From 42eb47d42f041054140b8e08ffc6ba85e9f092f2 Mon Sep 17 00:00:00 2001 From: Alexander Mohr Date: Tue, 5 Jul 2022 11:35:58 +0200 Subject: uncompressed-api: allow uncompressed_update only for independent blocks Signed-off-by: Alexander Mohr --- doc/lz4_manual.html | 12 ---- lib/lz4.c | 24 ++------ lib/lz4.h | 11 ---- lib/lz4frame.c | 172 ++++++++++++++++++++++------------------------------ lib/lz4frame.h | 1 + lib/lz4hc.c | 27 ++------- 6 files changed, 84 insertions(+), 163 deletions(-) diff --git a/doc/lz4_manual.html b/doc/lz4_manual.html index 700cb84..037cfc0 100644 --- a/doc/lz4_manual.html +++ b/doc/lz4_manual.html @@ -391,18 +391,6 @@ int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream);


-
LZ4LIB_STATIC_API int LZ4_getDictSize (const LZ4_stream_t* LZ4_dict, int dictSize);
-
 Get the size of the dictionary. This can be used for adding data without
- compression to the LZ4 archive. If linked blocked mode is used the memory
- of the dictionary is kept free.
- This way uncompressed data does not influence the effectiveness of the
- dictionary.
- @param LZ4_dict Pointer to the dictionary to get the size of.
- @param dictSize The maximum dictionary size. (Normally 64 KB).
- @return The size of the dictionary.
- 
-

- 

  It's possible to have input and output sharing the same buffer,
  for highly constrained memory environments.
diff --git a/lib/lz4.c b/lib/lz4.c
index 289162d..a2272cf 100644
--- a/lib/lz4.c
+++ b/lib/lz4.c
@@ -1679,25 +1679,6 @@ int LZ4_compress_forceExtDict (LZ4_stream_t* LZ4_dict, const char* source, char*
     return result;
 }
 
-/*! LZ4_getDictSize():
- * Get the size of the dictionary. This can be used for adding data without
- * compression to the LZ4 archive. If linked blocked mode is used the memory
- * of the dictionary is kept free.
- * This way uncompressed data does not influence the effectiveness of the
- * dictionary.
- * @param LZ4_dict Pointer to the dictionary to get the size of.
- * @param dictSize The maximum dictionary size. (Normally 64 KB).
- * @return The size of the dictionary.
- */
-int LZ4_getDictSize (const LZ4_stream_t* LZ4_dict, int dictSize)
-{
-  const LZ4_stream_t_internal* const dict = &LZ4_dict->internal_donotuse;
-
-  if ((U32)dictSize > 64 KB) { dictSize = 64 KB; } /* useless to define a dictionary > 64 KB */
-  if ((U32)dictSize > dict->dictSize) { dictSize = (int)dict->dictSize; }
-
-  return dictSize;
-}
 
 /*! LZ4_saveDict() :
  *  If previously compressed data block is not guaranteed to remain available at its memory location,
@@ -1709,9 +1690,12 @@ int LZ4_getDictSize (const LZ4_stream_t* LZ4_dict, int dictSize)
 int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize)
 {
     LZ4_stream_t_internal* const dict = &LZ4_dict->internal_donotuse;
-    dictSize = LZ4_getDictSize(LZ4_dict, dictSize);
+
     DEBUGLOG(5, "LZ4_saveDict : dictSize=%i, safeBuffer=%p", dictSize, safeBuffer);
 
+    if ((U32)dictSize > 64 KB) { dictSize = 64 KB; } /* useless to define a dictionary > 64 KB */
+    if ((U32)dictSize > dict->dictSize) { dictSize = (int)dict->dictSize; }
+
     if (safeBuffer == NULL) assert(dictSize == 0);
     if (dictSize > 0) {
         const BYTE* const previousDictEnd = dict->dictionary + dict->dictSize;
diff --git a/lib/lz4.h b/lib/lz4.h
index ce2288e..6c068c6 100644
--- a/lib/lz4.h
+++ b/lib/lz4.h
@@ -509,17 +509,6 @@ LZ4LIB_STATIC_API int LZ4_compress_fast_extState_fastReset (void* state, const c
  */
 LZ4LIB_STATIC_API void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dictionaryStream);
 
-/*! LZ4_getDictSize():
- * Get the size of the dictionary. This can be used for adding data without
- * compression to the LZ4 archive. If linked blocked mode is used the memory
- * of the dictionary is kept free.
- * This way uncompressed data does not influence the effectiveness of the
- * dictionary.
- * @param LZ4_dict Pointer to the dictionary to get the size of.
- * @param dictSize The maximum dictionary size. (Normally 64 KB).
- * @return The size of the dictionary.
- */
-LZ4LIB_STATIC_API int LZ4_getDictSize (const LZ4_stream_t* LZ4_dict, int dictSize);
 
 /*! In-place compression and decompression
  *
diff --git a/lib/lz4frame.c b/lib/lz4frame.c
index f32ed1d..3042d6f 100644
--- a/lib/lz4frame.c
+++ b/lib/lz4frame.c
@@ -841,23 +841,12 @@ static compressFunc_t LZ4F_selectCompression(LZ4F_blockMode_t blockMode, int lev
     return LZ4F_compressBlockHC_continue;
 }
 
-static int LZ4F_maxDictSize(void) {
-  return 64 KB;
-}
-
 /* Save history (up to 64KB) into @tmpBuff */
 static int LZ4F_localSaveDict(LZ4F_cctx_t* cctxPtr)
 {
     if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN)
-        return LZ4_saveDict ((LZ4_stream_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), LZ4F_maxDictSize());
-    return LZ4_saveDictHC ((LZ4_streamHC_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), LZ4F_maxDictSize());
-}
-
-static int LZ4F_localDictSize(LZ4F_cctx_t* cctxPtr)
-{
-  if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN)
-    return LZ4_getDictSize ((LZ4_stream_t*)(cctxPtr->lz4CtxPtr), LZ4F_maxDictSize());
-  return LZ4_getDictHCSize ((LZ4_streamHC_t*)(cctxPtr->lz4CtxPtr), LZ4F_maxDictSize());
+        return LZ4_saveDict ((LZ4_stream_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), 64 KB);
+    return LZ4_saveDictHC ((LZ4_streamHC_t*)(cctxPtr->lz4CtxPtr), (char*)(cctxPtr->tmpBuff), 64 KB);
 }
 
 typedef enum { notDone, fromTmpBuffer, fromSrcBuffer } LZ4F_lastBlockStatus;
@@ -886,69 +875,64 @@ static size_t LZ4F_compressUpdateImpl(LZ4F_cctx* cctxPtr,
     size_t const blockSize = cctxPtr->maxBlockSize;
     const BYTE* srcPtr = (const BYTE*)srcBuffer;
     const BYTE* const srcEnd = srcPtr + srcSize;
-    BYTE* dstStart = (BYTE*)dstBuffer;
+    BYTE* const dstStart = (BYTE*)dstBuffer;
     BYTE* dstPtr = dstStart;
     LZ4F_lastBlockStatus lastBlockCompressed = notDone;
     compressFunc_t const compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel);
-    size_t bytesWritten = 0;
+    size_t bytesWritten;
     DEBUGLOG(4, "LZ4F_compressUpdate (srcSize=%zu)", srcSize);
 
-    /* flush currently written block, to continue with new block compression */
-    if (cctxPtr->blockCompression != blockCompression) {
-      bytesWritten = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr);
-      dstStart = (BYTE*)dstBuffer + bytesWritten;
-      dstPtr = dstStart;
-      dstCapacity -= bytesWritten;
-      cctxPtr->blockCompression = blockCompression;
-    }
-
     RETURN_ERROR_IF(cctxPtr->cStage != 1, compressionState_uninitialized);   /* state must be initialized and waiting for next block */
-
-    if (blockCompression == LZ4B_COMPRESSED &&
-        dstCapacity < LZ4F_compressBound_internal(srcSize, &(cctxPtr->prefs), cctxPtr->tmpInSize))
-      RETURN_ERROR(dstMaxSize_tooSmall);
+    if (dstCapacity < LZ4F_compressBound_internal(srcSize, &(cctxPtr->prefs), cctxPtr->tmpInSize))
+        RETURN_ERROR(dstMaxSize_tooSmall);
 
     if (blockCompression == LZ4B_UNCOMPRESSED && dstCapacity < srcSize)
-      RETURN_ERROR(dstMaxSize_tooSmall);
+        RETURN_ERROR(dstMaxSize_tooSmall);
+
+    /* flush currently written block, to continue with new block compression */
+    if (cctxPtr->blockCompression != blockCompression) {
+        bytesWritten = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr);
+        dstPtr += bytesWritten;
+        cctxPtr->blockCompression = blockCompression;
+    }
 
     if (compressOptionsPtr == NULL) compressOptionsPtr = &k_cOptionsNull;
 
     /* complete tmp buffer */
     if (cctxPtr->tmpInSize > 0) {   /* some data already within tmp buffer */
-      size_t const sizeToCopy = blockSize - cctxPtr->tmpInSize;
-      assert(blockSize > cctxPtr->tmpInSize);
-      if (sizeToCopy > srcSize) {
-        /* add src to tmpIn buffer */
-        memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, srcSize);
-        srcPtr = srcEnd;
-        cctxPtr->tmpInSize += srcSize;
-        /* still needs some CRC */
-      } else {
-        /* complete tmpIn block and then compress it */
-        lastBlockCompressed = fromTmpBuffer;
-        memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, sizeToCopy);
-        srcPtr += sizeToCopy;
+        size_t const sizeToCopy = blockSize - cctxPtr->tmpInSize;
+        assert(blockSize > cctxPtr->tmpInSize);
+        if (sizeToCopy > srcSize) {
+            /* add src to tmpIn buffer */
+            memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, srcSize);
+            srcPtr = srcEnd;
+            cctxPtr->tmpInSize += srcSize;
+            /* still needs some CRC */
+        } else {
+            /* complete tmpIn block and then compress it */
+            lastBlockCompressed = fromTmpBuffer;
+            memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, sizeToCopy);
+            srcPtr += sizeToCopy;
+
+            dstPtr += LZ4F_makeBlock(dstPtr,
+                                     cctxPtr->tmpIn, blockSize,
+                                     compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel,
+                                     cctxPtr->cdict,
+                                     cctxPtr->prefs.frameInfo.blockChecksumFlag, blockCompression);
+            if (cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) cctxPtr->tmpIn += blockSize;
+            cctxPtr->tmpInSize = 0;
+    }   }
 
+    while ((size_t)(srcEnd - srcPtr) >= blockSize) {
+        /* compress full blocks */
+        lastBlockCompressed = fromSrcBuffer;
         dstPtr += LZ4F_makeBlock(dstPtr,
-                                 cctxPtr->tmpIn, blockSize,
+                                 srcPtr, blockSize,
                                  compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel,
                                  cctxPtr->cdict,
-                                 cctxPtr->prefs.frameInfo.blockChecksumFlag, blockCompression);
-
-        if (cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) cctxPtr->tmpIn += blockSize;
-        cctxPtr->tmpInSize = 0;
-      }
-    }
-
-    while ((size_t)(srcEnd - srcPtr) >= blockSize) {
-      /* compress full blocks */
-      lastBlockCompressed = fromSrcBuffer;
-      dstPtr += LZ4F_makeBlock(dstPtr,
-                               srcPtr, blockSize,
-                               compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel,
-                               cctxPtr->cdict,
-                               cctxPtr->prefs.frameInfo.blockChecksumFlag, blockCompression);
-      srcPtr += blockSize;
+                                 cctxPtr->prefs.frameInfo.blockChecksumFlag,
+                                 blockCompression);
+        srcPtr += blockSize;
     }
 
     if ((cctxPtr->prefs.autoFlush) && (srcPtr < srcEnd)) {
@@ -960,27 +944,20 @@ static size_t LZ4F_compressUpdateImpl(LZ4F_cctx* cctxPtr,
                                  cctxPtr->cdict,
                                  cctxPtr->prefs.frameInfo.blockChecksumFlag,
                                  blockCompression);
-      srcPtr = srcEnd;
+        srcPtr = srcEnd;
     }
 
     /* preserve dictionary within @tmpBuff whenever necessary */
     if ((cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) && (lastBlockCompressed==fromSrcBuffer)) {
-      if (compressOptionsPtr->stableSrc) {
-        cctxPtr->tmpIn = cctxPtr->tmpBuff;  /* src is stable : dictionary remains in src across invocations */
-      } else {
-        int realDictSize;
-        if (blockCompression == LZ4B_COMPRESSED) {
-          realDictSize = LZ4F_localSaveDict(cctxPtr);
+        /* linked blocks are only supported in compressed mode, see LZ4F_uncompressedUpdate */
+        assert(blockCompression == LZ4B_COMPRESSED);
+        if (compressOptionsPtr->stableSrc) {
+            cctxPtr->tmpIn = cctxPtr->tmpBuff;  /* src is stable : dictionary remains in src across invocations */
         } else {
-          /* only keep the space of the dictionary, so dict data is kept for the next compressedUpdate
-           * this is only relevant if linked block mode
-           * */
-          realDictSize = LZ4F_localDictSize(cctxPtr);
+            int const realDictSize = LZ4F_localSaveDict(cctxPtr);
+            assert(0 <= realDictSize && realDictSize <= 64 KB);
+            cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize;
         }
-
-        assert(0 <= realDictSize && realDictSize <= 64 KB);
-        cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize;
-      }
     }
 
     /* keep tmpIn within limits */
@@ -989,30 +966,24 @@ static size_t LZ4F_compressUpdateImpl(LZ4F_cctx* cctxPtr,
     {
         /* only preserve 64KB within internal buffer. Ensures there is enough room for next block.
          * note: this situation necessarily implies lastBlockCompressed==fromTmpBuffer */
-        int realDictSize;
-        if (blockCompression == LZ4B_COMPRESSED) {
-          realDictSize = LZ4F_localSaveDict(cctxPtr);
-        } else {
-          /* only keep the space of the dictionary, so dict data is kept for the next compressedUpdate*/
-          realDictSize = LZ4F_maxDictSize();
-        }
+        int const realDictSize = LZ4F_localSaveDict(cctxPtr);
         cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize;
         assert((cctxPtr->tmpIn + blockSize) <= (cctxPtr->tmpBuff + cctxPtr->maxBufferSize));
     }
 
     /* some input data left, necessarily < blockSize */
     if (srcPtr < srcEnd) {
-       /* fill tmp buffer */
-       size_t const sizeToCopy = (size_t)(srcEnd - srcPtr);
-       memcpy(cctxPtr->tmpIn, srcPtr, sizeToCopy);
-       cctxPtr->tmpInSize = sizeToCopy;
+        /* fill tmp buffer */
+        size_t const sizeToCopy = (size_t)(srcEnd - srcPtr);
+        memcpy(cctxPtr->tmpIn, srcPtr, sizeToCopy);
+        cctxPtr->tmpInSize = sizeToCopy;
     }
 
     if (cctxPtr->prefs.frameInfo.contentChecksumFlag == LZ4F_contentChecksumEnabled)
-      (void)XXH32_update(&(cctxPtr->xxh), srcBuffer, srcSize);
+        (void)XXH32_update(&(cctxPtr->xxh), srcBuffer, srcSize);
 
     cctxPtr->totalInSize += srcSize;
-    return bytesWritten + (size_t)(dstPtr - dstStart);
+    return (size_t)(dstPtr - dstStart);
 }
 
 /*! LZ4F_compressUpdate() :
@@ -1032,10 +1003,10 @@ size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr,
                      const void* srcBuffer, size_t srcSize,
                      const LZ4F_compressOptions_t* compressOptionsPtr)
 {
-   return LZ4F_compressUpdateImpl(cctxPtr,
-                                 dstBuffer, dstCapacity,
-                                 srcBuffer, srcSize,
-                                 compressOptionsPtr, LZ4B_COMPRESSED);
+     return LZ4F_compressUpdateImpl(cctxPtr,
+                                   dstBuffer, dstCapacity,
+                                   srcBuffer, srcSize,
+                                   compressOptionsPtr, LZ4B_COMPRESSED);
 }
 
 /*! LZ4F_compressUpdate() :
@@ -1044,6 +1015,7 @@ size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr,
  *  src data is either buffered or compressed into @dstBuffer.
  *  If previously an uncompressed block was written, buffered data is flushed
  *  before appending compressed data is continued.
+ *  This is only supported when LZ4F_blockIndependent is used
  * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr).
  * @compressOptionsPtr is optional : provide NULL to mean "default".
  * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered.
@@ -1051,13 +1023,14 @@ size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr,
  *  After an error, the state is left in a UB state, and must be re-initialized.
  */
 size_t LZ4F_uncompressedUpdate(LZ4F_cctx* cctxPtr,
-                                               void* dstBuffer, size_t dstCapacity,
-                                               const void* srcBuffer, size_t srcSize,
-                                               const LZ4F_compressOptions_t* compressOptionsPtr) {
-  return LZ4F_compressUpdateImpl(cctxPtr,
-                                 dstBuffer, dstCapacity,
-                                 srcBuffer, srcSize,
-                                 compressOptionsPtr, LZ4B_UNCOMPRESSED);
+                               void* dstBuffer, size_t dstCapacity,
+                         const void* srcBuffer, size_t srcSize,
+                         const LZ4F_compressOptions_t* compressOptionsPtr) {
+    assert(cctxPtr->prefs.frameInfo.blockMode == LZ4F_blockIndependent);
+    return LZ4F_compressUpdateImpl(cctxPtr,
+                                   dstBuffer, dstCapacity,
+                                   srcBuffer, srcSize,
+                                   compressOptionsPtr, LZ4B_UNCOMPRESSED);
 }
 
 
@@ -1090,7 +1063,8 @@ size_t LZ4F_flush(LZ4F_cctx* cctxPtr,
                              cctxPtr->tmpIn, cctxPtr->tmpInSize,
                              compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel,
                              cctxPtr->cdict,
-                             cctxPtr->prefs.frameInfo.blockChecksumFlag, cctxPtr->blockCompression);
+                             cctxPtr->prefs.frameInfo.blockChecksumFlag,
+                             cctxPtr->blockCompression);
     assert(((void)"flush overflows dstBuffer!", (size_t)(dstPtr - dstStart) <= dstCapacity));
 
     if (cctxPtr->prefs.frameInfo.blockMode == LZ4F_blockLinked)
diff --git a/lib/lz4frame.h b/lib/lz4frame.h
index 7d81fa0..691ad50 100644
--- a/lib/lz4frame.h
+++ b/lib/lz4frame.h
@@ -552,6 +552,7 @@ LZ4FLIB_STATIC_API size_t LZ4F_getBlockSize(unsigned);
  *  After an error, the state is left in a UB state, and must be re-initialized or freed.
  *  If previously a compressed block was written, buffered data is flushed
  *  before appending uncompressed data is continued.
+ *  This is only supported when LZ4F_blockIndependent is used
  * `cOptPtr` is optional : NULL can be provided, in which case all options are set to default.
  * @return : number of bytes written into `dstBuffer` (it can be zero, meaning input data was just buffered).
  *           or an error code if it fails (which can be tested using LZ4F_isError())
diff --git a/lib/lz4hc.c b/lib/lz4hc.c
index b5bc880..99650a6 100644
--- a/lib/lz4hc.c
+++ b/lib/lz4hc.c
@@ -1154,26 +1154,6 @@ int LZ4_compress_HC_continue_destSize (LZ4_streamHC_t* LZ4_streamHCPtr, const ch
     return LZ4_compressHC_continue_generic(LZ4_streamHCPtr, src, dst, srcSizePtr, targetDestSize, fillOutput);
 }
 
-/*! LZ4_getDictHCSize():
- * Get the size of the dictionary. This can be used for adding data without
- * compression to the LZ4 archive. If linked blocked mode is used the memory
- * of the dictionary is kept free.
- * This way uncompressed data does not influence the effectiveness of the
- * dictionary.
- * @param LZ4_dict Pointer to the dictionary to get the size of.
- * @param dictSize The maximum dictionary size. (Normally 64 KB).
- * @return The size of the dictionary.
- */
-int LZ4_getDictHCSize(const LZ4_streamHC_t* LZ4_streamHCPtr, int dictSize) {
-  const LZ4HC_CCtx_internal* const streamPtr = &LZ4_streamHCPtr->internal_donotuse;
-  int const prefixSize = (int)(streamPtr->end - (streamPtr->base + streamPtr->dictLimit));
-  DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize);
-  assert(prefixSize >= 0);
-  if (dictSize > 64 KB) dictSize = 64 KB;
-  if (dictSize < 4) dictSize = 0;
-  if (dictSize > prefixSize) dictSize = prefixSize;
-  return dictSize;
-}
 
 
 /* LZ4_saveDictHC :
@@ -1184,7 +1164,12 @@ int LZ4_getDictHCSize(const LZ4_streamHC_t* LZ4_streamHCPtr, int dictSize) {
 int LZ4_saveDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, char* safeBuffer, int dictSize)
 {
     LZ4HC_CCtx_internal* const streamPtr = &LZ4_streamHCPtr->internal_donotuse;
-    dictSize = LZ4_getDictHCSize(LZ4_streamHCPtr, dictSize);
+    int const prefixSize = (int)(streamPtr->end - (streamPtr->base + streamPtr->dictLimit));
+    DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize);
+    assert(prefixSize >= 0);
+    if (dictSize > 64 KB) dictSize = 64 KB;
+    if (dictSize < 4) dictSize = 0;
+    if (dictSize > prefixSize) dictSize = prefixSize;
     if (safeBuffer == NULL) assert(dictSize == 0);
     if (dictSize > 0)
         memmove(safeBuffer, streamPtr->end - dictSize, dictSize);
-- 
cgit v0.12