wire up bmi2 support
This commit is contained in:
@@ -38,8 +38,9 @@ const char* HUF_getErrorName(size_t code) { return ERR_getErrorName(code); }
|
||||
/*-**************************************************************
|
||||
* FSE NCount encoding-decoding
|
||||
****************************************************************/
|
||||
size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize)
|
||||
FORCE_INLINE_TEMPLATE
|
||||
size_t FSE_readNCount_body(short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize)
|
||||
{
|
||||
const BYTE* const istart = (const BYTE*) headerBuffer;
|
||||
const BYTE* const iend = istart + hbSize;
|
||||
@@ -175,6 +176,43 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
|
||||
return ip-istart;
|
||||
}
|
||||
|
||||
static size_t FSE_readNCount_body_default(
|
||||
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize)
|
||||
{
|
||||
return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
|
||||
}
|
||||
|
||||
#if DYNAMIC_BMI2
|
||||
TARGET_ATTRIBUTE("bmi2") static size_t FSE_readNCount_body_bmi2(
|
||||
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize)
|
||||
{
|
||||
return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
|
||||
}
|
||||
#endif
|
||||
|
||||
size_t FSE_readNCount_bmi2(
|
||||
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize, int bmi2)
|
||||
{
|
||||
#if DYNAMIC_BMI2
|
||||
if (bmi2) {
|
||||
return FSE_readNCount_body_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
|
||||
}
|
||||
#endif
|
||||
(void)bmi2;
|
||||
return FSE_readNCount_body_default(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
|
||||
}
|
||||
|
||||
size_t FSE_readNCount(
|
||||
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
|
||||
const void* headerBuffer, size_t hbSize)
|
||||
{
|
||||
return FSE_readNCount_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize, /* bmi2 */ 0);
|
||||
}
|
||||
|
||||
|
||||
/*! HUF_readStats() :
|
||||
Read compact Huffman tree, saved by HUF_writeCTable().
|
||||
`huffWeight` is destination buffer.
|
||||
@@ -187,13 +225,14 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
const void* src, size_t srcSize)
|
||||
{
|
||||
U32 wksp[HUF_READ_STATS_WORKSPACE_SIZE_U32];
|
||||
return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp));
|
||||
return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp), /* bmi2 */ 0);
|
||||
}
|
||||
|
||||
size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
FORCE_INLINE_TEMPLATE size_t HUF_readStats_body(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
U32* nbSymbolsPtr, U32* tableLogPtr,
|
||||
const void* src, size_t srcSize,
|
||||
void* workSpace, size_t wkspSize)
|
||||
void* workSpace, size_t wkspSize,
|
||||
int bmi2)
|
||||
{
|
||||
U32 weightTotal;
|
||||
const BYTE* ip = (const BYTE*) src;
|
||||
@@ -217,7 +256,7 @@ size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
} } }
|
||||
else { /* header compressed with FSE (normal case) */
|
||||
if (iSize+1 > srcSize) return ERROR(srcSize_wrong);
|
||||
oSize = FSE_decompress_wksp(huffWeight, hwSize-1, ip+1, iSize, 6, workSpace, wkspSize); /* max (hwSize-1) values decoded, as last one is implied */
|
||||
oSize = FSE_decompress_wksp_bmi2(huffWeight, hwSize-1, ip+1, iSize, 6, workSpace, wkspSize, bmi2); /* max (hwSize-1) values decoded, as last one is implied */
|
||||
if (FSE_isError(oSize)) return oSize;
|
||||
}
|
||||
|
||||
@@ -252,3 +291,36 @@ size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
*nbSymbolsPtr = (U32)(oSize+1);
|
||||
return iSize+1;
|
||||
}
|
||||
|
||||
static size_t HUF_readStats_body_default(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
U32* nbSymbolsPtr, U32* tableLogPtr,
|
||||
const void* src, size_t srcSize,
|
||||
void* workSpace, size_t wkspSize)
|
||||
{
|
||||
return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 0);
|
||||
}
|
||||
|
||||
#if DYNAMIC_BMI2
|
||||
static TARGET_ATTRIBUTE("bmi2") size_t HUF_readStats_body_bmi2(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
U32* nbSymbolsPtr, U32* tableLogPtr,
|
||||
const void* src, size_t srcSize,
|
||||
void* workSpace, size_t wkspSize)
|
||||
{
|
||||
return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 1);
|
||||
}
|
||||
#endif
|
||||
|
||||
size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
|
||||
U32* nbSymbolsPtr, U32* tableLogPtr,
|
||||
const void* src, size_t srcSize,
|
||||
void* workSpace, size_t wkspSize,
|
||||
int bmi2)
|
||||
{
|
||||
#if DYNAMIC_BMI2
|
||||
if (bmi2) {
|
||||
return HUF_readStats_body_bmi2(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
|
||||
}
|
||||
#endif
|
||||
(void)bmi2;
|
||||
return HUF_readStats_body_default(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
|
||||
}
|
||||
|
||||
@@ -228,6 +228,13 @@ FSE_PUBLIC_API size_t FSE_readNCount (short* normalizedCounter,
|
||||
unsigned* maxSymbolValuePtr, unsigned* tableLogPtr,
|
||||
const void* rBuffer, size_t rBuffSize);
|
||||
|
||||
/*! FSE_readNCount_bmi2():
|
||||
* Same as FSE_readNCount() but pass bmi2=1 when your CPU supports BMI2 and 0 otherwise.
|
||||
*/
|
||||
FSE_PUBLIC_API size_t FSE_readNCount_bmi2(short* normalizedCounter,
|
||||
unsigned* maxSymbolValuePtr, unsigned* tableLogPtr,
|
||||
const void* rBuffer, size_t rBuffSize, int bmi2);
|
||||
|
||||
/*! Constructor and Destructor of FSE_DTable.
|
||||
Note that its size depends on 'tableLog' */
|
||||
typedef unsigned FSE_DTable; /* don't allocate that. It's just a way to be more restrictive than void* */
|
||||
@@ -342,6 +349,9 @@ size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue);
|
||||
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize);
|
||||
/**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DECOMPRESS_WKSP_SIZE_U32(maxLog, maxSymbolValue)` */
|
||||
|
||||
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);
|
||||
/**< Same as FSE_decompress_wksp() but with dynamic BMI2 support. Pass 1 if your CPU supports BMI2 or 0 if it doesn't. */
|
||||
|
||||
typedef enum {
|
||||
FSE_repeat_none, /**< Cannot use the previous table */
|
||||
FSE_repeat_check, /**< Can use the previous table but it must be checked */
|
||||
|
||||
@@ -73,7 +73,7 @@ size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned
|
||||
return FSE_buildDTable_wksp(dt, normalizedCounter, maxSymbolValue, tableLog, wksp, sizeof(wksp));
|
||||
}
|
||||
|
||||
size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
|
||||
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);
|
||||
@@ -178,6 +178,11 @@ size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsi
|
||||
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
|
||||
|
||||
@@ -306,6 +311,15 @@ size_t FSE_decompress_usingDTable(void* dst, size_t originalSize,
|
||||
|
||||
|
||||
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
|
||||
{
|
||||
return FSE_decompress_wksp_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, /* bmi2 */ 0);
|
||||
}
|
||||
|
||||
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;
|
||||
@@ -315,7 +329,7 @@ size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size
|
||||
FSE_DTable* const dtable = (FSE_DTable*)workSpace;
|
||||
|
||||
/* normal FSE decoding mode */
|
||||
size_t const NCountLength = FSE_readNCount (counting, &maxSymbolValue, &tableLog, istart, cSrcSize);
|
||||
size_t const NCountLength = FSE_readNCount_bmi2(counting, &maxSymbolValue, &tableLog, istart, cSrcSize, bmi2);
|
||||
if (FSE_isError(NCountLength)) return NCountLength;
|
||||
if (tableLog > maxLog) return ERROR(tableLog_tooLarge);
|
||||
assert(NCountLength <= cSrcSize);
|
||||
@@ -326,9 +340,40 @@ size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size
|
||||
workSpace = dtable + FSE_DTABLE_SIZE_U32(tableLog);
|
||||
wkspSize -= FSE_DTABLE_SIZE(tableLog);
|
||||
|
||||
CHECK_F( FSE_buildDTable_wksp(dtable, counting, maxSymbolValue, tableLog, workSpace, wkspSize) );
|
||||
CHECK_F( FSE_buildDTable_internal(dtable, counting, maxSymbolValue, tableLog, workSpace, wkspSize) );
|
||||
|
||||
return FSE_decompress_usingDTable (dst, dstCapacity, ip, cSrcSize, dtable); /* always return, even if it is an error code */
|
||||
{
|
||||
const void* ptr = 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, dtable, 1);
|
||||
return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, dtable, 0);
|
||||
}
|
||||
}
|
||||
|
||||
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
|
||||
TARGET_ATTRIBUTE("bmi2") 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);
|
||||
}
|
||||
|
||||
|
||||
|
||||
+6
-1
@@ -231,13 +231,15 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize,
|
||||
/*! HUF_readStats_wksp() :
|
||||
* Same as HUF_readStats() but takes an external workspace which must be
|
||||
* 4-byte aligned and its size must be >= HUF_READ_STATS_WORKSPACE_SIZE.
|
||||
* If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0.
|
||||
*/
|
||||
#define HUF_READ_STATS_WORKSPACE_SIZE_U32 FSE_DECOMPRESS_WKSP_SIZE_U32(6, HUF_TABLELOG_MAX-1)
|
||||
#define HUF_READ_STATS_WORKSPACE_SIZE (HUF_READ_STATS_WORKSPACE_SIZE_U32 * sizeof(unsigned))
|
||||
size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize,
|
||||
U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
|
||||
const void* src, size_t srcSize,
|
||||
void* workspace, size_t wkspSize);
|
||||
void* workspace, size_t wkspSize,
|
||||
int bmi2);
|
||||
|
||||
/** HUF_readCTable() :
|
||||
* Loading a CTable saved with HUF_writeCTable() */
|
||||
@@ -345,6 +347,9 @@ size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstS
|
||||
#endif
|
||||
size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
|
||||
size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
|
||||
#ifndef HUF_FORCE_DECOMPRESS_X2
|
||||
size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2);
|
||||
#endif
|
||||
|
||||
#endif /* HUF_STATIC_LINKING_ONLY */
|
||||
|
||||
|
||||
Reference in New Issue
Block a user