  7,  3,128, ZSTD_btopt   },  /* level 19 */
-    { 25, 25, 23,  7,  3,128, ZSTD_btultra },  /* level 20 */
-    { 26, 26, 24,  7,  3,256, ZSTD_btultra },  /* level 21 */
-    { 27, 27, 25,  9,  3,512, ZSTD_btultra },  /* level 22 */
+    { 19, 13, 14,  1,  7,  0, ZSTD_fast    },  /* level  1 */
+    { 20, 15, 16,  1,  6,  0, ZSTD_fast    },  /* level  2 */
+    { 21, 16, 17,  1,  5,  1, ZSTD_dfast   },  /* level  3 */
+    { 21, 18, 18,  1,  5,  1, ZSTD_dfast   },  /* level  4 */
+    { 21, 18, 19,  2,  5,  2, ZSTD_greedy  },  /* level  5 */
+    { 21, 19, 19,  3,  5,  4, ZSTD_greedy  },  /* level  6 */
+    { 21, 19, 19,  3,  5,  8, ZSTD_lazy    },  /* level  7 */
+    { 21, 19, 19,  3,  5, 16, ZSTD_lazy2   },  /* level  8 */
+    { 21, 19, 20,  4,  5, 16, ZSTD_lazy2   },  /* level  9 */
+    { 22, 20, 21,  4,  5, 16, ZSTD_lazy2   },  /* level 10 */
+    { 22, 21, 22,  4,  5, 16, ZSTD_lazy2   },  /* level 11 */
+    { 22, 21, 22,  5,  5, 16, ZSTD_lazy2   },  /* level 12 */
+    { 22, 21, 22,  5,  5, 32, ZSTD_btlazy2 },  /* level 13 */
+    { 22, 22, 23,  5,  5, 32, ZSTD_btlazy2 },  /* level 14 */
+    { 22, 23, 23,  6,  5, 32, ZSTD_btlazy2 },  /* level 15 */
+    { 22, 22, 22,  5,  5, 48, ZSTD_btopt   },  /* level 16 */
+    { 23, 23, 22,  5,  4, 64, ZSTD_btopt   },  /* level 17 */
+    { 23, 23, 22,  6,  3, 64, ZSTD_btultra },  /* level 18 */
+    { 23, 24, 22,  7,  3,256, ZSTD_btultra2},  /* level 19 */
+    { 25, 25, 23,  7,  3,256, ZSTD_btultra2},  /* level 20 */
+    { 26, 26, 24,  7,  3,512, ZSTD_btultra2},  /* level 21 */
+    { 27, 27, 25,  9,  3,999, ZSTD_btultra2},  /* level 22 */
 },
 {   /* for srcSize <= 256 KB */
     /* W,  C,  H,  S,  L,  T, strat */
     { 18, 12, 13,  1,  5,  1, ZSTD_fast    },  /* base for negative levels */
-    { 18, 13, 14,  1,  6,  1, ZSTD_fast    },  /* level  1 */
-    { 18, 14, 13,  1,  5,  8, ZSTD_dfast   },  /* level  2 */
-    { 18, 16, 15,  1,  5,  8, ZSTD_dfast   },  /* level  3 */
-    { 18, 15, 17,  1,  5,  8, ZSTD_greedy  },  /* level  4.*/
-    { 18, 16, 17,  4,  5,  8, ZSTD_greedy  },  /* level  5.*/
-    { 18, 16, 17,  3,  5,  8, ZSTD_lazy    },  /* level  6.*/
-    { 18, 17, 17,  4,  4,  8, ZSTD_lazy    },  /* level  7 */
-    { 18, 17, 17,  4,  4,  8, ZSTD_lazy2   },  /* level  8 */
-    { 18, 17, 17,  5,  4,  8, ZSTD_lazy2   },  /* level  9 */
-    { 18, 17, 17,  6,  4,  8, ZSTD_lazy2   },  /* level 10 */
-    { 18, 18, 17,  6,  4,  8, ZSTD_lazy2   },  /* level 11.*/
-    { 18, 18, 17,  5,  4,  8, ZSTD_btlazy2 },  /* level 12.*/
-    { 18, 19, 17,  7,  4,  8, ZSTD_btlazy2 },  /* level 13 */
-    { 18, 18, 18,  4,  4, 16, ZSTD_btopt   },  /* level 14.*/
-    { 18, 18, 18,  4,  3, 16, ZSTD_btopt   },  /* level 15.*/
-    { 18, 19, 18,  6,  3, 32, ZSTD_btopt   },  /* level 16.*/
-    { 18, 19, 18,  8,  3, 64, ZSTD_btopt   },  /* level 17.*/
-    { 18, 19, 18,  9,  3,128, ZSTD_btopt   },  /* level 18.*/
-    { 18, 19, 18, 10,  3,256, ZSTD_btopt   },  /* level 19.*/
-    { 18, 19, 18, 11,  3,512, ZSTD_btultra },  /* level 20.*/
-    { 18, 19, 18, 12,  3,512, ZSTD_btultra },  /* level 21.*/
-    { 18, 19, 18, 13,  3,512, ZSTD_btultra },  /* level 22.*/
+    { 18, 13, 14,  1,  6,  0, ZSTD_fast    },  /* level  1 */
+    { 18, 14, 14,  1,  5,  1, ZSTD_dfast   },  /* level  2 */
+    { 18, 16, 16,  1,  4,  1, ZSTD_dfast   },  /* level  3 */
+    { 18, 16, 17,  2,  5,  2, ZSTD_greedy  },  /* level  4.*/
+    { 18, 18, 18,  3,  5,  2, ZSTD_greedy  },  /* level  5.*/
+    { 18, 18, 19,  3,  5,  4, ZSTD_lazy    },  /* level  6.*/
+    { 18, 18, 19,  4,  4,  4, ZSTD_lazy    },  /* level  7 */
+    { 18, 18, 19,  4,  4,  8, ZSTD_lazy2   },  /* level  8 */
+    { 18, 18, 19,  5,  4,  8, ZSTD_lazy2   },  /* level  9 */
+    { 18, 18, 19,  6,  4,  8, ZSTD_lazy2   },  /* level 10 */
+    { 18, 18, 19,  5,  4, 12, ZSTD_btlazy2 },  /* level 11.*/
+    { 18, 19, 19,  7,  4, 12, ZSTD_btlazy2 },  /* level 12.*/
+    { 18, 18, 19,  4,  4, 16, ZSTD_btopt   },  /* level 13 */
+    { 18, 18, 19,  4,  3, 32, ZSTD_btopt   },  /* level 14.*/
+    { 18, 18, 19,  6,  3,128, ZSTD_btopt   },  /* level 15.*/
+    { 18, 19, 19,  6,  3,128, ZSTD_btultra },  /* level 16.*/
+    { 18, 19, 19,  8,  3,256, ZSTD_btultra },  /* level 17.*/
+    { 18, 19, 19,  6,  3,128, ZSTD_btultra2},  /* level 18.*/
+    { 18, 19, 19,  8,  3,256, ZSTD_btultra2},  /* level 19.*/
+    { 18, 19, 19, 10,  3,512, ZSTD_btultra2},  /* level 20.*/
+    { 18, 19, 19, 12,  3,512, ZSTD_btultra2},  /* level 21.*/
+    { 18, 19, 19, 13,  3,999, ZSTD_btultra2},  /* level 22.*/
 },
 {   /* for srcSize <= 128 KB */
     /* W,  C,  H,  S,  L,  T, strat */
-    { 17, 12, 12,  1,  5,  1, ZSTD_fast    },  /* level  0 - not used */
-    { 17, 12, 13,  1,  6,  1, ZSTD_fast    },  /* level  1 */
-    { 17, 13, 16,  1,  5,  1, ZSTD_fast    },  /* level  2 */
-    { 17, 16, 16,  2,  5,  8, ZSTD_dfast   },  /* level  3 */
-    { 17, 13, 15,  3,  4,  8, ZSTD_greedy  },  /* level  4 */
-    { 17, 15, 17,  4,  4,  8, ZSTD_greedy  },  /* level  5 */
-    { 17, 16, 17,  3,  4,  8, ZSTD_lazy    },  /* level  6 */
-    { 17, 15, 17,  4,  4,  8, ZSTD_lazy2   },  /* level  7 */
+    { 17, 12, 12,  1,  5,  1, ZSTD_fast    },  /* base for negative levels */
+    { 17, 12, 13,  1,  6,  0, ZSTD_fast    },  /* level  1 */
+    { 17, 13, 15,  1,  5,  0, ZSTD_fast    },  /* level  2 */
+    { 17, 15, 16,  2,  5,  1, ZSTD_dfast   },  /* level  3 */
+    { 17, 17, 17,  2,  4,  1, ZSTD_dfast   },  /* level  4 */
+    { 17, 16, 17,  3,  4,  2, ZSTD_greedy  },  /* level  5 */
+    { 17, 17, 17,  3,  4,  4, ZSTD_lazy    },  /* level  6 */
+    { 17, 17, 17,  3,  4,  8, ZSTD_lazy2   },  /* level  7 */
     { 17, 17, 17,  4,  4,  8, ZSTD_lazy2   },  /* level  8 */
     { 17, 17, 17,  5,  4,  8, ZSTD_lazy2   },  /* level  9 */
     { 17, 17, 17,  6,  4,  8, ZSTD_lazy2   },  /* level 10 */
-    { 17, 17, 17,  7,  4,  8, ZSTD_lazy2   },  /* level 11 */
-    { 17, 17, 17,  8,  4,  8, ZSTD_lazy2   },  /* level 12 */
-    { 17, 18, 17,  6,  4,  8, ZSTD_btlazy2 },  /* level 13.*/
-    { 17, 17, 17,  7,  3,  8, ZSTD_btopt   },  /* level 14.*/
-    { 17, 17, 17,  7,  3, 16, ZSTD_btopt   },  /* level 15.*/
-    { 17, 18, 17,  7,  3, 32, ZSTD_btopt   },  /* level 16.*/
-    { 17, 18, 17,  7,  3, 64, ZSTD_btopt   },  /* level 17.*/
-    { 17, 18, 17,  7,  3,256, ZSTD_btopt   },  /* level 18.*/
-    { 17, 18, 17,  8,  3,256, ZSTD_btopt   },  /* level 19.*/
-    { 17, 18, 17,  9,  3,256, ZSTD_btultra },  /* level 20.*/
-    { 17, 18, 17, 10,  3,256, ZSTD_btultra },  /* level 21.*/
-    { 17, 18, 17, 11,  3,512, ZSTD_btultra },  /* level 22.*/
+    { 17, 17, 17,  5,  4,  8, ZSTD_btlazy2 },  /* level 11 */
+    { 17, 18, 17,  7,  4, 12, ZSTD_btlazy2 },  /* level 12 */
+    { 17, 18, 17,  3,  4, 12, ZSTD_btopt   },  /* level 13.*/
+    { 17, 18, 17,  4,  3, 32, ZSTD_btopt   },  /* level 14.*/
+    { 17, 18, 17,  6,  3,256, ZSTD_btopt   },  /* level 15.*/
+    { 17, 18, 17,  6,  3,128, ZSTD_btultra },  /* level 16.*/
+    { 17, 18, 17,  8,  3,256, ZSTD_btultra },  /* level 17.*/
+    { 17, 18, 17, 10,  3,512, ZSTD_btultra },  /* level 18.*/
+    { 17, 18, 17,  5,  3,256, ZSTD_btultra2},  /* level 19.*/
+    { 17, 18, 17,  7,  3,512, ZSTD_btultra2},  /* level 20.*/
+    { 17, 18, 17,  9,  3,512, ZSTD_btultra2},  /* level 21.*/
+    { 17, 18, 17, 11,  3,999, ZSTD_btultra2},  /* level 22.*/
 },
 {   /* for srcSize <= 16 KB */
     /* W,  C,  H,  S,  L,  T, strat */
     { 14, 12, 13,  1,  5,  1, ZSTD_fast    },  /* base for negative levels */
-    { 14, 14, 14,  1,  6,  1, ZSTD_fast    },  /* level  1 */
-    { 14, 14, 14,  1,  4,  1, ZSTD_fast    },  /* level  2 */
-    { 14, 14, 14,  1,  4,  6, ZSTD_dfast   },  /* level  3.*/
-    { 14, 14, 14,  4,  4,  6, ZSTD_greedy  },  /* level  4.*/
-    { 14, 14, 14,  3,  4,  6, ZSTD_lazy    },  /* level  5.*/
-    { 14, 14, 14,  4,  4,  6, ZSTD_lazy2   },  /* level  6 */
-    { 14, 14, 14,  5,  4,  6, ZSTD_lazy2   },  /* level  7 */
-    { 14, 14, 14,  6,  4,  6, ZSTD_lazy2   },  /* level  8.*/
-    { 14, 15, 14,  6,  4,  6, ZSTD_btlazy2 },  /* level  9.*/
-    { 14, 15, 14,  3,  3,  6, ZSTD_btopt   },  /* level 10.*/
-    { 14, 15, 14,  6,  3,  8, ZSTD_btopt   },  /* level 11.*/
-    { 14, 15, 14,  6,  3, 16, ZSTD_btopt   },  /* level 12.*/
-    { 14, 15, 14,  6,  3, 24, ZSTD_btopt   },  /* level 13.*/
-    { 14, 15, 15,  6,  3, 48, ZSTD_btopt   },  /* level 14.*/
-    { 14, 15, 15,  6,  3, 64, ZSTD_btopt   },  /* level 15.*/
-    { 14, 15, 15,  6,  3, 96, ZSTD_btopt   },  /* level 16.*/
-    { 14, 15, 15,  6,  3,128, ZSTD_btopt   },  /* level 17.*/
-    { 14, 15, 15,  6,  3,256, ZSTD_btopt   },  /* level 18.*/
-    { 14, 15, 15,  7,  3,256, ZSTD_btopt   },  /* level 19.*/
-    { 14, 15, 15,  8,  3,256, ZSTD_btultra },  /* level 20.*/
-    { 14, 15, 15,  9,  3,256, ZSTD_btultra },  /* level 21.*/
-    { 14, 15, 15, 10,  3,256, ZSTD_btultra },  /* level 22.*/
+    { 14, 14, 15,  1,  5,  0, ZSTD_fast    },  /* level  1 */
+    { 14, 14, 15,  1,  4,  0, ZSTD_fast    },  /* level  2 */
+    { 14, 14, 15,  2,  4,  1, ZSTD_dfast   },  /* level  3 */
+    { 14, 14, 14,  4,  4,  2, ZSTD_greedy  },  /* level  4 */
+    { 14, 14, 14,  3,  4,  4, ZSTD_lazy    },  /* level  5.*/
+    { 14, 14, 14,  4,  4,  8, ZSTD_lazy2   },  /* level  6 */
+    { 14, 14, 14,  6,  4,  8, ZSTD_lazy2   },  /* level  7 */
+    { 14, 14, 14,  8,  4,  8, ZSTD_lazy2   },  /* level  8.*/
+    { 14, 15, 14,  5,  4,  8, ZSTD_btlazy2 },  /* level  9.*/
+    { 14, 15, 14,  9,  4,  8, ZSTD_btlazy2 },  /* level 10.*/
+    { 14, 15, 14,  3,  4, 12, ZSTD_btopt   },  /* level 11.*/
+    { 14, 15, 14,  4,  3, 24, ZSTD_btopt   },  /* level 12.*/
+    { 14, 15, 14,  5,  3, 32, ZSTD_btultra },  /* level 13.*/
+    { 14, 15, 15,  6,  3, 64, ZSTD_btultra },  /* level 14.*/
+    { 14, 15, 15,  7,  3,256, ZSTD_btultra },  /* level 15.*/
+    { 14, 15, 15,  5,  3, 48, ZSTD_btultra2},  /* level 16.*/
+    { 14, 15, 15,  6,  3,128, ZSTD_btultra2},  /* level 17.*/
+    { 14, 15, 15,  7,  3,256, ZSTD_btultra2},  /* level 18.*/
+    { 14, 15, 15,  8,  3,256, ZSTD_btultra2},  /* level 19.*/
+    { 14, 15, 15,  8,  3,512, ZSTD_btultra2},  /* level 20.*/
+    { 14, 15, 15,  9,  3,512, ZSTD_btultra2},  /* level 21.*/
+    { 14, 15, 15, 10,  3,999, ZSTD_btultra2},  /* level 22.*/
 },
 };
 
 /*! ZSTD_getCParams() :
-*  @return ZSTD_compressionParameters structure for a selected compression level, srcSize and dictSize.
-*   Size values are optional, provide 0 if not known or unused */
+ * @return ZSTD_compressionParameters structure for a selected compression level, srcSize and dictSize.
+ *  Size values are optional, provide 0 if not known or unused */
 ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize)
 {
     size_t const addedSize = srcSizeHint ? 0 : 500;
-    U64 const rSize = srcSizeHint+dictSize ? srcSizeHint+dictSize+addedSize : (U64)-1;
-    U32 const tableID = (rSize <= 256 KB) + (rSize <= 128 KB) + (rSize <= 16 KB);   /* intentional underflow for srcSizeHint == 0 */
+    U64 const rSize = srcSizeHint+dictSize ? srcSizeHint+dictSize+addedSize : ZSTD_CONTENTSIZE_UNKNOWN;  /* intentional overflow for srcSizeHint == ZSTD_CONTENTSIZE_UNKNOWN */
+    U32 const tableID = (rSize <= 256 KB) + (rSize <= 128 KB) + (rSize <= 16 KB);
     int row = compressionLevel;
     DEBUGLOG(5, "ZSTD_getCParams (cLevel=%i)", compressionLevel);
     if (compressionLevel == 0) row = ZSTD_CLEVEL_DEFAULT;   /* 0 == default */
@@ -3431,13 +4456,14 @@
     if (compressionLevel > ZSTD_MAX_CLEVEL) row = ZSTD_MAX_CLEVEL;
     {   ZSTD_compressionParameters cp = ZSTD_defaultCParameters[tableID][row];
         if (compressionLevel < 0) cp.targetLength = (unsigned)(-compressionLevel);   /* acceleration factor */
-        return ZSTD_adjustCParams_internal(cp, srcSizeHint, dictSize); }
-
+        return ZSTD_adjustCParams_internal(cp, srcSizeHint, dictSize);               /* refine parameters based on srcSize & dictSize */
+    }
 }
 
 /*! ZSTD_getParams() :
-*   same as ZSTD_getCParams(), but @return a `ZSTD_parameters` object (instead of `ZSTD_compressionParameters`).
-*   All fields of `ZSTD_frameParameters` are set to default (0) */
+ *  same idea as ZSTD_getCParams()
+ * @return a `ZSTD_parameters` structure (instead of `ZSTD_compressionParameters`).
+ *  Fields of `ZSTD_frameParameters` are set to default values */
 ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize) {
     ZSTD_parameters params;
     ZSTD_compressionParameters const cParams = ZSTD_getCParams(compressionLevel, srcSizeHint, dictSize);