minor cleaning of huff0

Update code documentation, and properly names a few "magic constants".
Also, HUF_compress_internal() gets a cleaner way
to determine size of tables inside workspace.

diff --git a/lib/common/huf.h b/lib/common/huf.h
index cadee595dc16fa2651a9f76428209456ddf1ac65..4e7eee0cba82d55f84a4974abfea921e49dec12b 100644 (file)
--- a/lib/common/huf.h
+++ b/lib/common/huf.h
@@ -67,7 +67,6 @@ HUF_compress() :
     `srcSize` must be <= `HUF_BLOCKSIZE_MAX` == 128 KB.
     @return : size of compressed data (<= `dstCapacity`).
     Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!!
-                     if return == 1, srcData is a single repeated byte symbol (RLE compression).
                      if HUF_isError(return), compression failed (more details using HUF_getErrorName())
 */
 HUF_PUBLIC_API size_t HUF_compress(void* dst, size_t dstCapacity,
@@ -80,7 +79,7 @@ HUF_decompress() :
     `originalSize` : **must** be the ***exact*** size of original (uncompressed) data.
     Note : in contrast with FSE, HUF_decompress can regenerate
            RLE (cSrcSize==1) and uncompressed (cSrcSize==dstSize) data,
-           because it knows size to regenerate.
+           because it knows size to regenerate (originalSize).
     @return : size of regenerated data (== originalSize),
               or an error code, which can be tested using HUF_isError()
 */
@@ -101,6 +100,7 @@ HUF_PUBLIC_API const char* HUF_getErrorName(size_t code);  /**< provides error c
 
 /** HUF_compress2() :
  *  Same as HUF_compress(), but offers direct control over `maxSymbolValue` and `tableLog`.
+ *  `maxSymbolValue` must be <= HUF_SYMBOLVALUE_MAX .
  *  `tableLog` must be `<= HUF_TABLELOG_MAX` . */
 HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog);
 
@@ -129,7 +129,7 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity, const
 /* ******************************************************************
  *  WARNING !!
  *  The following section contains advanced and experimental definitions
- *  which shall never be used in the context of dll
+ *  which shall never be used in the context of a dynamic library,
  *  because they are not guaranteed to remain stable in the future.
  *  Only consider them in association with static linking.
  *******************************************************************/
@@ -141,11 +141,11 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity, const
 
 
 /* *** Constants *** */
-#define HUF_TABLELOG_MAX      12       /* max configured tableLog (for static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */
-#define HUF_TABLELOG_DEFAULT  11       /* tableLog by default, when not specified */
+#define HUF_TABLELOG_MAX      12      /* max runtime value of tableLog (due to static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */
+#define HUF_TABLELOG_DEFAULT  11      /* default tableLog value when none specified */
 #define HUF_SYMBOLVALUE_MAX  255
 
-#define HUF_TABLELOG_ABSOLUTEMAX  15   /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */
+#define HUF_TABLELOG_ABSOLUTEMAX  15  /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */
 #if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX)
 #  error "HUF_TABLELOG_MAX is too large !"
 #endif
@@ -227,8 +227,9 @@ size_t HUF_compress4X_repeat(void* dst, size_t dstSize, const void* src, size_t
 
 /** HUF_buildCTable_wksp() :
  *  Same as HUF_buildCTable(), but using externally allocated scratch buffer.
- *  `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of 1024 unsigned.
+ * `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of HUF_CTABLE_WORKSPACE_SIZE_U32 unsigned.
  */
+ #define HUF_CTABLE_WORKSPACE_SIZE_U32 (2*HUF_SYMBOLVALUE_MAX +1 +1)
 size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize);
 
 /*! HUF_readStats() :
@@ -236,8 +237,8 @@ size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValu
     `huffWeight` is destination buffer.
     @return : size read from `src` , or an error Code .
     Note : Needed by HUF_readCTable() and HUF_readDTableXn() . */
-size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
-                     U32* nbSymbolsPtr, U32* tableLogPtr,
+size_t HUF_readStats(BYTE* huffWeight, size_t hwSize,
+                     U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
                      const void* src, size_t srcSize);
 
 /** HUF_readCTable() :

diff --git a/lib/compress/huf_compress.c b/lib/compress/huf_compress.c
index 10320f75f75527efb5cfa77712654985d63ccca1..59c5c6c853cde5af0f19b9cf309e72058340f9a4 100644 (file)
--- a/lib/compress/huf_compress.c
+++ b/lib/compress/huf_compress.c
@@ -335,10 +335,10 @@ static void HUF_sort(nodeElt* huffNode, const U32* count, U32 maxSymbolValue)
 
 /** HUF_buildCTable_wksp() :
  *  Same as HUF_buildCTable(), but using externally allocated scratch buffer.
- *  `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of 1024 unsigned.
+ *  `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of HUF_CTABLE_WORKSPACE_SIZE_U32 unsigned.
  */
 #define STARTNODE (HUF_SYMBOLVALUE_MAX+1)
-typedef nodeElt huffNodeTable[2*HUF_SYMBOLVALUE_MAX+1 +1];
+typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
 size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
 {
     nodeElt* const huffNode0 = (nodeElt*)workSpace;
@@ -349,9 +349,10 @@ size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValu
     U32 nodeRoot;
 
     /* safety checks */
-    if (wkspSize < sizeof(huffNodeTable)) return ERROR(GENERIC);   /* workSpace is not large enough */
+    if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC);  /* must be aligned on 4-bytes boundaries */
+    if (wkspSize < sizeof(huffNodeTable)) return ERROR(workSpace_tooSmall);
     if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
-    if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(GENERIC);
+    if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
     memset(huffNode0, 0, sizeof(huffNodeTable));
 
     /* sort, decreasing order */
@@ -508,8 +509,14 @@ static size_t HUF_compressCTable_internal(
     return op-ostart;
 }
 
+typedef struct {
+    U32 count[HUF_SYMBOLVALUE_MAX + 1];
+    HUF_CElt CTable[HUF_SYMBOLVALUE_MAX + 1];
+    huffNodeTable nodeTable;
+} HUF_compress_tables_t;
 
-/* `workSpace` must a table of at least 1024 unsigned */
+/* HUF_compress_internal() :
+ * `workSpace` must a table of at least HUF_WORKSPACE_SIZE_U32 unsigned */
 static size_t HUF_compress_internal (
                 void* dst, size_t dstSize,
                 const void* src, size_t srcSize,
@@ -519,76 +526,81 @@ static size_t HUF_compress_internal (
                 HUF_CElt* oldHufTable, HUF_repeat* repeat, int preferRepeat,
                 const int bmi2)
 {
+    HUF_compress_tables_t* const table = (HUF_compress_tables_t*)workSpace;
     BYTE* const ostart = (BYTE*)dst;
     BYTE* const oend = ostart + dstSize;
     BYTE* op = ostart;
 
-    U32* count;
-    size_t const countSize = sizeof(U32) * (HUF_SYMBOLVALUE_MAX + 1);
-    HUF_CElt* CTable;
-    size_t const CTableSize = sizeof(HUF_CElt) * (HUF_SYMBOLVALUE_MAX + 1);
-
     /* checks & inits */
-    if (wkspSize < sizeof(huffNodeTable) + countSize + CTableSize) return ERROR(GENERIC);
-    if (!srcSize) return 0;  /* Uncompressed (note : 1 means rle, so first byte must be correct) */
-    if (!dstSize) return 0;  /* cannot fit within dst budget */
+    if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC);  /* must be aligned on 4-bytes boundaries */
+    if (wkspSize < sizeof(table)) return ERROR(workSpace_tooSmall);
+    if (!srcSize) return 0;  /* Uncompressed */
+    if (!dstSize) return 0;  /* cannot fit anything within dst budget */
     if (srcSize > HUF_BLOCKSIZE_MAX) return ERROR(srcSize_wrong);   /* current block size limit */
     if (huffLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
+    if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
     if (!maxSymbolValue) maxSymbolValue = HUF_SYMBOLVALUE_MAX;
     if (!huffLog) huffLog = HUF_TABLELOG_DEFAULT;
 
-    count = (U32*)workSpace;
-    workSpace = (BYTE*)workSpace + countSize;
-    wkspSize -= countSize;
-    CTable = (HUF_CElt*)workSpace;
-    workSpace = (BYTE*)workSpace + CTableSize;
-    wkspSize -= CTableSize;
-
-    /* Heuristic : If we don't need to check the validity of the old table use the old table for small inputs */
+    /* Heuristic : If old table is valid, use it for small inputs */
     if (preferRepeat && repeat && *repeat == HUF_repeat_valid) {
-        return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable, bmi2);
+        return HUF_compressCTable_internal(ostart, op, oend,
+                                           src, srcSize,
+                                           singleStream, oldHufTable, bmi2);
     }
 
     /* Scan input and build symbol stats */
-    {   CHECK_V_F(largest, FSE_count_wksp (count, &maxSymbolValue, (const BYTE*)src, srcSize, (U32*)workSpace) );
+    {   CHECK_V_F(largest, FSE_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, table->count) );
         if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; }   /* single symbol, rle */
-        if (largest <= (srcSize >> 7)+1) return 0;   /* Fast heuristic : not compressible enough */
+        if (largest <= (srcSize >> 7)+1) return 0;   /* heuristic : probably not compressible enough */
     }
 
     /* Check validity of previous table */
-    if (repeat && *repeat == HUF_repeat_check && !HUF_validateCTable(oldHufTable, count, maxSymbolValue)) {
+    if ( repeat
+      && *repeat == HUF_repeat_check
+      && !HUF_validateCTable(oldHufTable, table->count, maxSymbolValue)) {
         *repeat = HUF_repeat_none;
     }
     /* Heuristic : use existing table for small inputs */
     if (preferRepeat && repeat && *repeat != HUF_repeat_none) {
-        return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable, bmi2);
+        return HUF_compressCTable_internal(ostart, op, oend,
+                                           src, srcSize,
+                                           singleStream, oldHufTable, bmi2);
     }
 
     /* Build Huffman Tree */
     huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
-    {   CHECK_V_F(maxBits, HUF_buildCTable_wksp (CTable, count, maxSymbolValue, huffLog, workSpace, wkspSize) );
+    {   CHECK_V_F(maxBits, HUF_buildCTable_wksp(table->CTable, table->count,
+                                                maxSymbolValue, huffLog,
+                                                table->nodeTable, sizeof(table->nodeTable)) );
         huffLog = (U32)maxBits;
-        /* Zero the unused symbols so we can check it for validity */
-        memset(CTable + maxSymbolValue + 1, 0, CTableSize - (maxSymbolValue + 1) * sizeof(HUF_CElt));
+        /* Zero unused symbols in CTable, so we can check it for validity */
+        memset(table->CTable + (maxSymbolValue + 1), 0,
+               sizeof(table->CTable) - ((maxSymbolValue + 1) * sizeof(HUF_CElt)));
     }
 
     /* Write table description header */
-    {   CHECK_V_F(hSize, HUF_writeCTable (op, dstSize, CTable, maxSymbolValue, huffLog) );
-        /* Check if using the previous table will be beneficial */
+    {   CHECK_V_F(hSize, HUF_writeCTable (op, dstSize, table->CTable, maxSymbolValue, huffLog) );
+        /* Check if using previous huffman table is beneficial */
         if (repeat && *repeat != HUF_repeat_none) {
-            size_t const oldSize = HUF_estimateCompressedSize(oldHufTable, count, maxSymbolValue);
-            size_t const newSize = HUF_estimateCompressedSize(CTable, count, maxSymbolValue);
+            size_t const oldSize = HUF_estimateCompressedSize(oldHufTable, table->count, maxSymbolValue);
+            size_t const newSize = HUF_estimateCompressedSize(table->CTable, table->count, maxSymbolValue);
             if (oldSize <= hSize + newSize || hSize + 12 >= srcSize) {
-                return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable, bmi2);
-            }
-        }
-        /* Use the new table */
+                return HUF_compressCTable_internal(ostart, op, oend,
+                                                   src, srcSize,
+                                                   singleStream, oldHufTable, bmi2);
+        }   }
+
+        /* Use the new huffman table */
         if (hSize + 12ul >= srcSize) { return 0; }
         op += hSize;
         if (repeat) { *repeat = HUF_repeat_none; }
-        if (oldHufTable) { memcpy(oldHufTable, CTable, CTableSize); } /* Save the new table */
+        if (oldHufTable)
+            memcpy(oldHufTable, table->CTable, sizeof(table->CTable));  /* Save new table */
     }
-    return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, CTable, bmi2);
+    return HUF_compressCTable_internal(ostart, op, oend,
+                                       src, srcSize,
+                                       singleStream, table->CTable, bmi2);
 }
 
 
@@ -597,7 +609,10 @@ size_t HUF_compress1X_wksp (void* dst, size_t dstSize,
                       unsigned maxSymbolValue, unsigned huffLog,
                       void* workSpace, size_t wkspSize)
 {
-    return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 1 /* single stream */, workSpace, wkspSize, NULL, NULL, 0, /* bmi2 */ 0);
+    return HUF_compress_internal(dst, dstSize, src, srcSize,
+                                 maxSymbolValue, huffLog, 1 /*single stream*/,
+                                 workSpace, wkspSize,
+                                 NULL, NULL, 0, 0 /*bmi2*/);
 }
 
 size_t HUF_compress1X_repeat (void* dst, size_t dstSize,
@@ -606,7 +621,10 @@ size_t HUF_compress1X_repeat (void* dst, size_t dstSize,
                       void* workSpace, size_t wkspSize,
                       HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2)
 {
-    return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 1 /* single stream */, workSpace, wkspSize, hufTable, repeat, preferRepeat, bmi2);
+    return HUF_compress_internal(dst, dstSize, src, srcSize,
+                                 maxSymbolValue, huffLog, 1 /*single stream*/,
+                                 workSpace, wkspSize, hufTable,
+                                 repeat, preferRepeat, bmi2);
 }
 
 size_t HUF_compress1X (void* dst, size_t dstSize,
@@ -625,7 +643,10 @@ size_t HUF_compress4X_wksp (void* dst, size_t dstSize,
                       unsigned maxSymbolValue, unsigned huffLog,
                       void* workSpace, size_t wkspSize)
 {
-    return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 0 /* 4 streams */, workSpace, wkspSize, NULL, NULL, 0, /* bmi2 */ 0);
+    return HUF_compress_internal(dst, dstSize, src, srcSize,
+                                 maxSymbolValue, huffLog, 0 /*4 streams*/,
+                                 workSpace, wkspSize,
+                                 NULL, NULL, 0, 0 /*bmi2*/);
 }
 
 /* HUF_compress4X_repeat():
@@ -637,7 +658,10 @@ size_t HUF_compress4X_repeat (void* dst, size_t dstSize,
                       void* workSpace, size_t wkspSize,
                       HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2)
 {
-    return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 0 /* 4 streams */, workSpace, wkspSize, hufTable, repeat, preferRepeat, bmi2);
+    return HUF_compress_internal(dst, dstSize, src, srcSize,
+                                 maxSymbolValue, huffLog, 0 /* 4 streams */,
+                                 workSpace, wkspSize,
+                                 hufTable, repeat, preferRepeat, bmi2);
 }
 
 size_t HUF_compress2 (void* dst, size_t dstSize,
@@ -650,5 +674,5 @@ size_t HUF_compress2 (void* dst, size_t dstSize,
 
 size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSize)
 {
-    return HUF_compress2(dst, maxDstSize, src, (U32)srcSize, 255, HUF_TABLELOG_DEFAULT);
+    return HUF_compress2(dst, maxDstSize, src, srcSize, 255, HUF_TABLELOG_DEFAULT);
 }

diff --git a/lib/decompress/huf_decompress.c b/lib/decompress/huf_decompress.c
index 9afec6d5088a4b60cea6159ad74de2686594d0f0..d2e1bb4ac0c171018d1076621d461a60130c6351 100644 (file)
--- a/lib/decompress/huf_decompress.c
+++ b/lib/decompress/huf_decompress.c
@@ -168,18 +168,19 @@ HUF_decodeStreamX2(BYTE* p, BIT_DStream_t* const bitDPtr, BYTE* const pEnd, cons
     BYTE* const pStart = p;
 
     /* up to 4 symbols at a time */
-    while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) && (p <= pEnd-4)) {
+    while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd-3)) {
         HUF_DECODE_SYMBOLX2_2(p, bitDPtr);
         HUF_DECODE_SYMBOLX2_1(p, bitDPtr);
         HUF_DECODE_SYMBOLX2_2(p, bitDPtr);
         HUF_DECODE_SYMBOLX2_0(p, bitDPtr);
     }
 
-    /* closer to the end */
-    while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) && (p < pEnd))
-        HUF_DECODE_SYMBOLX2_0(p, bitDPtr);
+    /* [0-3] symbols remaining */
+    if (MEM_32bits())
+        while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd))
+            HUF_DECODE_SYMBOLX2_0(p, bitDPtr);
 
-    /* no more data to retrieve from bitstream, hence no need to reload */
+    /* no more data to retrieve from bitstream, no need to reload */
     while (p < pEnd)
         HUF_DECODE_SYMBOLX2_0(p, bitDPtr);
 
@@ -282,7 +283,7 @@ HUF_decompress4X2_usingDTable_internal_body(
 
         /* check corruption */
         /* note : should not be necessary : op# advance in lock step, and we control op4.
-         *        but curiously, binary generated by gcc 7.2 & 7.3 with -mbmi2 runs faster when >=2 tests are present */
+         *        but curiously, binary generated by gcc 7.2 & 7.3 with -mbmi2 runs faster when >=1 test is present */
         if (op1 > opStart2) return ERROR(corruption_detected);
         if (op2 > opStart3) return ERROR(corruption_detected);
         if (op3 > opStart4) return ERROR(corruption_detected);