libarchive-static 1.0.6 → 1.1.0

data/ext/{zlib-1.2.5 → zlib-1.2.13}/deflate.c

@@ -1,5 +1,5 @@
 /* deflate.c -- compress data using the deflation algorithm
- * Copyright (C) 1995-2010 Jean-loup Gailly and Mark Adler
+ * Copyright (C) 1995-2022 Jean-loup Gailly and Mark Adler
  * For conditions of distribution and use, see copyright notice in zlib.h
  */
 
@@ -37,7 +37,7 @@
  *  REFERENCES
  *
  *      Deutsch, L.P.,"DEFLATE Compressed Data Format Specification".
- *      Available in http://www.ietf.org/rfc/rfc1951.txt
+ *      Available in http://tools.ietf.org/html/rfc1951
  *
  *      A description of the Rabin and Karp algorithm is given in the book
  *         "Algorithms" by R. Sedgewick, Addison-Wesley, p252.
@@ -52,7 +52,7 @@
 #include "deflate.h"
 
 const char deflate_copyright[] =
-   " deflate 1.2.5 Copyright 1995-2010 Jean-loup Gailly and Mark Adler ";
+   " deflate 1.2.13 Copyright 1995-2022 Jean-loup Gailly and Mark Adler ";
 /*
   If you use the zlib library in a product, an acknowledgment is welcome
   in the documentation of your product. If for some reason you cannot
@@ -73,6 +73,8 @@ typedef enum {
 typedef block_state (*compress_func) OF((deflate_state *s, int flush));
 /* Compression function. Returns the block state after the call. */
 
+local int deflateStateCheck      OF((z_streamp strm));
+local void slide_hash     OF((deflate_state *s));
 local void fill_window    OF((deflate_state *s));
 local block_state deflate_stored OF((deflate_state *s, int flush));
 local block_state deflate_fast   OF((deflate_state *s, int flush));
@@ -84,15 +86,10 @@ local block_state deflate_huff   OF((deflate_state *s, int flush));
 local void lm_init        OF((deflate_state *s));
 local void putShortMSB    OF((deflate_state *s, uInt b));
 local void flush_pending  OF((z_streamp strm));
-local int read_buf        OF((z_streamp strm, Bytef *buf, unsigned size));
-#ifdef ASMV
-      void match_init OF((void)); /* asm code initialization */
-      uInt longest_match  OF((deflate_state *s, IPos cur_match));
-#else
+local unsigned read_buf   OF((z_streamp strm, Bytef *buf, unsigned size));
 local uInt longest_match  OF((deflate_state *s, IPos cur_match));
-#endif
 
-#ifdef DEBUG
+#ifdef ZLIB_DEBUG
 local  void check_match OF((deflate_state *s, IPos start, IPos match,
                             int length));
 #endif
@@ -148,20 +145,16 @@ local const config configuration_table[10] = {
  * meaning.
  */
 
-#define EQUAL 0
-/* result of memcmp for equal strings */
-
-#ifndef NO_DUMMY_DECL
-struct static_tree_desc_s {int dummy;}; /* for buggy compilers */
-#endif
+/* rank Z_BLOCK between Z_NO_FLUSH and Z_PARTIAL_FLUSH */
+#define RANK(f) (((f) * 2) - ((f) > 4 ? 9 : 0))
 
 /* ===========================================================================
  * Update a hash value with the given input byte
- * IN  assertion: all calls to to UPDATE_HASH are made with consecutive
- *    input characters, so that a running hash key can be computed from the
- *    previous key instead of complete recalculation each time.
+ * IN  assertion: all calls to UPDATE_HASH are made with consecutive input
+ *    characters, so that a running hash key can be computed from the previous
+ *    key instead of complete recalculation each time.
  */
-#define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask)
+#define UPDATE_HASH(s,h,c) (h = (((h) << s->hash_shift) ^ (c)) & s->hash_mask)
 
 
 /* ===========================================================================
@@ -170,9 +163,9 @@ struct static_tree_desc_s {int dummy;}; /* for buggy compilers */
  * the previous length of the hash chain.
  * If this file is compiled with -DFASTEST, the compression level is forced
  * to 1, and no hash chains are maintained.
- * IN  assertion: all calls to to INSERT_STRING are made with consecutive
- *    input characters and the first MIN_MATCH bytes of str are valid
- *    (except for the last MIN_MATCH-1 bytes of the input file).
+ * IN  assertion: all calls to INSERT_STRING are made with consecutive input
+ *    characters and the first MIN_MATCH bytes of str are valid (except for
+ *    the last MIN_MATCH-1 bytes of the input file).
  */
 #ifdef FASTEST
 #define INSERT_STRING(s, str, match_head) \
@@ -191,8 +184,42 @@ struct static_tree_desc_s {int dummy;}; /* for buggy compilers */
  * prev[] will be initialized on the fly.
  */
 #define CLEAR_HASH(s) \
-    s->head[s->hash_size-1] = NIL; \
-    zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head));
+    do { \
+        s->head[s->hash_size - 1] = NIL; \
+        zmemzero((Bytef *)s->head, \
+                 (unsigned)(s->hash_size - 1)*sizeof(*s->head)); \
+    } while (0)
+
+/* ===========================================================================
+ * Slide the hash table when sliding the window down (could be avoided with 32
+ * bit values at the expense of memory usage). We slide even when level == 0 to
+ * keep the hash table consistent if we switch back to level > 0 later.
+ */
+local void slide_hash(s)
+    deflate_state *s;
+{
+    unsigned n, m;
+    Posf *p;
+    uInt wsize = s->w_size;
+
+    n = s->hash_size;
+    p = &s->head[n];
+    do {
+        m = *--p;
+        *p = (Pos)(m >= wsize ? m - wsize : NIL);
+    } while (--n);
+    n = wsize;
+#ifndef FASTEST
+    p = &s->prev[n];
+    do {
+        m = *--p;
+        *p = (Pos)(m >= wsize ? m - wsize : NIL);
+        /* If n is not on any hash chain, prev[n] is garbage but
+         * its value will never be used.
+         */
+    } while (--n);
+#endif
+}
 
 /* ========================================================================= */
 int ZEXPORT deflateInit_(strm, level, version, stream_size)
@@ -222,11 +249,6 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
     int wrap = 1;
     static const char my_version[] = ZLIB_VERSION;
 
-    ushf *overlay;
-    /* We overlay pending_buf and d_buf+l_buf. This works since the average
-     * output size for (length,distance) codes is <= 24 bits.
-     */
-
     if (version == Z_NULL || version[0] != my_version[0] ||
         stream_size != sizeof(z_stream)) {
         return Z_VERSION_ERROR;
@@ -235,10 +257,19 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
 
     strm->msg = Z_NULL;
     if (strm->zalloc == (alloc_func)0) {
+#ifdef Z_SOLO
+        return Z_STREAM_ERROR;
+#else
         strm->zalloc = zcalloc;
         strm->opaque = (voidpf)0;
+#endif
     }
-    if (strm->zfree == (free_func)0) strm->zfree = zcfree;
+    if (strm->zfree == (free_func)0)
+#ifdef Z_SOLO
+        return Z_STREAM_ERROR;
+#else
+        strm->zfree = zcfree;
+#endif
 
 #ifdef FASTEST
     if (level != 0) level = 1;
@@ -248,6 +279,8 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
 
     if (windowBits < 0) { /* suppress zlib wrapper */
         wrap = 0;
+        if (windowBits < -15)
+            return Z_STREAM_ERROR;
         windowBits = -windowBits;
     }
 #ifdef GZIP
@@ -258,7 +291,7 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
 #endif
     if (memLevel < 1 || memLevel > MAX_MEM_LEVEL || method != Z_DEFLATED ||
         windowBits < 8 || windowBits > 15 || level < 0 || level > 9 ||
-        strategy < 0 || strategy > Z_FIXED) {
+        strategy < 0 || strategy > Z_FIXED || (windowBits == 8 && wrap != 1)) {
         return Z_STREAM_ERROR;
     }
     if (windowBits == 8) windowBits = 9;  /* until 256-byte window bug fixed */
@@ -266,17 +299,18 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
     if (s == Z_NULL) return Z_MEM_ERROR;
     strm->state = (struct internal_state FAR *)s;
     s->strm = strm;
+    s->status = INIT_STATE;     /* to pass state test in deflateReset() */
 
     s->wrap = wrap;
     s->gzhead = Z_NULL;
-    s->w_bits = windowBits;
+    s->w_bits = (uInt)windowBits;
     s->w_size = 1 << s->w_bits;
     s->w_mask = s->w_size - 1;
 
-    s->hash_bits = memLevel + 7;
+    s->hash_bits = (uInt)memLevel + 7;
     s->hash_size = 1 << s->hash_bits;
     s->hash_mask = s->hash_size - 1;
-    s->hash_shift =  ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH);
+    s->hash_shift =  ((s->hash_bits + MIN_MATCH-1) / MIN_MATCH);
 
     s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte));
     s->prev   = (Posf *)  ZALLOC(strm, s->w_size, sizeof(Pos));
@@ -286,19 +320,61 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
 
     s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */
 
-    overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2);
-    s->pending_buf = (uchf *) overlay;
-    s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L);
+    /* We overlay pending_buf and sym_buf. This works since the average size
+     * for length/distance pairs over any compressed block is assured to be 31
+     * bits or less.
+     *
+     * Analysis: The longest fixed codes are a length code of 8 bits plus 5
+     * extra bits, for lengths 131 to 257. The longest fixed distance codes are
+     * 5 bits plus 13 extra bits, for distances 16385 to 32768. The longest
+     * possible fixed-codes length/distance pair is then 31 bits total.
+     *
+     * sym_buf starts one-fourth of the way into pending_buf. So there are
+     * three bytes in sym_buf for every four bytes in pending_buf. Each symbol
+     * in sym_buf is three bytes -- two for the distance and one for the
+     * literal/length. As each symbol is consumed, the pointer to the next
+     * sym_buf value to read moves forward three bytes. From that symbol, up to
+     * 31 bits are written to pending_buf. The closest the written pending_buf
+     * bits gets to the next sym_buf symbol to read is just before the last
+     * code is written. At that time, 31*(n - 2) bits have been written, just
+     * after 24*(n - 2) bits have been consumed from sym_buf. sym_buf starts at
+     * 8*n bits into pending_buf. (Note that the symbol buffer fills when n - 1
+     * symbols are written.) The closest the writing gets to what is unread is
+     * then n + 14 bits. Here n is lit_bufsize, which is 16384 by default, and
+     * can range from 128 to 32768.
+     *
+     * Therefore, at a minimum, there are 142 bits of space between what is
+     * written and what is read in the overlain buffers, so the symbols cannot
+     * be overwritten by the compressed data. That space is actually 139 bits,
+     * due to the three-bit fixed-code block header.
+     *
+     * That covers the case where either Z_FIXED is specified, forcing fixed
+     * codes, or when the use of fixed codes is chosen, because that choice
+     * results in a smaller compressed block than dynamic codes. That latter
+     * condition then assures that the above analysis also covers all dynamic
+     * blocks. A dynamic-code block will only be chosen to be emitted if it has
+     * fewer bits than a fixed-code block would for the same set of symbols.
+     * Therefore its average symbol length is assured to be less than 31. So
+     * the compressed data for a dynamic block also cannot overwrite the
+     * symbols from which it is being constructed.
+     */
+
+    s->pending_buf = (uchf *) ZALLOC(strm, s->lit_bufsize, 4);
+    s->pending_buf_size = (ulg)s->lit_bufsize * 4;
 
     if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL ||
         s->pending_buf == Z_NULL) {
         s->status = FINISH_STATE;
-        strm->msg = (char*)ERR_MSG(Z_MEM_ERROR);
+        strm->msg = ERR_MSG(Z_MEM_ERROR);
         deflateEnd (strm);
         return Z_MEM_ERROR;
     }
-    s->d_buf = overlay + s->lit_bufsize/sizeof(ush);
-    s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize;
+    s->sym_buf = s->pending_buf + s->lit_bufsize;
+    s->sym_end = (s->lit_bufsize - 1) * 3;
+    /* We avoid equality with lit_bufsize*3 because of wraparound at 64K
+     * on 16 bit machines and because stored blocks are restricted to
+     * 64K-1 bytes.
+     */
 
     s->level = level;
     s->strategy = strategy;
@@ -307,56 +383,129 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
     return deflateReset(strm);
 }
 
+/* =========================================================================
+ * Check for a valid deflate stream state. Return 0 if ok, 1 if not.
+ */
+local int deflateStateCheck(strm)
+    z_streamp strm;
+{
+    deflate_state *s;
+    if (strm == Z_NULL ||
+        strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0)
+        return 1;
+    s = strm->state;
+    if (s == Z_NULL || s->strm != strm || (s->status != INIT_STATE &&
+#ifdef GZIP
+                                           s->status != GZIP_STATE &&
+#endif
+                                           s->status != EXTRA_STATE &&
+                                           s->status != NAME_STATE &&
+                                           s->status != COMMENT_STATE &&
+                                           s->status != HCRC_STATE &&
+                                           s->status != BUSY_STATE &&
+                                           s->status != FINISH_STATE))
+        return 1;
+    return 0;
+}
+
 /* ========================================================================= */
-int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength)
+int ZEXPORT deflateSetDictionary(strm, dictionary, dictLength)
     z_streamp strm;
     const Bytef *dictionary;
     uInt  dictLength;
 {
     deflate_state *s;
-    uInt length = dictLength;
-    uInt n;
-    IPos hash_head = 0;
+    uInt str, n;
+    int wrap;
+    unsigned avail;
+    z_const unsigned char *next;
 
-    if (strm == Z_NULL || strm->state == Z_NULL || dictionary == Z_NULL ||
-        strm->state->wrap == 2 ||
-        (strm->state->wrap == 1 && strm->state->status != INIT_STATE))
+    if (deflateStateCheck(strm) || dictionary == Z_NULL)
         return Z_STREAM_ERROR;
-
     s = strm->state;
-    if (s->wrap)
-        strm->adler = adler32(strm->adler, dictionary, dictLength);
+    wrap = s->wrap;
+    if (wrap == 2 || (wrap == 1 && s->status != INIT_STATE) || s->lookahead)
+        return Z_STREAM_ERROR;
 
-    if (length < MIN_MATCH) return Z_OK;
-    if (length > s->w_size) {
-        length = s->w_size;
-        dictionary += dictLength - length; /* use the tail of the dictionary */
+    /* when using zlib wrappers, compute Adler-32 for provided dictionary */
+    if (wrap == 1)
+        strm->adler = adler32(strm->adler, dictionary, dictLength);
+    s->wrap = 0;                    /* avoid computing Adler-32 in read_buf */
+
+    /* if dictionary would fill window, just replace the history */
+    if (dictLength >= s->w_size) {
+        if (wrap == 0) {            /* already empty otherwise */
+            CLEAR_HASH(s);
+            s->strstart = 0;
+            s->block_start = 0L;
+            s->insert = 0;
+        }
+        dictionary += dictLength - s->w_size;  /* use the tail */
+        dictLength = s->w_size;
     }
-    zmemcpy(s->window, dictionary, length);
-    s->strstart = length;
-    s->block_start = (long)length;
 
-    /* Insert all strings in the hash table (except for the last two bytes).
-     * s->lookahead stays null, so s->ins_h will be recomputed at the next
-     * call of fill_window.
-     */
-    s->ins_h = s->window[0];
-    UPDATE_HASH(s, s->ins_h, s->window[1]);
-    for (n = 0; n <= length - MIN_MATCH; n++) {
-        INSERT_STRING(s, n, hash_head);
+    /* insert dictionary into window and hash */
+    avail = strm->avail_in;
+    next = strm->next_in;
+    strm->avail_in = dictLength;
+    strm->next_in = (z_const Bytef *)dictionary;
+    fill_window(s);
+    while (s->lookahead >= MIN_MATCH) {
+        str = s->strstart;
+        n = s->lookahead - (MIN_MATCH-1);
+        do {
+            UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]);
+#ifndef FASTEST
+            s->prev[str & s->w_mask] = s->head[s->ins_h];
+#endif
+            s->head[s->ins_h] = (Pos)str;
+            str++;
+        } while (--n);
+        s->strstart = str;
+        s->lookahead = MIN_MATCH-1;
+        fill_window(s);
     }
-    if (hash_head) hash_head = 0;  /* to make compiler happy */
+    s->strstart += s->lookahead;
+    s->block_start = (long)s->strstart;
+    s->insert = s->lookahead;
+    s->lookahead = 0;
+    s->match_length = s->prev_length = MIN_MATCH-1;
+    s->match_available = 0;
+    strm->next_in = next;
+    strm->avail_in = avail;
+    s->wrap = wrap;
+    return Z_OK;
+}
+
+/* ========================================================================= */
+int ZEXPORT deflateGetDictionary(strm, dictionary, dictLength)
+    z_streamp strm;
+    Bytef *dictionary;
+    uInt  *dictLength;
+{
+    deflate_state *s;
+    uInt len;
+
+    if (deflateStateCheck(strm))
+        return Z_STREAM_ERROR;
+    s = strm->state;
+    len = s->strstart + s->lookahead;
+    if (len > s->w_size)
+        len = s->w_size;
+    if (dictionary != Z_NULL && len)
+        zmemcpy(dictionary, s->window + s->strstart + s->lookahead - len, len);
+    if (dictLength != Z_NULL)
+        *dictLength = len;
     return Z_OK;
 }
 
 /* ========================================================================= */
-int ZEXPORT deflateReset (strm)
+int ZEXPORT deflateResetKeep(strm)
     z_streamp strm;
 {
     deflate_state *s;
 
-    if (strm == Z_NULL || strm->state == Z_NULL ||
-        strm->zalloc == (alloc_func)0 || strm->zfree == (free_func)0) {
+    if (deflateStateCheck(strm)) {
         return Z_STREAM_ERROR;
     }
 
@@ -371,40 +520,84 @@ int ZEXPORT deflateReset (strm)
     if (s->wrap < 0) {
         s->wrap = -s->wrap; /* was made negative by deflate(..., Z_FINISH); */
     }
-    s->status = s->wrap ? INIT_STATE : BUSY_STATE;
+    s->status =
+#ifdef GZIP
+        s->wrap == 2 ? GZIP_STATE :
+#endif
+        INIT_STATE;
     strm->adler =
 #ifdef GZIP
         s->wrap == 2 ? crc32(0L, Z_NULL, 0) :
 #endif
         adler32(0L, Z_NULL, 0);
-    s->last_flush = Z_NO_FLUSH;
+    s->last_flush = -2;
 
     _tr_init(s);
-    lm_init(s);
 
     return Z_OK;
 }
 
 /* ========================================================================= */
-int ZEXPORT deflateSetHeader (strm, head)
+int ZEXPORT deflateReset(strm)
+    z_streamp strm;
+{
+    int ret;
+
+    ret = deflateResetKeep(strm);
+    if (ret == Z_OK)
+        lm_init(strm->state);
+    return ret;
+}
+
+/* ========================================================================= */
+int ZEXPORT deflateSetHeader(strm, head)
     z_streamp strm;
     gz_headerp head;
 {
-    if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
-    if (strm->state->wrap != 2) return Z_STREAM_ERROR;
+    if (deflateStateCheck(strm) || strm->state->wrap != 2)
+        return Z_STREAM_ERROR;
     strm->state->gzhead = head;
     return Z_OK;
 }
 
 /* ========================================================================= */
-int ZEXPORT deflatePrime (strm, bits, value)
+int ZEXPORT deflatePending(strm, pending, bits)
+    unsigned *pending;
+    int *bits;
+    z_streamp strm;
+{
+    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
+    if (pending != Z_NULL)
+        *pending = strm->state->pending;
+    if (bits != Z_NULL)
+        *bits = strm->state->bi_valid;
+    return Z_OK;
+}
+
+/* ========================================================================= */
+int ZEXPORT deflatePrime(strm, bits, value)
     z_streamp strm;
     int bits;
     int value;
 {
-    if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
-    strm->state->bi_valid = bits;
-    strm->state->bi_buf = (ush)(value & ((1 << bits) - 1));
+    deflate_state *s;
+    int put;
+
+    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
+    s = strm->state;
+    if (bits < 0 || bits > 16 ||
+        s->sym_buf < s->pending_out + ((Buf_size + 7) >> 3))
+        return Z_BUF_ERROR;
+    do {
+        put = Buf_size - s->bi_valid;
+        if (put > bits)
+            put = bits;
+        s->bi_buf |= (ush)((value & ((1 << put) - 1)) << s->bi_valid);
+        s->bi_valid += put;
+        _tr_flush_bits(s);
+        value >>= put;
+        bits -= put;
+    } while (bits);
     return Z_OK;
 }
 
@@ -416,9 +609,8 @@ int ZEXPORT deflateParams(strm, level, strategy)
 {
     deflate_state *s;
     compress_func func;
-    int err = Z_OK;
 
-    if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
+    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
     s = strm->state;
 
 #ifdef FASTEST
@@ -432,11 +624,22 @@ int ZEXPORT deflateParams(strm, level, strategy)
     func = configuration_table[s->level].func;
 
     if ((strategy != s->strategy || func != configuration_table[level].func) &&
-        strm->total_in != 0) {
+        s->last_flush != -2) {
         /* Flush the last buffer: */
-        err = deflate(strm, Z_BLOCK);
+        int err = deflate(strm, Z_BLOCK);
+        if (err == Z_STREAM_ERROR)
+            return err;
+        if (strm->avail_in || (s->strstart - s->block_start) + s->lookahead)
+            return Z_BUF_ERROR;
     }
     if (s->level != level) {
+        if (s->level == 0 && s->matches != 0) {
+            if (s->matches == 1)
+                slide_hash(s);
+            else
+                CLEAR_HASH(s);
+            s->matches = 0;
+        }
         s->level = level;
         s->max_lazy_match   = configuration_table[level].max_lazy;
         s->good_match       = configuration_table[level].good_length;
@@ -444,7 +647,7 @@ int ZEXPORT deflateParams(strm, level, strategy)
         s->max_chain_length = configuration_table[level].max_chain;
     }
     s->strategy = strategy;
-    return err;
+    return Z_OK;
 }
 
 /* ========================================================================= */
@@ -457,47 +660,60 @@ int ZEXPORT deflateTune(strm, good_length, max_lazy, nice_length, max_chain)
 {
     deflate_state *s;
 
-    if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
+    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
     s = strm->state;
-    s->good_match = good_length;
-    s->max_lazy_match = max_lazy;
+    s->good_match = (uInt)good_length;
+    s->max_lazy_match = (uInt)max_lazy;
     s->nice_match = nice_length;
-    s->max_chain_length = max_chain;
+    s->max_chain_length = (uInt)max_chain;
     return Z_OK;
 }
 
 /* =========================================================================
- * For the default windowBits of 15 and memLevel of 8, this function returns
- * a close to exact, as well as small, upper bound on the compressed size.
- * They are coded as constants here for a reason--if the #define's are
- * changed, then this function needs to be changed as well.  The return
- * value for 15 and 8 only works for those exact settings.
+ * For the default windowBits of 15 and memLevel of 8, this function returns a
+ * close to exact, as well as small, upper bound on the compressed size. This
+ * is an expansion of ~0.03%, plus a small constant.
+ *
+ * For any setting other than those defaults for windowBits and memLevel, one
+ * of two worst case bounds is returned. This is at most an expansion of ~4% or
+ * ~13%, plus a small constant.
+ *
+ * Both the 0.03% and 4% derive from the overhead of stored blocks. The first
+ * one is for stored blocks of 16383 bytes (memLevel == 8), whereas the second
+ * is for stored blocks of 127 bytes (the worst case memLevel == 1). The
+ * expansion results from five bytes of header for each stored block.
  *
- * For any setting other than those defaults for windowBits and memLevel,
- * the value returned is a conservative worst case for the maximum expansion
- * resulting from using fixed blocks instead of stored blocks, which deflate
- * can emit on compressed data for some combinations of the parameters.
+ * The larger expansion of 13% results from a window size less than or equal to
+ * the symbols buffer size (windowBits <= memLevel + 7). In that case some of
+ * the data being compressed may have slid out of the sliding window, impeding
+ * a stored block from being emitted. Then the only choice is a fixed or
+ * dynamic block, where a fixed block limits the maximum expansion to 9 bits
+ * per 8-bit byte, plus 10 bits for every block. The smallest block size for
+ * which this can occur is 255 (memLevel == 2).
  *
- * This function could be more sophisticated to provide closer upper bounds for
- * every combination of windowBits and memLevel.  But even the conservative
- * upper bound of about 14% expansion does not seem onerous for output buffer
- * allocation.
+ * Shifts are used to approximate divisions, for speed.
  */
 uLong ZEXPORT deflateBound(strm, sourceLen)
     z_streamp strm;
     uLong sourceLen;
 {
     deflate_state *s;
-    uLong complen, wraplen;
-    Bytef *str;
+    uLong fixedlen, storelen, wraplen;
 
-    /* conservative upper bound for compressed data */
-    complen = sourceLen +
-              ((sourceLen + 7) >> 3) + ((sourceLen + 63) >> 6) + 5;
+    /* upper bound for fixed blocks with 9-bit literals and length 255
+       (memLevel == 2, which is the lowest that may not use stored blocks) --
+       ~13% overhead plus a small constant */
+    fixedlen = sourceLen + (sourceLen >> 3) + (sourceLen >> 8) +
+               (sourceLen >> 9) + 4;
 
-    /* if can't get parameters, return conservative bound plus zlib wrapper */
-    if (strm == Z_NULL || strm->state == Z_NULL)
-        return complen + 6;
+    /* upper bound for stored blocks with length 127 (memLevel == 1) --
+       ~4% overhead plus a small constant */
+    storelen = sourceLen + (sourceLen >> 5) + (sourceLen >> 7) +
+               (sourceLen >> 11) + 7;
+
+    /* if can't get parameters, return larger bound plus a zlib wrapper */
+    if (deflateStateCheck(strm))
+        return (fixedlen > storelen ? fixedlen : storelen) + 6;
 
     /* compute wrapper length */
     s = strm->state;
@@ -508,9 +724,11 @@ uLong ZEXPORT deflateBound(strm, sourceLen)
     case 1:                                 /* zlib wrapper */
         wraplen = 6 + (s->strstart ? 4 : 0);
         break;
+#ifdef GZIP
     case 2:                                 /* gzip wrapper */
         wraplen = 18;
         if (s->gzhead != Z_NULL) {          /* user-supplied gzip header */
+            Bytef *str;
             if (s->gzhead->extra != Z_NULL)
                 wraplen += 2 + s->gzhead->extra_len;
             str = s->gzhead->name;
@@ -527,15 +745,17 @@ uLong ZEXPORT deflateBound(strm, sourceLen)
                 wraplen += 2;
         }
         break;
+#endif
     default:                                /* for compiler happiness */
         wraplen = 6;
     }
 
-    /* if not default parameters, return conservative bound */
+    /* if not default parameters, return one of the conservative bounds */
     if (s->w_bits != 15 || s->hash_bits != 8 + 7)
-        return complen + wraplen;
+        return (s->w_bits <= s->hash_bits ? fixedlen : storelen) + wraplen;
 
-    /* default settings: return tight bound for that case */
+    /* default settings: return tight bound for that case -- ~0.03% overhead
+       plus a small constant */
     return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
            (sourceLen >> 25) + 13 - 6 + wraplen;
 }
@@ -545,7 +765,7 @@ uLong ZEXPORT deflateBound(strm, sourceLen)
  * IN assertion: the stream state is correct and there is enough room in
  * pending_buf.
  */
-local void putShortMSB (s, b)
+local void putShortMSB(s, b)
     deflate_state *s;
     uInt b;
 {
@@ -554,262 +774,276 @@ local void putShortMSB (s, b)
 }
 
 /* =========================================================================
- * Flush as much pending output as possible. All deflate() output goes
- * through this function so some applications may wish to modify it
- * to avoid allocating a large strm->next_out buffer and copying into it.
- * (See also read_buf()).
+ * Flush as much pending output as possible. All deflate() output, except for
+ * some deflate_stored() output, goes through this function so some
+ * applications may wish to modify it to avoid allocating a large
+ * strm->next_out buffer and copying into it. (See also read_buf()).
  */
 local void flush_pending(strm)
     z_streamp strm;
 {
-    unsigned len = strm->state->pending;
+    unsigned len;
+    deflate_state *s = strm->state;
 
+    _tr_flush_bits(s);
+    len = s->pending;
     if (len > strm->avail_out) len = strm->avail_out;
     if (len == 0) return;
 
-    zmemcpy(strm->next_out, strm->state->pending_out, len);
+    zmemcpy(strm->next_out, s->pending_out, len);
     strm->next_out  += len;
-    strm->state->pending_out  += len;
+    s->pending_out  += len;
     strm->total_out += len;
-    strm->avail_out  -= len;
-    strm->state->pending -= len;
-    if (strm->state->pending == 0) {
-        strm->state->pending_out = strm->state->pending_buf;
+    strm->avail_out -= len;
+    s->pending      -= len;
+    if (s->pending == 0) {
+        s->pending_out = s->pending_buf;
     }
 }
 
+/* ===========================================================================
+ * Update the header CRC with the bytes s->pending_buf[beg..s->pending - 1].
+ */
+#define HCRC_UPDATE(beg) \
+    do { \
+        if (s->gzhead->hcrc && s->pending > (beg)) \
+            strm->adler = crc32(strm->adler, s->pending_buf + (beg), \
+                                s->pending - (beg)); \
+    } while (0)
+
 /* ========================================================================= */
-int ZEXPORT deflate (strm, flush)
+int ZEXPORT deflate(strm, flush)
     z_streamp strm;
     int flush;
 {
     int old_flush; /* value of flush param for previous deflate call */
     deflate_state *s;
 
-    if (strm == Z_NULL || strm->state == Z_NULL ||
-        flush > Z_BLOCK || flush < 0) {
+    if (deflateStateCheck(strm) || flush > Z_BLOCK || flush < 0) {
         return Z_STREAM_ERROR;
     }
     s = strm->state;
 
     if (strm->next_out == Z_NULL ||
-        (strm->next_in == Z_NULL && strm->avail_in != 0) ||
+        (strm->avail_in != 0 && strm->next_in == Z_NULL) ||
         (s->status == FINISH_STATE && flush != Z_FINISH)) {
         ERR_RETURN(strm, Z_STREAM_ERROR);
     }
     if (strm->avail_out == 0) ERR_RETURN(strm, Z_BUF_ERROR);
 
-    s->strm = strm; /* just in case */
     old_flush = s->last_flush;
     s->last_flush = flush;
 
+    /* Flush as much pending output as possible */
+    if (s->pending != 0) {
+        flush_pending(strm);
+        if (strm->avail_out == 0) {
+            /* Since avail_out is 0, deflate will be called again with
+             * more output space, but possibly with both pending and
+             * avail_in equal to zero. There won't be anything to do,
+             * but this is not an error situation so make sure we
+             * return OK instead of BUF_ERROR at next call of deflate:
+             */
+            s->last_flush = -1;
+            return Z_OK;
+        }
+
+    /* Make sure there is something to do and avoid duplicate consecutive
+     * flushes. For repeated and useless calls with Z_FINISH, we keep
+     * returning Z_STREAM_END instead of Z_BUF_ERROR.
+     */
+    } else if (strm->avail_in == 0 && RANK(flush) <= RANK(old_flush) &&
+               flush != Z_FINISH) {
+        ERR_RETURN(strm, Z_BUF_ERROR);
+    }
+
+    /* User must not provide more input after the first FINISH: */
+    if (s->status == FINISH_STATE && strm->avail_in != 0) {
+        ERR_RETURN(strm, Z_BUF_ERROR);
+    }
+
     /* Write the header */
+    if (s->status == INIT_STATE && s->wrap == 0)
+        s->status = BUSY_STATE;
     if (s->status == INIT_STATE) {
-#ifdef GZIP
-        if (s->wrap == 2) {
-            strm->adler = crc32(0L, Z_NULL, 0);
-            put_byte(s, 31);
-            put_byte(s, 139);
-            put_byte(s, 8);
-            if (s->gzhead == Z_NULL) {
-                put_byte(s, 0);
-                put_byte(s, 0);
-                put_byte(s, 0);
-                put_byte(s, 0);
-                put_byte(s, 0);
-                put_byte(s, s->level == 9 ? 2 :
-                            (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
-                             4 : 0));
-                put_byte(s, OS_CODE);
-                s->status = BUSY_STATE;
-            }
-            else {
-                put_byte(s, (s->gzhead->text ? 1 : 0) +
-                            (s->gzhead->hcrc ? 2 : 0) +
-                            (s->gzhead->extra == Z_NULL ? 0 : 4) +
-                            (s->gzhead->name == Z_NULL ? 0 : 8) +
-                            (s->gzhead->comment == Z_NULL ? 0 : 16)
-                        );
-                put_byte(s, (Byte)(s->gzhead->time & 0xff));
-                put_byte(s, (Byte)((s->gzhead->time >> 8) & 0xff));
-                put_byte(s, (Byte)((s->gzhead->time >> 16) & 0xff));
-                put_byte(s, (Byte)((s->gzhead->time >> 24) & 0xff));
-                put_byte(s, s->level == 9 ? 2 :
-                            (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
-                             4 : 0));
-                put_byte(s, s->gzhead->os & 0xff);
-                if (s->gzhead->extra != Z_NULL) {
-                    put_byte(s, s->gzhead->extra_len & 0xff);
-                    put_byte(s, (s->gzhead->extra_len >> 8) & 0xff);
-                }
-                if (s->gzhead->hcrc)
-                    strm->adler = crc32(strm->adler, s->pending_buf,
-                                        s->pending);
-                s->gzindex = 0;
-                s->status = EXTRA_STATE;
-            }
-        }
+        /* zlib header */
+        uInt header = (Z_DEFLATED + ((s->w_bits - 8) << 4)) << 8;
+        uInt level_flags;
+
+        if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2)
+            level_flags = 0;
+        else if (s->level < 6)
+            level_flags = 1;
+        else if (s->level == 6)
+            level_flags = 2;
         else
-#endif
-        {
-            uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8;
-            uInt level_flags;
-
-            if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2)
-                level_flags = 0;
-            else if (s->level < 6)
-                level_flags = 1;
-            else if (s->level == 6)
-                level_flags = 2;
-            else
-                level_flags = 3;
-            header |= (level_flags << 6);
-            if (s->strstart != 0) header |= PRESET_DICT;
-            header += 31 - (header % 31);
+            level_flags = 3;
+        header |= (level_flags << 6);
+        if (s->strstart != 0) header |= PRESET_DICT;
+        header += 31 - (header % 31);
 
+        putShortMSB(s, header);
+
+        /* Save the adler32 of the preset dictionary: */
+        if (s->strstart != 0) {
+            putShortMSB(s, (uInt)(strm->adler >> 16));
+            putShortMSB(s, (uInt)(strm->adler & 0xffff));
+        }
+        strm->adler = adler32(0L, Z_NULL, 0);
+        s->status = BUSY_STATE;
+
+        /* Compression must start with an empty pending buffer */
+        flush_pending(strm);
+        if (s->pending != 0) {
+            s->last_flush = -1;
+            return Z_OK;
+        }
+    }
+#ifdef GZIP
+    if (s->status == GZIP_STATE) {
+        /* gzip header */
+        strm->adler = crc32(0L, Z_NULL, 0);
+        put_byte(s, 31);
+        put_byte(s, 139);
+        put_byte(s, 8);
+        if (s->gzhead == Z_NULL) {
+            put_byte(s, 0);
+            put_byte(s, 0);
+            put_byte(s, 0);
+            put_byte(s, 0);
+            put_byte(s, 0);
+            put_byte(s, s->level == 9 ? 2 :
+                     (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
+                      4 : 0));
+            put_byte(s, OS_CODE);
             s->status = BUSY_STATE;
-            putShortMSB(s, header);
 
-            /* Save the adler32 of the preset dictionary: */
-            if (s->strstart != 0) {
-                putShortMSB(s, (uInt)(strm->adler >> 16));
-                putShortMSB(s, (uInt)(strm->adler & 0xffff));
+            /* Compression must start with an empty pending buffer */
+            flush_pending(strm);
+            if (s->pending != 0) {
+                s->last_flush = -1;
+                return Z_OK;
+            }
+        }
+        else {
+            put_byte(s, (s->gzhead->text ? 1 : 0) +
+                     (s->gzhead->hcrc ? 2 : 0) +
+                     (s->gzhead->extra == Z_NULL ? 0 : 4) +
+                     (s->gzhead->name == Z_NULL ? 0 : 8) +
+                     (s->gzhead->comment == Z_NULL ? 0 : 16)
+                     );
+            put_byte(s, (Byte)(s->gzhead->time & 0xff));
+            put_byte(s, (Byte)((s->gzhead->time >> 8) & 0xff));
+            put_byte(s, (Byte)((s->gzhead->time >> 16) & 0xff));
+            put_byte(s, (Byte)((s->gzhead->time >> 24) & 0xff));
+            put_byte(s, s->level == 9 ? 2 :
+                     (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2 ?
+                      4 : 0));
+            put_byte(s, s->gzhead->os & 0xff);
+            if (s->gzhead->extra != Z_NULL) {
+                put_byte(s, s->gzhead->extra_len & 0xff);
+                put_byte(s, (s->gzhead->extra_len >> 8) & 0xff);
             }
-            strm->adler = adler32(0L, Z_NULL, 0);
+            if (s->gzhead->hcrc)
+                strm->adler = crc32(strm->adler, s->pending_buf,
+                                    s->pending);
+            s->gzindex = 0;
+            s->status = EXTRA_STATE;
         }
     }
-#ifdef GZIP
     if (s->status == EXTRA_STATE) {
         if (s->gzhead->extra != Z_NULL) {
-            uInt beg = s->pending;  /* start of bytes to update crc */
-
-            while (s->gzindex < (s->gzhead->extra_len & 0xffff)) {
-                if (s->pending == s->pending_buf_size) {
-                    if (s->gzhead->hcrc && s->pending > beg)
-                        strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                            s->pending - beg);
-                    flush_pending(strm);
-                    beg = s->pending;
-                    if (s->pending == s->pending_buf_size)
-                        break;
+            ulg beg = s->pending;   /* start of bytes to update crc */
+            uInt left = (s->gzhead->extra_len & 0xffff) - s->gzindex;
+            while (s->pending + left > s->pending_buf_size) {
+                uInt copy = s->pending_buf_size - s->pending;
+                zmemcpy(s->pending_buf + s->pending,
+                        s->gzhead->extra + s->gzindex, copy);
+                s->pending = s->pending_buf_size;
+                HCRC_UPDATE(beg);
+                s->gzindex += copy;
+                flush_pending(strm);
+                if (s->pending != 0) {
+                    s->last_flush = -1;
+                    return Z_OK;
                 }
-                put_byte(s, s->gzhead->extra[s->gzindex]);
-                s->gzindex++;
-            }
-            if (s->gzhead->hcrc && s->pending > beg)
-                strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                    s->pending - beg);
-            if (s->gzindex == s->gzhead->extra_len) {
-                s->gzindex = 0;
-                s->status = NAME_STATE;
+                beg = 0;
+                left -= copy;
             }
+            zmemcpy(s->pending_buf + s->pending,
+                    s->gzhead->extra + s->gzindex, left);
+            s->pending += left;
+            HCRC_UPDATE(beg);
+            s->gzindex = 0;
         }
-        else
-            s->status = NAME_STATE;
+        s->status = NAME_STATE;
     }
     if (s->status == NAME_STATE) {
         if (s->gzhead->name != Z_NULL) {
-            uInt beg = s->pending;  /* start of bytes to update crc */
+            ulg beg = s->pending;   /* start of bytes to update crc */
             int val;
-
             do {
                 if (s->pending == s->pending_buf_size) {
-                    if (s->gzhead->hcrc && s->pending > beg)
-                        strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                            s->pending - beg);
+                    HCRC_UPDATE(beg);
                     flush_pending(strm);
-                    beg = s->pending;
-                    if (s->pending == s->pending_buf_size) {
-                        val = 1;
-                        break;
+                    if (s->pending != 0) {
+                        s->last_flush = -1;
+                        return Z_OK;
                     }
+                    beg = 0;
                 }
                 val = s->gzhead->name[s->gzindex++];
                 put_byte(s, val);
             } while (val != 0);
-            if (s->gzhead->hcrc && s->pending > beg)
-                strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                    s->pending - beg);
-            if (val == 0) {
-                s->gzindex = 0;
-                s->status = COMMENT_STATE;
-            }
+            HCRC_UPDATE(beg);
+            s->gzindex = 0;
         }
-        else
-            s->status = COMMENT_STATE;
+        s->status = COMMENT_STATE;
     }
     if (s->status == COMMENT_STATE) {
         if (s->gzhead->comment != Z_NULL) {
-            uInt beg = s->pending;  /* start of bytes to update crc */
+            ulg beg = s->pending;   /* start of bytes to update crc */
             int val;
-
             do {
                 if (s->pending == s->pending_buf_size) {
-                    if (s->gzhead->hcrc && s->pending > beg)
-                        strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                            s->pending - beg);
+                    HCRC_UPDATE(beg);
                     flush_pending(strm);
-                    beg = s->pending;
-                    if (s->pending == s->pending_buf_size) {
-                        val = 1;
-                        break;
+                    if (s->pending != 0) {
+                        s->last_flush = -1;
+                        return Z_OK;
                     }
+                    beg = 0;
                 }
                 val = s->gzhead->comment[s->gzindex++];
                 put_byte(s, val);
             } while (val != 0);
-            if (s->gzhead->hcrc && s->pending > beg)
-                strm->adler = crc32(strm->adler, s->pending_buf + beg,
-                                    s->pending - beg);
-            if (val == 0)
-                s->status = HCRC_STATE;
+            HCRC_UPDATE(beg);
         }
-        else
-            s->status = HCRC_STATE;
+        s->status = HCRC_STATE;
     }
     if (s->status == HCRC_STATE) {
         if (s->gzhead->hcrc) {
-            if (s->pending + 2 > s->pending_buf_size)
+            if (s->pending + 2 > s->pending_buf_size) {
                 flush_pending(strm);
-            if (s->pending + 2 <= s->pending_buf_size) {
-                put_byte(s, (Byte)(strm->adler & 0xff));
-                put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
-                strm->adler = crc32(0L, Z_NULL, 0);
-                s->status = BUSY_STATE;
+                if (s->pending != 0) {
+                    s->last_flush = -1;
+                    return Z_OK;
+                }
             }
+            put_byte(s, (Byte)(strm->adler & 0xff));
+            put_byte(s, (Byte)((strm->adler >> 8) & 0xff));
+            strm->adler = crc32(0L, Z_NULL, 0);
         }
-        else
-            s->status = BUSY_STATE;
-    }
-#endif
+        s->status = BUSY_STATE;
 
-    /* Flush as much pending output as possible */
-    if (s->pending != 0) {
+        /* Compression must start with an empty pending buffer */
         flush_pending(strm);
-        if (strm->avail_out == 0) {
-            /* Since avail_out is 0, deflate will be called again with
-             * more output space, but possibly with both pending and
-             * avail_in equal to zero. There won't be anything to do,
-             * but this is not an error situation so make sure we
-             * return OK instead of BUF_ERROR at next call of deflate:
-             */
+        if (s->pending != 0) {
             s->last_flush = -1;
             return Z_OK;
         }
-
-    /* Make sure there is something to do and avoid duplicate consecutive
-     * flushes. For repeated and useless calls with Z_FINISH, we keep
-     * returning Z_STREAM_END instead of Z_BUF_ERROR.
-     */
-    } else if (strm->avail_in == 0 && flush <= old_flush &&
-               flush != Z_FINISH) {
-        ERR_RETURN(strm, Z_BUF_ERROR);
-    }
-
-    /* User must not provide more input after the first FINISH: */
-    if (s->status == FINISH_STATE && strm->avail_in != 0) {
-        ERR_RETURN(strm, Z_BUF_ERROR);
     }
+#endif
 
     /* Start a new block or continue the current one.
      */
@@ -817,9 +1051,10 @@ int ZEXPORT deflate (strm, flush)
         (flush != Z_NO_FLUSH && s->status != FINISH_STATE)) {
         block_state bstate;
 
-        bstate = s->strategy == Z_HUFFMAN_ONLY ? deflate_huff(s, flush) :
-                    (s->strategy == Z_RLE ? deflate_rle(s, flush) :
-                        (*(configuration_table[s->level].func))(s, flush));
+        bstate = s->level == 0 ? deflate_stored(s, flush) :
+                 s->strategy == Z_HUFFMAN_ONLY ? deflate_huff(s, flush) :
+                 s->strategy == Z_RLE ? deflate_rle(s, flush) :
+                 (*(configuration_table[s->level].func))(s, flush);
 
         if (bstate == finish_started || bstate == finish_done) {
             s->status = FINISH_STATE;
@@ -850,6 +1085,7 @@ int ZEXPORT deflate (strm, flush)
                     if (s->lookahead == 0) {
                         s->strstart = 0;
                         s->block_start = 0L;
+                        s->insert = 0;
                     }
                 }
             }
@@ -860,7 +1096,6 @@ int ZEXPORT deflate (strm, flush)
             }
         }
     }
-    Assert(strm->avail_out > 0, "bug2");
 
     if (flush != Z_FINISH) return Z_OK;
     if (s->wrap <= 0) return Z_STREAM_END;
@@ -892,23 +1127,14 @@ int ZEXPORT deflate (strm, flush)
 }
 
 /* ========================================================================= */
-int ZEXPORT deflateEnd (strm)
+int ZEXPORT deflateEnd(strm)
     z_streamp strm;
 {
     int status;
 
-    if (strm == Z_NULL || strm->state == Z_NULL) return Z_STREAM_ERROR;
+    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
 
     status = strm->state->status;
-    if (status != INIT_STATE &&
-        status != EXTRA_STATE &&
-        status != NAME_STATE &&
-        status != COMMENT_STATE &&
-        status != HCRC_STATE &&
-        status != BUSY_STATE &&
-        status != FINISH_STATE) {
-      return Z_STREAM_ERROR;
-    }
 
     /* Deallocate in reverse order of allocations: */
     TRY_FREE(strm, strm->state->pending_buf);
@@ -927,7 +1153,7 @@ int ZEXPORT deflateEnd (strm)
  * To simplify the source, this is not supported for 16-bit MSDOS (which
  * doesn't have enough memory anyway to duplicate compression states).
  */
-int ZEXPORT deflateCopy (dest, source)
+int ZEXPORT deflateCopy(dest, source)
     z_streamp dest;
     z_streamp source;
 {
@@ -936,28 +1162,26 @@ int ZEXPORT deflateCopy (dest, source)
 #else
     deflate_state *ds;
     deflate_state *ss;
-    ushf *overlay;
 
 
-    if (source == Z_NULL || dest == Z_NULL || source->state == Z_NULL) {
+    if (deflateStateCheck(source) || dest == Z_NULL) {
         return Z_STREAM_ERROR;
     }
 
     ss = source->state;
 
-    zmemcpy(dest, source, sizeof(z_stream));
+    zmemcpy((voidpf)dest, (voidpf)source, sizeof(z_stream));
 
     ds = (deflate_state *) ZALLOC(dest, 1, sizeof(deflate_state));
     if (ds == Z_NULL) return Z_MEM_ERROR;
     dest->state = (struct internal_state FAR *) ds;
-    zmemcpy(ds, ss, sizeof(deflate_state));
+    zmemcpy((voidpf)ds, (voidpf)ss, sizeof(deflate_state));
     ds->strm = dest;
 
     ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte));
     ds->prev   = (Posf *)  ZALLOC(dest, ds->w_size, sizeof(Pos));
     ds->head   = (Posf *)  ZALLOC(dest, ds->hash_size, sizeof(Pos));
-    overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2);
-    ds->pending_buf = (uchf *) overlay;
+    ds->pending_buf = (uchf *) ZALLOC(dest, ds->lit_bufsize, 4);
 
     if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL ||
         ds->pending_buf == Z_NULL) {
@@ -966,13 +1190,12 @@ int ZEXPORT deflateCopy (dest, source)
     }
     /* following zmemcpy do not work for 16-bit MSDOS */
     zmemcpy(ds->window, ss->window, ds->w_size * 2 * sizeof(Byte));
-    zmemcpy(ds->prev, ss->prev, ds->w_size * sizeof(Pos));
-    zmemcpy(ds->head, ss->head, ds->hash_size * sizeof(Pos));
+    zmemcpy((voidpf)ds->prev, (voidpf)ss->prev, ds->w_size * sizeof(Pos));
+    zmemcpy((voidpf)ds->head, (voidpf)ss->head, ds->hash_size * sizeof(Pos));
     zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size);
 
     ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf);
-    ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush);
-    ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize;
+    ds->sym_buf = ds->pending_buf + ds->lit_bufsize;
 
     ds->l_desc.dyn_tree = ds->dyn_ltree;
     ds->d_desc.dyn_tree = ds->dyn_dtree;
@@ -989,7 +1212,7 @@ int ZEXPORT deflateCopy (dest, source)
  * allocating a large strm->next_in buffer and copying from it.
  * (See also flush_pending()).
  */
-local int read_buf(strm, buf, size)
+local unsigned read_buf(strm, buf, size)
     z_streamp strm;
     Bytef *buf;
     unsigned size;
@@ -1001,25 +1224,25 @@ local int read_buf(strm, buf, size)
 
     strm->avail_in  -= len;
 
+    zmemcpy(buf, strm->next_in, len);
     if (strm->state->wrap == 1) {
-        strm->adler = adler32(strm->adler, strm->next_in, len);
+        strm->adler = adler32(strm->adler, buf, len);
     }
 #ifdef GZIP
     else if (strm->state->wrap == 2) {
-        strm->adler = crc32(strm->adler, strm->next_in, len);
+        strm->adler = crc32(strm->adler, buf, len);
     }
 #endif
-    zmemcpy(buf, strm->next_in, len);
     strm->next_in  += len;
     strm->total_in += len;
 
-    return (int)len;
+    return len;
 }
 
 /* ===========================================================================
  * Initialize the "longest match" routines for a new zlib stream
  */
-local void lm_init (s)
+local void lm_init(s)
     deflate_state *s;
 {
     s->window_size = (ulg)2L*s->w_size;
@@ -1036,14 +1259,10 @@ local void lm_init (s)
     s->strstart = 0;
     s->block_start = 0L;
     s->lookahead = 0;
+    s->insert = 0;
     s->match_length = s->prev_length = MIN_MATCH-1;
     s->match_available = 0;
     s->ins_h = 0;
-#ifndef FASTEST
-#ifdef ASMV
-    match_init(); /* initialize the asm code */
-#endif
-#endif
 }
 
 #ifndef FASTEST
@@ -1056,19 +1275,15 @@ local void lm_init (s)
  *   string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
  * OUT assertion: the match length is not greater than s->lookahead.
  */
-#ifndef ASMV
-/* For 80x86 and 680x0, an optimized version will be provided in match.asm or
- * match.S. The code will be functionally equivalent.
- */
 local uInt longest_match(s, cur_match)
     deflate_state *s;
     IPos cur_match;                             /* current match */
 {
     unsigned chain_length = s->max_chain_length;/* max hash chain length */
     register Bytef *scan = s->window + s->strstart; /* current string */
-    register Bytef *match;                       /* matched string */
+    register Bytef *match;                      /* matched string */
     register int len;                           /* length of current match */
-    int best_len = s->prev_length;              /* best match length so far */
+    int best_len = (int)s->prev_length;         /* best match length so far */
     int nice_match = s->nice_match;             /* stop if match long enough */
     IPos limit = s->strstart > (IPos)MAX_DIST(s) ?
         s->strstart - (IPos)MAX_DIST(s) : NIL;
@@ -1084,10 +1299,10 @@ local uInt longest_match(s, cur_match)
      */
     register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1;
     register ush scan_start = *(ushf*)scan;
-    register ush scan_end   = *(ushf*)(scan+best_len-1);
+    register ush scan_end   = *(ushf*)(scan + best_len - 1);
 #else
     register Bytef *strend = s->window + s->strstart + MAX_MATCH;
-    register Byte scan_end1  = scan[best_len-1];
+    register Byte scan_end1  = scan[best_len - 1];
     register Byte scan_end   = scan[best_len];
 #endif
 
@@ -1103,9 +1318,10 @@ local uInt longest_match(s, cur_match)
     /* Do not look for matches beyond the end of the input. This is necessary
      * to make deflate deterministic.
      */
-    if ((uInt)nice_match > s->lookahead) nice_match = s->lookahead;
+    if ((uInt)nice_match > s->lookahead) nice_match = (int)s->lookahead;
 
-    Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
+    Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
+           "need lookahead");
 
     do {
         Assert(cur_match < s->strstart, "no future");
@@ -1123,43 +1339,44 @@ local uInt longest_match(s, cur_match)
         /* This code assumes sizeof(unsigned short) == 2. Do not use
          * UNALIGNED_OK if your compiler uses a different size.
          */
-        if (*(ushf*)(match+best_len-1) != scan_end ||
+        if (*(ushf*)(match + best_len - 1) != scan_end ||
             *(ushf*)match != scan_start) continue;
 
         /* It is not necessary to compare scan[2] and match[2] since they are
          * always equal when the other bytes match, given that the hash keys
          * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at
-         * strstart+3, +5, ... up to strstart+257. We check for insufficient
+         * strstart + 3, + 5, up to strstart + 257. We check for insufficient
          * lookahead only every 4th comparison; the 128th check will be made
-         * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is
+         * at strstart + 257. If MAX_MATCH-2 is not a multiple of 8, it is
          * necessary to put more guard bytes at the end of the window, or
          * to check more often for insufficient lookahead.
          */
         Assert(scan[2] == match[2], "scan[2]?");
         scan++, match++;
         do {
-        } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
+        } while (*(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
                  scan < strend);
         /* The funny "do {}" generates better code on most compilers */
 
-        /* Here, scan <= window+strstart+257 */
-        Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+        /* Here, scan <= window + strstart + 257 */
+        Assert(scan <= s->window + (unsigned)(s->window_size - 1),
+               "wild scan");
         if (*scan == *match) scan++;
 
-        len = (MAX_MATCH - 1) - (int)(strend-scan);
+        len = (MAX_MATCH - 1) - (int)(strend - scan);
         scan = strend - (MAX_MATCH-1);
 
 #else /* UNALIGNED_OK */
 
-        if (match[best_len]   != scan_end  ||
-            match[best_len-1] != scan_end1 ||
-            *match            != *scan     ||
-            *++match          != scan[1])      continue;
+        if (match[best_len]     != scan_end  ||
+            match[best_len - 1] != scan_end1 ||
+            *match              != *scan     ||
+            *++match            != scan[1])      continue;
 
-        /* The check at best_len-1 can be removed because it will be made
+        /* The check at best_len - 1 can be removed because it will be made
          * again later. (This heuristic is not always a win.)
          * It is not necessary to compare scan[2] and match[2] since they
          * are always equal when the other bytes match, given that
@@ -1169,7 +1386,7 @@ local uInt longest_match(s, cur_match)
         Assert(*scan == *match, "match[2]?");
 
         /* We check for insufficient lookahead only every 8th comparison;
-         * the 256th check will be made at strstart+258.
+         * the 256th check will be made at strstart + 258.
          */
         do {
         } while (*++scan == *++match && *++scan == *++match &&
@@ -1178,7 +1395,8 @@ local uInt longest_match(s, cur_match)
                  *++scan == *++match && *++scan == *++match &&
                  scan < strend);
 
-        Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+        Assert(scan <= s->window + (unsigned)(s->window_size - 1),
+               "wild scan");
 
         len = MAX_MATCH - (int)(strend - scan);
         scan = strend - MAX_MATCH;
@@ -1190,9 +1408,9 @@ local uInt longest_match(s, cur_match)
             best_len = len;
             if (len >= nice_match) break;
 #ifdef UNALIGNED_OK
-            scan_end = *(ushf*)(scan+best_len-1);
+            scan_end = *(ushf*)(scan + best_len - 1);
 #else
-            scan_end1  = scan[best_len-1];
+            scan_end1  = scan[best_len - 1];
             scan_end   = scan[best_len];
 #endif
         }
@@ -1202,7 +1420,6 @@ local uInt longest_match(s, cur_match)
     if ((uInt)best_len <= s->lookahead) return (uInt)best_len;
     return s->lookahead;
 }
-#endif /* ASMV */
 
 #else /* FASTEST */
 
@@ -1223,7 +1440,8 @@ local uInt longest_match(s, cur_match)
      */
     Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");
 
-    Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
+    Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
+           "need lookahead");
 
     Assert(cur_match < s->strstart, "no future");
 
@@ -1233,7 +1451,7 @@ local uInt longest_match(s, cur_match)
      */
     if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1;
 
-    /* The check at best_len-1 can be removed because it will be made
+    /* The check at best_len - 1 can be removed because it will be made
      * again later. (This heuristic is not always a win.)
      * It is not necessary to compare scan[2] and match[2] since they
      * are always equal when the other bytes match, given that
@@ -1243,7 +1461,7 @@ local uInt longest_match(s, cur_match)
     Assert(*scan == *match, "match[2]?");
 
     /* We check for insufficient lookahead only every 8th comparison;
-     * the 256th check will be made at strstart+258.
+     * the 256th check will be made at strstart + 258.
      */
     do {
     } while (*++scan == *++match && *++scan == *++match &&
@@ -1252,7 +1470,7 @@ local uInt longest_match(s, cur_match)
              *++scan == *++match && *++scan == *++match &&
              scan < strend);
 
-    Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+    Assert(scan <= s->window + (unsigned)(s->window_size - 1), "wild scan");
 
     len = MAX_MATCH - (int)(strend - scan);
 
@@ -1264,7 +1482,11 @@ local uInt longest_match(s, cur_match)
 
 #endif /* FASTEST */
 
-#ifdef DEBUG
+#ifdef ZLIB_DEBUG
+
+#define EQUAL 0
+/* result of memcmp for equal strings */
+
 /* ===========================================================================
  * Check that the match at match_start is indeed a match.
  */
@@ -1284,13 +1506,13 @@ local void check_match(s, start, match, length)
         z_error("invalid match");
     }
     if (z_verbose > 1) {
-        fprintf(stderr,"\\[%d,%d]", start-match, length);
+        fprintf(stderr,"\\[%d,%d]", start - match, length);
         do { putc(s->window[start++], stderr); } while (--length != 0);
     }
 }
 #else
 #  define check_match(s, start, match, length)
-#endif /* DEBUG */
+#endif /* ZLIB_DEBUG */
 
 /* ===========================================================================
  * Fill the window when the lookahead becomes insufficient.
@@ -1305,11 +1527,12 @@ local void check_match(s, start, match, length)
 local void fill_window(s)
     deflate_state *s;
 {
-    register unsigned n, m;
-    register Posf *p;
+    unsigned n;
     unsigned more;    /* Amount of free space at the end of the window. */
     uInt wsize = s->w_size;
 
+    Assert(s->lookahead < MIN_LOOKAHEAD, "already enough lookahead");
+
     do {
         more = (unsigned)(s->window_size -(ulg)s->lookahead -(ulg)s->strstart);
 
@@ -1329,40 +1552,18 @@ local void fill_window(s)
         /* If the window is almost full and there is insufficient lookahead,
          * move the upper half to the lower one to make room in the upper half.
          */
-        if (s->strstart >= wsize+MAX_DIST(s)) {
+        if (s->strstart >= wsize + MAX_DIST(s)) {
 
-            zmemcpy(s->window, s->window+wsize, (unsigned)wsize);
+            zmemcpy(s->window, s->window + wsize, (unsigned)wsize - more);
             s->match_start -= wsize;
             s->strstart    -= wsize; /* we now have strstart >= MAX_DIST */
             s->block_start -= (long) wsize;
-
-            /* Slide the hash table (could be avoided with 32 bit values
-               at the expense of memory usage). We slide even when level == 0
-               to keep the hash table consistent if we switch back to level > 0
-               later. (Using level 0 permanently is not an optimal usage of
-               zlib, so we don't care about this pathological case.)
-             */
-            n = s->hash_size;
-            p = &s->head[n];
-            do {
-                m = *--p;
-                *p = (Pos)(m >= wsize ? m-wsize : NIL);
-            } while (--n);
-
-            n = wsize;
-#ifndef FASTEST
-            p = &s->prev[n];
-            do {
-                m = *--p;
-                *p = (Pos)(m >= wsize ? m-wsize : NIL);
-                /* If n is not on any hash chain, prev[n] is garbage but
-                 * its value will never be used.
-                 */
-            } while (--n);
-#endif
+            if (s->insert > s->strstart)
+                s->insert = s->strstart;
+            slide_hash(s);
             more += wsize;
         }
-        if (s->strm->avail_in == 0) return;
+        if (s->strm->avail_in == 0) break;
 
         /* If there was no sliding:
          *    strstart <= WSIZE+MAX_DIST-1 && lookahead <= MIN_LOOKAHEAD - 1 &&
@@ -1381,12 +1582,24 @@ local void fill_window(s)
         s->lookahead += n;
 
         /* Initialize the hash value now that we have some input: */
-        if (s->lookahead >= MIN_MATCH) {
-            s->ins_h = s->window[s->strstart];
-            UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
+        if (s->lookahead + s->insert >= MIN_MATCH) {
+            uInt str = s->strstart - s->insert;
+            s->ins_h = s->window[str];
+            UPDATE_HASH(s, s->ins_h, s->window[str + 1]);
 #if MIN_MATCH != 3
             Call UPDATE_HASH() MIN_MATCH-3 more times
 #endif
+            while (s->insert) {
+                UPDATE_HASH(s, s->ins_h, s->window[str + MIN_MATCH-1]);
+#ifndef FASTEST
+                s->prev[str & s->w_mask] = s->head[s->ins_h];
+#endif
+                s->head[s->ins_h] = (Pos)str;
+                str++;
+                s->insert--;
+                if (s->lookahead + s->insert < MIN_MATCH)
+                    break;
+            }
         }
         /* If the whole input has less than MIN_MATCH bytes, ins_h is garbage,
          * but this is not important since only literal bytes will be emitted.
@@ -1427,6 +1640,9 @@ local void fill_window(s)
             s->high_water += init;
         }
     }
+
+    Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
+           "not enough room for search");
 }
 
 /* ===========================================================================
@@ -1450,64 +1666,205 @@ local void fill_window(s)
    if (s->strm->avail_out == 0) return (last) ? finish_started : need_more; \
 }
 
+/* Maximum stored block length in deflate format (not including header). */
+#define MAX_STORED 65535
+
+/* Minimum of a and b. */
+#define MIN(a, b) ((a) > (b) ? (b) : (a))
+
 /* ===========================================================================
  * Copy without compression as much as possible from the input stream, return
  * the current block state.
- * This function does not insert new strings in the dictionary since
- * uncompressible data is probably not useful. This function is used
- * only for the level=0 compression option.
- * NOTE: this function should be optimized to avoid extra copying from
- * window to pending_buf.
+ *
+ * In case deflateParams() is used to later switch to a non-zero compression
+ * level, s->matches (otherwise unused when storing) keeps track of the number
+ * of hash table slides to perform. If s->matches is 1, then one hash table
+ * slide will be done when switching. If s->matches is 2, the maximum value
+ * allowed here, then the hash table will be cleared, since two or more slides
+ * is the same as a clear.
+ *
+ * deflate_stored() is written to minimize the number of times an input byte is
+ * copied. It is most efficient with large input and output buffers, which
+ * maximizes the opportunities to have a single copy from next_in to next_out.
  */
 local block_state deflate_stored(s, flush)
     deflate_state *s;
     int flush;
 {
-    /* Stored blocks are limited to 0xffff bytes, pending_buf is limited
-     * to pending_buf_size, and each stored block has a 5 byte header:
+    /* Smallest worthy block size when not flushing or finishing. By default
+     * this is 32K. This can be as small as 507 bytes for memLevel == 1. For
+     * large input and output buffers, the stored block size will be larger.
      */
-    ulg max_block_size = 0xffff;
-    ulg max_start;
-
-    if (max_block_size > s->pending_buf_size - 5) {
-        max_block_size = s->pending_buf_size - 5;
-    }
-
-    /* Copy as much as possible from input to output: */
-    for (;;) {
-        /* Fill the window as much as possible: */
-        if (s->lookahead <= 1) {
+    unsigned min_block = MIN(s->pending_buf_size - 5, s->w_size);
 
-            Assert(s->strstart < s->w_size+MAX_DIST(s) ||
-                   s->block_start >= (long)s->w_size, "slide too late");
+    /* Copy as many min_block or larger stored blocks directly to next_out as
+     * possible. If flushing, copy the remaining available input to next_out as
+     * stored blocks, if there is enough space.
+     */
+    unsigned len, left, have, last = 0;
+    unsigned used = s->strm->avail_in;
+    do {
+        /* Set len to the maximum size block that we can copy directly with the
+         * available input data and output space. Set left to how much of that
+         * would be copied from what's left in the window.
+         */
+        len = MAX_STORED;       /* maximum deflate stored block length */
+        have = (s->bi_valid + 42) >> 3;         /* number of header bytes */
+        if (s->strm->avail_out < have)          /* need room for header */
+            break;
+            /* maximum stored block length that will fit in avail_out: */
+        have = s->strm->avail_out - have;
+        left = s->strstart - s->block_start;    /* bytes left in window */
+        if (len > (ulg)left + s->strm->avail_in)
+            len = left + s->strm->avail_in;     /* limit len to the input */
+        if (len > have)
+            len = have;                         /* limit len to the output */
+
+        /* If the stored block would be less than min_block in length, or if
+         * unable to copy all of the available input when flushing, then try
+         * copying to the window and the pending buffer instead. Also don't
+         * write an empty block when flushing -- deflate() does that.
+         */
+        if (len < min_block && ((len == 0 && flush != Z_FINISH) ||
+                                flush == Z_NO_FLUSH ||
+                                len != left + s->strm->avail_in))
+            break;
 
-            fill_window(s);
-            if (s->lookahead == 0 && flush == Z_NO_FLUSH) return need_more;
+        /* Make a dummy stored block in pending to get the header bytes,
+         * including any pending bits. This also updates the debugging counts.
+         */
+        last = flush == Z_FINISH && len == left + s->strm->avail_in ? 1 : 0;
+        _tr_stored_block(s, (char *)0, 0L, last);
+
+        /* Replace the lengths in the dummy stored block with len. */
+        s->pending_buf[s->pending - 4] = len;
+        s->pending_buf[s->pending - 3] = len >> 8;
+        s->pending_buf[s->pending - 2] = ~len;
+        s->pending_buf[s->pending - 1] = ~len >> 8;
+
+        /* Write the stored block header bytes. */
+        flush_pending(s->strm);
+
+#ifdef ZLIB_DEBUG
+        /* Update debugging counts for the data about to be copied. */
+        s->compressed_len += len << 3;
+        s->bits_sent += len << 3;
+#endif
 
-            if (s->lookahead == 0) break; /* flush the current block */
+        /* Copy uncompressed bytes from the window to next_out. */
+        if (left) {
+            if (left > len)
+                left = len;
+            zmemcpy(s->strm->next_out, s->window + s->block_start, left);
+            s->strm->next_out += left;
+            s->strm->avail_out -= left;
+            s->strm->total_out += left;
+            s->block_start += left;
+            len -= left;
         }
-        Assert(s->block_start >= 0L, "block gone");
-
-        s->strstart += s->lookahead;
-        s->lookahead = 0;
-
-        /* Emit a stored block if pending_buf will be full: */
-        max_start = s->block_start + max_block_size;
-        if (s->strstart == 0 || (ulg)s->strstart >= max_start) {
-            /* strstart == 0 is possible when wraparound on 16-bit machine */
-            s->lookahead = (uInt)(s->strstart - max_start);
-            s->strstart = (uInt)max_start;
-            FLUSH_BLOCK(s, 0);
+
+        /* Copy uncompressed bytes directly from next_in to next_out, updating
+         * the check value.
+         */
+        if (len) {
+            read_buf(s->strm, s->strm->next_out, len);
+            s->strm->next_out += len;
+            s->strm->avail_out -= len;
+            s->strm->total_out += len;
         }
-        /* Flush if we may have to slide, otherwise block_start may become
-         * negative and the data will be gone:
+    } while (last == 0);
+
+    /* Update the sliding window with the last s->w_size bytes of the copied
+     * data, or append all of the copied data to the existing window if less
+     * than s->w_size bytes were copied. Also update the number of bytes to
+     * insert in the hash tables, in the event that deflateParams() switches to
+     * a non-zero compression level.
+     */
+    used -= s->strm->avail_in;      /* number of input bytes directly copied */
+    if (used) {
+        /* If any input was used, then no unused input remains in the window,
+         * therefore s->block_start == s->strstart.
          */
-        if (s->strstart - (uInt)s->block_start >= MAX_DIST(s)) {
-            FLUSH_BLOCK(s, 0);
+        if (used >= s->w_size) {    /* supplant the previous history */
+            s->matches = 2;         /* clear hash */
+            zmemcpy(s->window, s->strm->next_in - s->w_size, s->w_size);
+            s->strstart = s->w_size;
+            s->insert = s->strstart;
+        }
+        else {
+            if (s->window_size - s->strstart <= used) {
+                /* Slide the window down. */
+                s->strstart -= s->w_size;
+                zmemcpy(s->window, s->window + s->w_size, s->strstart);
+                if (s->matches < 2)
+                    s->matches++;   /* add a pending slide_hash() */
+                if (s->insert > s->strstart)
+                    s->insert = s->strstart;
+            }
+            zmemcpy(s->window + s->strstart, s->strm->next_in - used, used);
+            s->strstart += used;
+            s->insert += MIN(used, s->w_size - s->insert);
         }
+        s->block_start = s->strstart;
+    }
+    if (s->high_water < s->strstart)
+        s->high_water = s->strstart;
+
+    /* If the last block was written to next_out, then done. */
+    if (last)
+        return finish_done;
+
+    /* If flushing and all input has been consumed, then done. */
+    if (flush != Z_NO_FLUSH && flush != Z_FINISH &&
+        s->strm->avail_in == 0 && (long)s->strstart == s->block_start)
+        return block_done;
+
+    /* Fill the window with any remaining input. */
+    have = s->window_size - s->strstart;
+    if (s->strm->avail_in > have && s->block_start >= (long)s->w_size) {
+        /* Slide the window down. */
+        s->block_start -= s->w_size;
+        s->strstart -= s->w_size;
+        zmemcpy(s->window, s->window + s->w_size, s->strstart);
+        if (s->matches < 2)
+            s->matches++;           /* add a pending slide_hash() */
+        have += s->w_size;          /* more space now */
+        if (s->insert > s->strstart)
+            s->insert = s->strstart;
+    }
+    if (have > s->strm->avail_in)
+        have = s->strm->avail_in;
+    if (have) {
+        read_buf(s->strm, s->window + s->strstart, have);
+        s->strstart += have;
+        s->insert += MIN(have, s->w_size - s->insert);
+    }
+    if (s->high_water < s->strstart)
+        s->high_water = s->strstart;
+
+    /* There was not enough avail_out to write a complete worthy or flushed
+     * stored block to next_out. Write a stored block to pending instead, if we
+     * have enough input for a worthy block, or if flushing and there is enough
+     * room for the remaining input as a stored block in the pending buffer.
+     */
+    have = (s->bi_valid + 42) >> 3;         /* number of header bytes */
+        /* maximum stored block length that will fit in pending: */
+    have = MIN(s->pending_buf_size - have, MAX_STORED);
+    min_block = MIN(have, s->w_size);
+    left = s->strstart - s->block_start;
+    if (left >= min_block ||
+        ((left || flush == Z_FINISH) && flush != Z_NO_FLUSH &&
+         s->strm->avail_in == 0 && left <= have)) {
+        len = MIN(left, have);
+        last = flush == Z_FINISH && s->strm->avail_in == 0 &&
+               len == left ? 1 : 0;
+        _tr_stored_block(s, (charf *)s->window + s->block_start, len, last);
+        s->block_start += len;
+        flush_pending(s->strm);
     }
-    FLUSH_BLOCK(s, flush == Z_FINISH);
-    return flush == Z_FINISH ? finish_done : block_done;
+
+    /* We've done all we can with the available input and output. */
+    return last ? finish_started : need_more;
 }
 
 /* ===========================================================================
@@ -1538,7 +1895,7 @@ local block_state deflate_fast(s, flush)
             if (s->lookahead == 0) break; /* flush the current block */
         }
 
-        /* Insert the string window[strstart .. strstart+2] in the
+        /* Insert the string window[strstart .. strstart + 2] in the
          * dictionary, and set hash_head to the head of the hash chain:
          */
         hash_head = NIL;
@@ -1586,7 +1943,7 @@ local block_state deflate_fast(s, flush)
                 s->strstart += s->match_length;
                 s->match_length = 0;
                 s->ins_h = s->window[s->strstart];
-                UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
+                UPDATE_HASH(s, s->ins_h, s->window[s->strstart + 1]);
 #if MIN_MATCH != 3
                 Call UPDATE_HASH() MIN_MATCH-3 more times
 #endif
@@ -1597,14 +1954,20 @@ local block_state deflate_fast(s, flush)
         } else {
             /* No match, output a literal byte */
             Tracevv((stderr,"%c", s->window[s->strstart]));
-            _tr_tally_lit (s, s->window[s->strstart], bflush);
+            _tr_tally_lit(s, s->window[s->strstart], bflush);
             s->lookahead--;
             s->strstart++;
         }
         if (bflush) FLUSH_BLOCK(s, 0);
     }
-    FLUSH_BLOCK(s, flush == Z_FINISH);
-    return flush == Z_FINISH ? finish_done : block_done;
+    s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1;
+    if (flush == Z_FINISH) {
+        FLUSH_BLOCK(s, 1);
+        return finish_done;
+    }
+    if (s->sym_next)
+        FLUSH_BLOCK(s, 0);
+    return block_done;
 }
 
 #ifndef FASTEST
@@ -1635,7 +1998,7 @@ local block_state deflate_slow(s, flush)
             if (s->lookahead == 0) break; /* flush the current block */
         }
 
-        /* Insert the string window[strstart .. strstart+2] in the
+        /* Insert the string window[strstart .. strstart + 2] in the
          * dictionary, and set hash_head to the head of the hash chain:
          */
         hash_head = NIL;
@@ -1677,17 +2040,17 @@ local block_state deflate_slow(s, flush)
             uInt max_insert = s->strstart + s->lookahead - MIN_MATCH;
             /* Do not insert strings in hash table beyond this. */
 
-            check_match(s, s->strstart-1, s->prev_match, s->prev_length);
+            check_match(s, s->strstart - 1, s->prev_match, s->prev_length);
 
-            _tr_tally_dist(s, s->strstart -1 - s->prev_match,
+            _tr_tally_dist(s, s->strstart - 1 - s->prev_match,
                            s->prev_length - MIN_MATCH, bflush);
 
             /* Insert in hash table all strings up to the end of the match.
-             * strstart-1 and strstart are already inserted. If there is not
+             * strstart - 1 and strstart are already inserted. If there is not
              * enough lookahead, the last two strings are not inserted in
              * the hash table.
              */
-            s->lookahead -= s->prev_length-1;
+            s->lookahead -= s->prev_length - 1;
             s->prev_length -= 2;
             do {
                 if (++s->strstart <= max_insert) {
@@ -1705,8 +2068,8 @@ local block_state deflate_slow(s, flush)
              * single literal. If there was a match but the current match
              * is longer, truncate the previous match to a single literal.
              */
-            Tracevv((stderr,"%c", s->window[s->strstart-1]));
-            _tr_tally_lit(s, s->window[s->strstart-1], bflush);
+            Tracevv((stderr,"%c", s->window[s->strstart - 1]));
+            _tr_tally_lit(s, s->window[s->strstart - 1], bflush);
             if (bflush) {
                 FLUSH_BLOCK_ONLY(s, 0);
             }
@@ -1724,12 +2087,18 @@ local block_state deflate_slow(s, flush)
     }
     Assert (flush != Z_NO_FLUSH, "no flush?");
     if (s->match_available) {
-        Tracevv((stderr,"%c", s->window[s->strstart-1]));
-        _tr_tally_lit(s, s->window[s->strstart-1], bflush);
+        Tracevv((stderr,"%c", s->window[s->strstart - 1]));
+        _tr_tally_lit(s, s->window[s->strstart - 1], bflush);
         s->match_available = 0;
     }
-    FLUSH_BLOCK(s, flush == Z_FINISH);
-    return flush == Z_FINISH ? finish_done : block_done;
+    s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1;
+    if (flush == Z_FINISH) {
+        FLUSH_BLOCK(s, 1);
+        return finish_done;
+    }
+    if (s->sym_next)
+        FLUSH_BLOCK(s, 0);
+    return block_done;
 }
 #endif /* FASTEST */
 
@@ -1749,11 +2118,11 @@ local block_state deflate_rle(s, flush)
     for (;;) {
         /* Make sure that we always have enough lookahead, except
          * at the end of the input file. We need MAX_MATCH bytes
-         * for the longest encodable run.
+         * for the longest run, plus one for the unrolled loop.
          */
-        if (s->lookahead < MAX_MATCH) {
+        if (s->lookahead <= MAX_MATCH) {
             fill_window(s);
-            if (s->lookahead < MAX_MATCH && flush == Z_NO_FLUSH) {
+            if (s->lookahead <= MAX_MATCH && flush == Z_NO_FLUSH) {
                 return need_more;
             }
             if (s->lookahead == 0) break; /* flush the current block */
@@ -1772,10 +2141,12 @@ local block_state deflate_rle(s, flush)
                          prev == *++scan && prev == *++scan &&
                          prev == *++scan && prev == *++scan &&
                          scan < strend);
-                s->match_length = MAX_MATCH - (int)(strend - scan);
+                s->match_length = MAX_MATCH - (uInt)(strend - scan);
                 if (s->match_length > s->lookahead)
                     s->match_length = s->lookahead;
             }
+            Assert(scan <= s->window + (uInt)(s->window_size - 1),
+                   "wild scan");
         }
 
         /* Emit match if have run of MIN_MATCH or longer, else emit literal */
@@ -1790,14 +2161,20 @@ local block_state deflate_rle(s, flush)
         } else {
             /* No match, output a literal byte */
             Tracevv((stderr,"%c", s->window[s->strstart]));
-            _tr_tally_lit (s, s->window[s->strstart], bflush);
+            _tr_tally_lit(s, s->window[s->strstart], bflush);
             s->lookahead--;
             s->strstart++;
         }
         if (bflush) FLUSH_BLOCK(s, 0);
     }
-    FLUSH_BLOCK(s, flush == Z_FINISH);
-    return flush == Z_FINISH ? finish_done : block_done;
+    s->insert = 0;
+    if (flush == Z_FINISH) {
+        FLUSH_BLOCK(s, 1);
+        return finish_done;
+    }
+    if (s->sym_next)
+        FLUSH_BLOCK(s, 0);
+    return block_done;
 }
 
 /* ===========================================================================
@@ -1824,11 +2201,17 @@ local block_state deflate_huff(s, flush)
         /* Output a literal byte */
         s->match_length = 0;
         Tracevv((stderr,"%c", s->window[s->strstart]));
-        _tr_tally_lit (s, s->window[s->strstart], bflush);
+        _tr_tally_lit(s, s->window[s->strstart], bflush);
         s->lookahead--;
         s->strstart++;
         if (bflush) FLUSH_BLOCK(s, 0);
     }
-    FLUSH_BLOCK(s, flush == Z_FINISH);
-    return flush == Z_FINISH ? finish_done : block_done;
+    s->insert = 0;
+    if (flush == Z_FINISH) {
+        FLUSH_BLOCK(s, 1);
+        return finish_done;
+    }
+    if (s->sym_next)
+        FLUSH_BLOCK(s, 0);
+    return block_done;
 }