[pcsx_rearmed.git] / deps / libchdr / deps / zstd-1.5.5 / lib / common / fse_decompress.c

/* ******************************************************************
 * FSE : Finite State Entropy decoder
 * Copyright (c) Meta Platforms, Inc. and affiliates.
 *
 *  You can contact the author at :
 *  - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
 *  - Public forum : https://groups.google.com/forum/#!forum/lz4c
 *
 * This source code is licensed under both the BSD-style license (found in the
 * LICENSE file in the root directory of this source tree) and the GPLv2 (found
 * in the COPYING file in the root directory of this source tree).
 * You may select, at your option, one of the above-listed licenses.
****************************************************************** */


/* **************************************************************
*  Includes
****************************************************************/
#include "debug.h"      /* assert */
#include "bitstream.h"
#include "compiler.h"
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#include "error_private.h"
#define ZSTD_DEPS_NEED_MALLOC
#include "zstd_deps.h"
#include "bits.h"       /* ZSTD_highbit32 */


/* **************************************************************
*  Error Management
****************************************************************/
#define FSE_isError ERR_isError
#define FSE_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c)   /* use only *after* variable declarations */


/* **************************************************************
*  Templates
****************************************************************/
/*
  designed to be included
  for type-specific functions (template emulation in C)
  Objective is to write these functions only once, for improved maintenance
*/

/* safety checks */
#ifndef FSE_FUNCTION_EXTENSION
#  error "FSE_FUNCTION_EXTENSION must be defined"
#endif
#ifndef FSE_FUNCTION_TYPE
#  error "FSE_FUNCTION_TYPE must be defined"
#endif

/* Function names */
#define FSE_CAT(X,Y) X##Y
#define FSE_FUNCTION_NAME(X,Y) FSE_CAT(X,Y)
#define FSE_TYPE_NAME(X,Y) FSE_CAT(X,Y)

static size_t FSE_buildDTable_internal(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
{
    void* const tdPtr = dt+1;   /* because *dt is unsigned, 32-bits aligned on 32-bits */
    FSE_DECODE_TYPE* const tableDecode = (FSE_DECODE_TYPE*) (tdPtr);
    U16* symbolNext = (U16*)workSpace;
    BYTE* spread = (BYTE*)(symbolNext + maxSymbolValue + 1);

    U32 const maxSV1 = maxSymbolValue + 1;
    U32 const tableSize = 1 << tableLog;
    U32 highThreshold = tableSize-1;

    /* Sanity Checks */
    if (FSE_BUILD_DTABLE_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(maxSymbolValue_tooLarge);
    if (maxSymbolValue > FSE_MAX_SYMBOL_VALUE) return ERROR(maxSymbolValue_tooLarge);
    if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);

    /* Init, lay down lowprob symbols */
    {   FSE_DTableHeader DTableH;
        DTableH.tableLog = (U16)tableLog;
        DTableH.fastMode = 1;
        {   S16 const largeLimit= (S16)(1 << (tableLog-1));
            U32 s;
            for (s=0; s<maxSV1; s++) {
                if (normalizedCounter[s]==-1) {
                    tableDecode[highThreshold--].symbol = (FSE_FUNCTION_TYPE)s;
                    symbolNext[s] = 1;
                } else {
                    if (normalizedCounter[s] >= largeLimit) DTableH.fastMode=0;
                    symbolNext[s] = normalizedCounter[s];
        }   }   }
        ZSTD_memcpy(dt, &DTableH, sizeof(DTableH));
    }

    /* Spread symbols */
    if (highThreshold == tableSize - 1) {
        size_t const tableMask = tableSize-1;
        size_t const step = FSE_TABLESTEP(tableSize);
        /* First lay down the symbols in order.
         * We use a uint64_t to lay down 8 bytes at a time. This reduces branch
         * misses since small blocks generally have small table logs, so nearly
         * all symbols have counts <= 8. We ensure we have 8 bytes at the end of
         * our buffer to handle the over-write.
         */
        {
            U64 const add = 0x0101010101010101ull;
            size_t pos = 0;
            U64 sv = 0;
            U32 s;
            for (s=0; s<maxSV1; ++s, sv += add) {
                int i;
                int const n = normalizedCounter[s];
                MEM_write64(spread + pos, sv);
                for (i = 8; i < n; i += 8) {
                    MEM_write64(spread + pos + i, sv);
                }
                pos += n;
            }
        }
        /* Now we spread those positions across the table.
         * The benefit of doing it in two stages is that we avoid the
         * variable size inner loop, which caused lots of branch misses.
         * Now we can run through all the positions without any branch misses.
         * We unroll the loop twice, since that is what empirically worked best.
         */
        {
            size_t position = 0;
            size_t s;
            size_t const unroll = 2;
            assert(tableSize % unroll == 0); /* FSE_MIN_TABLELOG is 5 */
            for (s = 0; s < (size_t)tableSize; s += unroll) {
                size_t u;
                for (u = 0; u < unroll; ++u) {
                    size_t const uPosition = (position + (u * step)) & tableMask;
                    tableDecode[uPosition].symbol = spread[s + u];
                }
                position = (position + (unroll * step)) & tableMask;
            }
            assert(position == 0);
        }
    } else {
        U32 const tableMask = tableSize-1;
        U32 const step = FSE_TABLESTEP(tableSize);
        U32 s, position = 0;
        for (s=0; s<maxSV1; s++) {
            int i;
            for (i=0; i<normalizedCounter[s]; i++) {
                tableDecode[position].symbol = (FSE_FUNCTION_TYPE)s;
                position = (position + step) & tableMask;
                while (position > highThreshold) position = (position + step) & tableMask;   /* lowprob area */
        }   }
        if (position!=0) return ERROR(GENERIC);   /* position must reach all cells once, otherwise normalizedCounter is incorrect */
    }

    /* Build Decoding table */
    {   U32 u;
        for (u=0; u<tableSize; u++) {
            FSE_FUNCTION_TYPE const symbol = (FSE_FUNCTION_TYPE)(tableDecode[u].symbol);
            U32 const nextState = symbolNext[symbol]++;
            tableDecode[u].nbBits = (BYTE) (tableLog - ZSTD_highbit32(nextState) );
            tableDecode[u].newState = (U16) ( (nextState << tableDecode[u].nbBits) - tableSize);
    }   }

    return 0;
}

size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
{
    return FSE_buildDTable_internal(dt, normalizedCounter, maxSymbolValue, tableLog, workSpace, wkspSize);
}


#ifndef FSE_COMMONDEFS_ONLY

/*-*******************************************************
*  Decompression (Byte symbols)
*********************************************************/

FORCE_INLINE_TEMPLATE size_t FSE_decompress_usingDTable_generic(
          void* dst, size_t maxDstSize,
    const void* cSrc, size_t cSrcSize,
    const FSE_DTable* dt, const unsigned fast)
{
    BYTE* const ostart = (BYTE*) dst;
    BYTE* op = ostart;
    BYTE* const omax = op + maxDstSize;
    BYTE* const olimit = omax-3;

    BIT_DStream_t bitD;
    FSE_DState_t state1;
    FSE_DState_t state2;

    /* Init */
    CHECK_F(BIT_initDStream(&bitD, cSrc, cSrcSize));

    FSE_initDState(&state1, &bitD, dt);
    FSE_initDState(&state2, &bitD, dt);

#define FSE_GETSYMBOL(statePtr) fast ? FSE_decodeSymbolFast(statePtr, &bitD) : FSE_decodeSymbol(statePtr, &bitD)

    /* 4 symbols per loop */
    for ( ; (BIT_reloadDStream(&bitD)==BIT_DStream_unfinished) & (op<olimit) ; op+=4) {
        op[0] = FSE_GETSYMBOL(&state1);

        if (FSE_MAX_TABLELOG*2+7 > sizeof(bitD.bitContainer)*8)    /* This test must be static */
            BIT_reloadDStream(&bitD);

        op[1] = FSE_GETSYMBOL(&state2);

        if (FSE_MAX_TABLELOG*4+7 > sizeof(bitD.bitContainer)*8)    /* This test must be static */
            { if (BIT_reloadDStream(&bitD) > BIT_DStream_unfinished) { op+=2; break; } }

        op[2] = FSE_GETSYMBOL(&state1);

        if (FSE_MAX_TABLELOG*2+7 > sizeof(bitD.bitContainer)*8)    /* This test must be static */
            BIT_reloadDStream(&bitD);

        op[3] = FSE_GETSYMBOL(&state2);
    }

    /* tail */
    /* note : BIT_reloadDStream(&bitD) >= FSE_DStream_partiallyFilled; Ends at exactly BIT_DStream_completed */
    while (1) {
        if (op>(omax-2)) return ERROR(dstSize_tooSmall);
        *op++ = FSE_GETSYMBOL(&state1);
        if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) {
            *op++ = FSE_GETSYMBOL(&state2);
            break;
        }

        if (op>(omax-2)) return ERROR(dstSize_tooSmall);
        *op++ = FSE_GETSYMBOL(&state2);
        if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) {
            *op++ = FSE_GETSYMBOL(&state1);
            break;
    }   }

    return op-ostart;
}

typedef struct {
    short ncount[FSE_MAX_SYMBOL_VALUE + 1];
    FSE_DTable dtable[1]; /* Dynamically sized */
} FSE_DecompressWksp;


FORCE_INLINE_TEMPLATE size_t FSE_decompress_wksp_body(
        void* dst, size_t dstCapacity,
        const void* cSrc, size_t cSrcSize,
        unsigned maxLog, void* workSpace, size_t wkspSize,
        int bmi2)
{
    const BYTE* const istart = (const BYTE*)cSrc;
    const BYTE* ip = istart;
    unsigned tableLog;
    unsigned maxSymbolValue = FSE_MAX_SYMBOL_VALUE;
    FSE_DecompressWksp* const wksp = (FSE_DecompressWksp*)workSpace;

    DEBUG_STATIC_ASSERT((FSE_MAX_SYMBOL_VALUE + 1) % 2 == 0);
    if (wkspSize < sizeof(*wksp)) return ERROR(GENERIC);

    /* normal FSE decoding mode */
    {
        size_t const NCountLength = FSE_readNCount_bmi2(wksp->ncount, &maxSymbolValue, &tableLog, istart, cSrcSize, bmi2);
        if (FSE_isError(NCountLength)) return NCountLength;
        if (tableLog > maxLog) return ERROR(tableLog_tooLarge);
        assert(NCountLength <= cSrcSize);
        ip += NCountLength;
        cSrcSize -= NCountLength;
    }

    if (FSE_DECOMPRESS_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(tableLog_tooLarge);
    assert(sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog) <= wkspSize);
    workSpace = (BYTE*)workSpace + sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog);
    wkspSize -= sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog);

    CHECK_F( FSE_buildDTable_internal(wksp->dtable, wksp->ncount, maxSymbolValue, tableLog, workSpace, wkspSize) );

    {
        const void* ptr = wksp->dtable;
        const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr;
        const U32 fastMode = DTableH->fastMode;

        /* select fast mode (static) */
        if (fastMode) return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 1);
        return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 0);
    }
}

/* Avoids the FORCE_INLINE of the _body() function. */
static size_t FSE_decompress_wksp_body_default(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
{
    return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 0);
}

#if DYNAMIC_BMI2
BMI2_TARGET_ATTRIBUTE static size_t FSE_decompress_wksp_body_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
{
    return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 1);
}
#endif

size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2)
{
#if DYNAMIC_BMI2
    if (bmi2) {
        return FSE_decompress_wksp_body_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
    }
#endif
    (void)bmi2;
    return FSE_decompress_wksp_body_default(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
}

#endif   /* FSE_COMMONDEFS_ONLY */
Commit	Line	Data
648db22b	1	/* ******************************************************************
	2	* FSE : Finite State Entropy decoder
	3	* Copyright (c) Meta Platforms, Inc. and affiliates.
	4	*
	5	* You can contact the author at :
	6	* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
	7	* - Public forum : https://groups.google.com/forum/#!forum/lz4c
	8	*
	9	* This source code is licensed under both the BSD-style license (found in the
	10	* LICENSE file in the root directory of this source tree) and the GPLv2 (found
	11	* in the COPYING file in the root directory of this source tree).
	12	* You may select, at your option, one of the above-listed licenses.
	13	****************************************************************** */
	14
	15
	16	/* **************************************************************
	17	* Includes
	18	****************************************************************/
	19	#include "debug.h" /* assert */
	20	#include "bitstream.h"
	21	#include "compiler.h"
	22	#define FSE_STATIC_LINKING_ONLY
	23	#include "fse.h"
	24	#include "error_private.h"
	25	#define ZSTD_DEPS_NEED_MALLOC
	26	#include "zstd_deps.h"
	27	#include "bits.h" /* ZSTD_highbit32 */
	28
	29
	30	/* **************************************************************
	31	* Error Management
	32	****************************************************************/
	33	#define FSE_isError ERR_isError
	34	#define FSE_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only after variable declarations */
	35
	36
	37	/* **************************************************************
	38	* Templates
	39	****************************************************************/
	40	/*
	41	designed to be included
	42	for type-specific functions (template emulation in C)
	43	Objective is to write these functions only once, for improved maintenance
	44	*/
	45
	46	/* safety checks */
	47	#ifndef FSE_FUNCTION_EXTENSION
	48	# error "FSE_FUNCTION_EXTENSION must be defined"
	49	#endif
	50	#ifndef FSE_FUNCTION_TYPE
	51	# error "FSE_FUNCTION_TYPE must be defined"
	52	#endif
	53
	54	/* Function names */
	55	#define FSE_CAT(X,Y) X##Y
	56	#define FSE_FUNCTION_NAME(X,Y) FSE_CAT(X,Y)
	57	#define FSE_TYPE_NAME(X,Y) FSE_CAT(X,Y)
	58
	59	static size_t FSE_buildDTable_internal(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
	60	{
	61	void* const tdPtr = dt+1; /* because dt is unsigned, 32-bits aligned on 32-bits /
	62	FSE_DECODE_TYPE* const tableDecode = (FSE_DECODE_TYPE*) (tdPtr);
	63	U16* symbolNext = (U16*)workSpace;
	64	BYTE* spread = (BYTE*)(symbolNext + maxSymbolValue + 1);
65
66	U32 const maxSV1 = maxSymbolValue + 1;
67	U32 const tableSize = 1 << tableLog;
68	U32 highThreshold = tableSize-1;
69
70	/* Sanity Checks */
71	if (FSE_BUILD_DTABLE_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(maxSymbolValue_tooLarge);
72	if (maxSymbolValue > FSE_MAX_SYMBOL_VALUE) return ERROR(maxSymbolValue_tooLarge);
73	if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
74
75	/* Init, lay down lowprob symbols */
76	{ FSE_DTableHeader DTableH;
77	DTableH.tableLog = (U16)tableLog;
78	DTableH.fastMode = 1;
79	{ S16 const largeLimit= (S16)(1 << (tableLog-1));
80	U32 s;
81	for (s=0; s<maxSV1; s++) {
82	if (normalizedCounter[s]==-1) {
83	tableDecode[highThreshold--].symbol = (FSE_FUNCTION_TYPE)s;
84	symbolNext[s] = 1;
85	} else {
86	if (normalizedCounter[s] >= largeLimit) DTableH.fastMode=0;
87	symbolNext[s] = normalizedCounter[s];
88	} } }
89	ZSTD_memcpy(dt, &DTableH, sizeof(DTableH));
90	}
91
92	/* Spread symbols */
93	if (highThreshold == tableSize - 1) {
94	size_t const tableMask = tableSize-1;
95	size_t const step = FSE_TABLESTEP(tableSize);
96	/* First lay down the symbols in order.
97	* We use a uint64_t to lay down 8 bytes at a time. This reduces branch
98	* misses since small blocks generally have small table logs, so nearly
99	* all symbols have counts <= 8. We ensure we have 8 bytes at the end of
100	* our buffer to handle the over-write.
101	*/
102	{
103	U64 const add = 0x0101010101010101ull;
104	size_t pos = 0;
105	U64 sv = 0;
106	U32 s;
107	for (s=0; s<maxSV1; ++s, sv += add) {
108	int i;
109	int const n = normalizedCounter[s];
110	MEM_write64(spread + pos, sv);
111	for (i = 8; i < n; i += 8) {
112	MEM_write64(spread + pos + i, sv);
113	}
114	pos += n;
115	}
116	}
117	/* Now we spread those positions across the table.
118	* The benefit of doing it in two stages is that we avoid the
119	* variable size inner loop, which caused lots of branch misses.
120	* Now we can run through all the positions without any branch misses.
121	* We unroll the loop twice, since that is what empirically worked best.
122	*/
123	{
124	size_t position = 0;
125	size_t s;
126	size_t const unroll = 2;
127	assert(tableSize % unroll == 0); /* FSE_MIN_TABLELOG is 5 */
128	for (s = 0; s < (size_t)tableSize; s += unroll) {
129	size_t u;
130	for (u = 0; u < unroll; ++u) {
131	size_t const uPosition = (position + (u * step)) & tableMask;
132	tableDecode[uPosition].symbol = spread[s + u];
133	}
134	position = (position + (unroll * step)) & tableMask;
135	}
136	assert(position == 0);
137	}
138	} else {
139	U32 const tableMask = tableSize-1;
140	U32 const step = FSE_TABLESTEP(tableSize);
141	U32 s, position = 0;
142	for (s=0; s<maxSV1; s++) {
143	int i;
144	for (i=0; i<normalizedCounter[s]; i++) {
145	tableDecode[position].symbol = (FSE_FUNCTION_TYPE)s;
146	position = (position + step) & tableMask;
147	while (position > highThreshold) position = (position + step) & tableMask; /* lowprob area */
148	} }
149	if (position!=0) return ERROR(GENERIC); /* position must reach all cells once, otherwise normalizedCounter is incorrect */
150	}
151
152	/* Build Decoding table */
153	{ U32 u;
154	for (u=0; u<tableSize; u++) {
155	FSE_FUNCTION_TYPE const symbol = (FSE_FUNCTION_TYPE)(tableDecode[u].symbol);
156	U32 const nextState = symbolNext[symbol]++;
157	tableDecode[u].nbBits = (BYTE) (tableLog - ZSTD_highbit32(nextState) );
158	tableDecode[u].newState = (U16) ( (nextState << tableDecode[u].nbBits) - tableSize);
159	} }
160
161	return 0;
162	}
163
164	size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
165	{
166	return FSE_buildDTable_internal(dt, normalizedCounter, maxSymbolValue, tableLog, workSpace, wkspSize);
167	}
168
169
170	#ifndef FSE_COMMONDEFS_ONLY
171
172	/-******************************************************
173	* Decompression (Byte symbols)
174	*********************************************************/
175
176	FORCE_INLINE_TEMPLATE size_t FSE_decompress_usingDTable_generic(
177	void* dst, size_t maxDstSize,
178	const void* cSrc, size_t cSrcSize,
179	const FSE_DTable* dt, const unsigned fast)
180	{
181	BYTE* const ostart = (BYTE*) dst;
182	BYTE* op = ostart;
183	BYTE* const omax = op + maxDstSize;
184	BYTE* const olimit = omax-3;
185
186	BIT_DStream_t bitD;
187	FSE_DState_t state1;
188	FSE_DState_t state2;
189
190	/* Init */
191	CHECK_F(BIT_initDStream(&bitD, cSrc, cSrcSize));
192
193	FSE_initDState(&state1, &bitD, dt);
194	FSE_initDState(&state2, &bitD, dt);
195
196	#define FSE_GETSYMBOL(statePtr) fast ? FSE_decodeSymbolFast(statePtr, &bitD) : FSE_decodeSymbol(statePtr, &bitD)
197
198	/* 4 symbols per loop */
199	for ( ; (BIT_reloadDStream(&bitD)==BIT_DStream_unfinished) & (op<olimit) ; op+=4) {
200	op[0] = FSE_GETSYMBOL(&state1);
201
202	if (FSE_MAX_TABLELOG2+7 > sizeof(bitD.bitContainer)8) /* This test must be static */
203	BIT_reloadDStream(&bitD);
204
205	op[1] = FSE_GETSYMBOL(&state2);
206
207	if (FSE_MAX_TABLELOG4+7 > sizeof(bitD.bitContainer)8) /* This test must be static */
208	{ if (BIT_reloadDStream(&bitD) > BIT_DStream_unfinished) { op+=2; break; } }
209
210	op[2] = FSE_GETSYMBOL(&state1);
211
212	if (FSE_MAX_TABLELOG2+7 > sizeof(bitD.bitContainer)8) /* This test must be static */
213	BIT_reloadDStream(&bitD);
214
215	op[3] = FSE_GETSYMBOL(&state2);
216	}
217
218	/* tail */
219	/* note : BIT_reloadDStream(&bitD) >= FSE_DStream_partiallyFilled; Ends at exactly BIT_DStream_completed */
220	while (1) {
221	if (op>(omax-2)) return ERROR(dstSize_tooSmall);
222	*op++ = FSE_GETSYMBOL(&state1);
223	if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) {
224	*op++ = FSE_GETSYMBOL(&state2);
225	break;
226	}
227
228	if (op>(omax-2)) return ERROR(dstSize_tooSmall);
229	*op++ = FSE_GETSYMBOL(&state2);
230	if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) {
231	*op++ = FSE_GETSYMBOL(&state1);
232	break;
233	} }
234
235	return op-ostart;
236	}
237
238	typedef struct {
239	short ncount[FSE_MAX_SYMBOL_VALUE + 1];
240	FSE_DTable dtable[1]; /* Dynamically sized */
241	} FSE_DecompressWksp;
242
243
244	FORCE_INLINE_TEMPLATE size_t FSE_decompress_wksp_body(
245	void* dst, size_t dstCapacity,
246	const void* cSrc, size_t cSrcSize,
247	unsigned maxLog, void* workSpace, size_t wkspSize,
248	int bmi2)
249	{
250	const BYTE* const istart = (const BYTE*)cSrc;
251	const BYTE* ip = istart;
252	unsigned tableLog;
253	unsigned maxSymbolValue = FSE_MAX_SYMBOL_VALUE;
254	FSE_DecompressWksp* const wksp = (FSE_DecompressWksp*)workSpace;
255
256	DEBUG_STATIC_ASSERT((FSE_MAX_SYMBOL_VALUE + 1) % 2 == 0);
257	if (wkspSize < sizeof(*wksp)) return ERROR(GENERIC);
258
259	/* normal FSE decoding mode */
260	{
261	size_t const NCountLength = FSE_readNCount_bmi2(wksp->ncount, &maxSymbolValue, &tableLog, istart, cSrcSize, bmi2);
262	if (FSE_isError(NCountLength)) return NCountLength;
263	if (tableLog > maxLog) return ERROR(tableLog_tooLarge);
264	assert(NCountLength <= cSrcSize);
265	ip += NCountLength;
266	cSrcSize -= NCountLength;
267	}
268
269	if (FSE_DECOMPRESS_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(tableLog_tooLarge);
270	assert(sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog) <= wkspSize);
271	workSpace = (BYTE)workSpace + sizeof(wksp) + FSE_DTABLE_SIZE(tableLog);
272	wkspSize -= sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog);
273
274	CHECK_F( FSE_buildDTable_internal(wksp->dtable, wksp->ncount, maxSymbolValue, tableLog, workSpace, wkspSize) );
275
276	{
277	const void* ptr = wksp->dtable;
278	const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr;
279	const U32 fastMode = DTableH->fastMode;
280
281	/* select fast mode (static) */
282	if (fastMode) return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 1);
283	return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 0);
284	}
285	}
286
287	/* Avoids the FORCE_INLINE of the _body() function. */
288	static size_t FSE_decompress_wksp_body_default(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
289	{
290	return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 0);
291	}
292
293	#if DYNAMIC_BMI2
294	BMI2_TARGET_ATTRIBUTE static size_t FSE_decompress_wksp_body_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
295	{
296	return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 1);
297	}
298	#endif
299
300	size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2)
301	{
302	#if DYNAMIC_BMI2
303	if (bmi2) {
304	return FSE_decompress_wksp_body_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
305	}
306	#endif
307	(void)bmi2;
308	return FSE_decompress_wksp_body_default(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
309	}
310
311	#endif /* FSE_COMMONDEFS_ONLY */