pg_lzcompress.c

Go to the documentation of this file.

/* ----------
 * pg_lzcompress.c -
 *
 *      This is an implementation of LZ compression for PostgreSQL.
 *      It uses a simple history table and generates 2-3 byte tags
 *      capable of backward copy information for 3-273 bytes with
 *      a max offset of 4095.
 *
 *      Entry routines:
 *
 *          int32
 *          pglz_compress(const char *source, int32 slen, char *dest,
 *                        const PGLZ_Strategy *strategy);
 *
 *              source is the input data to be compressed.
 *
 *              slen is the length of the input data.
 *
 *              dest is the output area for the compressed result.
 *                  It must be at least as big as PGLZ_MAX_OUTPUT(slen).
 *
 *              strategy is a pointer to some information controlling
 *                  the compression algorithm. If NULL, the compiled
 *                  in default strategy is used.
 *
 *              The return value is the number of bytes written in the
 *              buffer dest, or -1 if compression fails; in the latter
 *              case the contents of dest are undefined.
 *
 *          int32
 *          pglz_decompress(const char *source, int32 slen, char *dest,
 *                          int32 rawsize, bool check_complete)
 *
 *              source is the compressed input.
 *
 *              slen is the length of the compressed input.
 *
 *              dest is the area where the uncompressed data will be
 *                  written to. It is the callers responsibility to
 *                  provide enough space.
 *
 *                  The data is written to buff exactly as it was handed
 *                  to pglz_compress(). No terminating zero byte is added.
 *
 *              rawsize is the length of the uncompressed data.
 *
 *              check_complete is a flag to let us know if -1 should be
 *                  returned in cases where we don't reach the end of the
 *                  source or dest buffers, or not.  This should be false
 *                  if the caller is asking for only a partial result and
 *                  true otherwise.
 *
 *              The return value is the number of bytes written in the
 *              buffer dest, or -1 if decompression fails.
 *
 *      The decompression algorithm and internal data format:
 *
 *          It is made with the compressed data itself.
 *
 *          The data representation is easiest explained by describing
 *          the process of decompression.
 *
 *          If compressed_size == rawsize, then the data
 *          is stored uncompressed as plain bytes. Thus, the decompressor
 *          simply copies rawsize bytes to the destination.
 *
 *          Otherwise the first byte tells what to do the next 8 times.
 *          We call this the control byte.
 *
 *          An unset bit in the control byte means, that one uncompressed
 *          byte follows, which is copied from input to output.
 *
 *          A set bit in the control byte means, that a tag of 2-3 bytes
 *          follows. A tag contains information to copy some bytes, that
 *          are already in the output buffer, to the current location in
 *          the output. Let's call the three tag bytes T1, T2 and T3. The
 *          position of the data to copy is coded as an offset from the
 *          actual output position.
 *
 *          The offset is in the upper nibble of T1 and in T2.
 *          The length is in the lower nibble of T1.
 *
 *          So the 16 bits of a 2 byte tag are coded as
 *
 *              7---T1--0  7---T2--0
 *              OOOO LLLL  OOOO OOOO
 *
 *          This limits the offset to 1-4095 (12 bits) and the length
 *          to 3-18 (4 bits) because 3 is always added to it. To emit
 *          a tag of 2 bytes with a length of 2 only saves one control
 *          bit. But we lose one byte in the possible length of a tag.
 *
 *          In the actual implementation, the 2 byte tag's length is
 *          limited to 3-17, because the value 0xF in the length nibble
 *          has special meaning. It means, that the next following
 *          byte (T3) has to be added to the length value of 18. That
 *          makes total limits of 1-4095 for offset and 3-273 for length.
 *
 *          Now that we have successfully decoded a tag. We simply copy
 *          the output that occurred <offset> bytes back to the current
 *          output location in the specified <length>. Thus, a
 *          sequence of 200 spaces (think about bpchar fields) could be
 *          coded in 4 bytes. One literal space and a three byte tag to
 *          copy 199 bytes with a -1 offset. Whow - that's a compression
 *          rate of 98%! Well, the implementation needs to save the
 *          original data size too, so we need another 4 bytes for it
 *          and end up with a total compression rate of 96%, what's still
 *          worth a Whow.
 *
 *      The compression algorithm
 *
 *          The following uses numbers used in the default strategy.
 *
 *          The compressor works best for attributes of a size between
 *          1K and 1M. For smaller items there's not that much chance of
 *          redundancy in the character sequence (except for large areas
 *          of identical bytes like trailing spaces) and for bigger ones
 *          our 4K maximum look-back distance is too small.
 *
 *          The compressor creates a table for lists of positions.
 *          For each input position (except the last 3), a hash key is
 *          built from the 4 next input bytes and the position remembered
 *          in the appropriate list. Thus, the table points to linked
 *          lists of likely to be at least in the first 4 characters
 *          matching strings. This is done on the fly while the input
 *          is compressed into the output area.  Table entries are only
 *          kept for the last 4096 input positions, since we cannot use
 *          back-pointers larger than that anyway.  The size of the hash
 *          table is chosen based on the size of the input - a larger table
 *          has a larger startup cost, as it needs to be initialized to
 *          zero, but reduces the number of hash collisions on long inputs.
 *
 *          For each byte in the input, its hash key (built from this
 *          byte and the next 3) is used to find the appropriate list
 *          in the table. The lists remember the positions of all bytes
 *          that had the same hash key in the past in increasing backward
 *          offset order. Now for all entries in the used lists, the
 *          match length is computed by comparing the characters from the
 *          entries position with the characters from the actual input
 *          position.
 *
 *          The compressor starts with a so called "good_match" of 128.
 *          It is a "prefer speed against compression ratio" optimizer.
 *          So if the first entry looked at already has 128 or more
 *          matching characters, the lookup stops and that position is
 *          used for the next tag in the output.
 *
 *          For each subsequent entry in the history list, the "good_match"
 *          is lowered by 10%. So the compressor will be more happy with
 *          short matches the farer it has to go back in the history.
 *          Another "speed against ratio" preference characteristic of
 *          the algorithm.
 *
 *          Thus there are 3 stop conditions for the lookup of matches:
 *
 *              - a match >= good_match is found
 *              - there are no more history entries to look at
 *              - the next history entry is already too far back
 *                to be coded into a tag.
 *
 *          Finally the match algorithm checks that at least a match
 *          of 3 or more bytes has been found, because that is the smallest
 *          amount of copy information to code into a tag. If so, a tag
 *          is omitted and all the input bytes covered by that are just
 *          scanned for the history add's, otherwise a literal character
 *          is omitted and only his history entry added.
 *
 *      Acknowledgments:
 *
 *          Many thanks to Adisak Pochanayon, who's article about SLZ
 *          inspired me to write the PostgreSQL compression this way.
 *
 *          Jan Wieck
 *
 * Copyright (c) 1999-2021, PostgreSQL Global Development Group
 *
 * src/common/pg_lzcompress.c
 * ----------
 */
#ifndef FRONTEND
#include "postgres.h"
#else
#include "postgres_fe.h"
#endif

#include <limits.h>

#include "common/pg_lzcompress.h"


/* ----------
 * Local definitions
 * ----------
 */
#define PGLZ_MAX_HISTORY_LISTS  8192    /* must be power of 2 */
#define PGLZ_HISTORY_SIZE       4096
#define PGLZ_MAX_MATCH          273


/* ----------
 * PGLZ_HistEntry -
 *
 *      Linked list for the backward history lookup
 *
 * All the entries sharing a hash key are linked in a doubly linked list.
 * This makes it easy to remove an entry when it's time to recycle it
 * (because it's more than 4K positions old).
 * ----------
 */
typedef struct PGLZ_HistEntry
{
    struct PGLZ_HistEntry *next;    /* links for my hash key's list */
    struct PGLZ_HistEntry *prev;
    int         hindex;         /* my current hash key */
    const char *pos;            /* my input position */
} PGLZ_HistEntry;


/* ----------
 * The provided standard strategies
 * ----------
 */
static const PGLZ_Strategy strategy_default_data = {
    32,                         /* Data chunks less than 32 bytes are not
                                 * compressed */
    INT_MAX,                    /* No upper limit on what we'll try to
                                 * compress */
    25,                         /* Require 25% compression rate, or not worth
                                 * it */
    1024,                       /* Give up if no compression in the first 1KB */
    128,                        /* Stop history lookup if a match of 128 bytes
                                 * is found */
    10                          /* Lower good match size by 10% at every loop
                                 * iteration */
};
const PGLZ_Strategy *const PGLZ_strategy_default = &strategy_default_data;


static const PGLZ_Strategy strategy_always_data = {
    0,                          /* Chunks of any size are compressed */
    INT_MAX,
    0,                          /* It's enough to save one single byte */
    INT_MAX,                    /* Never give up early */
    128,                        /* Stop history lookup if a match of 128 bytes
                                 * is found */
    6                           /* Look harder for a good match */
};
const PGLZ_Strategy *const PGLZ_strategy_always = &strategy_always_data;


/* ----------
 * Statically allocated work arrays for history
 * ----------
 */
static int16 hist_start[PGLZ_MAX_HISTORY_LISTS];
static PGLZ_HistEntry hist_entries[PGLZ_HISTORY_SIZE + 1];

/*
 * Element 0 in hist_entries is unused, and means 'invalid'. Likewise,
 * INVALID_ENTRY_PTR in next/prev pointers mean 'invalid'.
 */
#define INVALID_ENTRY           0
#define INVALID_ENTRY_PTR       (&hist_entries[INVALID_ENTRY])

/* ----------
 * pglz_hist_idx -
 *
 *      Computes the history table slot for the lookup by the next 4
 *      characters in the input.
 *
 * NB: because we use the next 4 characters, we are not guaranteed to
 * find 3-character matches; they very possibly will be in the wrong
 * hash list.  This seems an acceptable tradeoff for spreading out the
 * hash keys more.
 * ----------
 */
#define pglz_hist_idx(_s,_e, _mask) (                                       \
            ((((_e) - (_s)) < 4) ? (int) (_s)[0] :                          \
             (((_s)[0] << 6) ^ ((_s)[1] << 4) ^                             \
              ((_s)[2] << 2) ^ (_s)[3])) & (_mask)              \
        )


/* ----------
 * pglz_hist_add -
 *
 *      Adds a new entry to the history table.
 *
 * If _recycle is true, then we are recycling a previously used entry,
 * and must first delink it from its old hashcode's linked list.
 *
 * NOTE: beware of multiple evaluations of macro's arguments, and note that
 * _hn and _recycle are modified in the macro.
 * ----------
 */
#define pglz_hist_add(_hs,_he,_hn,_recycle,_s,_e, _mask)    \
do {                                    \
            int __hindex = pglz_hist_idx((_s),(_e), (_mask));               \
            int16 *__myhsp = &(_hs)[__hindex];                              \
            PGLZ_HistEntry *__myhe = &(_he)[_hn];                           \
            if (_recycle) {                                                 \
                if (__myhe->prev == NULL)                                   \
                    (_hs)[__myhe->hindex] = __myhe->next - (_he);           \
                else                                                        \
                    __myhe->prev->next = __myhe->next;                      \
                if (__myhe->next != NULL)                                   \
                    __myhe->next->prev = __myhe->prev;                      \
            }                                                               \
            __myhe->next = &(_he)[*__myhsp];                                \
            __myhe->prev = NULL;                                            \
            __myhe->hindex = __hindex;                                      \
            __myhe->pos  = (_s);                                            \
            /* If there was an existing entry in this hash slot, link */    \
            /* this new entry to it. However, the 0th entry in the */       \
            /* entries table is unused, so we can freely scribble on it. */ \
            /* So don't bother checking if the slot was used - we'll */     \
            /* scribble on the unused entry if it was not, but that's */    \
            /* harmless. Avoiding the branch in this critical path */       \
            /* speeds this up a little bit. */                              \
            /* if (*__myhsp != INVALID_ENTRY) */                            \
                (_he)[(*__myhsp)].prev = __myhe;                            \
            *__myhsp = _hn;                                                 \
            if (++(_hn) >= PGLZ_HISTORY_SIZE + 1) {                         \
                (_hn) = 1;                                                  \
                (_recycle) = true;                                          \
            }                                                               \
} while (0)


/* ----------
 * pglz_out_ctrl -
 *
 *      Outputs the last and allocates a new control byte if needed.
 * ----------
 */
#define pglz_out_ctrl(__ctrlp,__ctrlb,__ctrl,__buf) \
do { \
    if ((__ctrl & 0xff) == 0)                                               \
    {                                                                       \
        *(__ctrlp) = __ctrlb;                                               \
        __ctrlp = (__buf)++;                                                \
        __ctrlb = 0;                                                        \
        __ctrl = 1;                                                         \
    }                                                                       \
} while (0)


/* ----------
 * pglz_out_literal -
 *
 *      Outputs a literal byte to the destination buffer including the
 *      appropriate control bit.
 * ----------
 */
#define pglz_out_literal(_ctrlp,_ctrlb,_ctrl,_buf,_byte) \
do { \
    pglz_out_ctrl(_ctrlp,_ctrlb,_ctrl,_buf);                                \
    *(_buf)++ = (unsigned char)(_byte);                                     \
    _ctrl <<= 1;                                                            \
} while (0)


/* ----------
 * pglz_out_tag -
 *
 *      Outputs a backward reference tag of 2-4 bytes (depending on
 *      offset and length) to the destination buffer including the
 *      appropriate control bit.
 * ----------
 */
#define pglz_out_tag(_ctrlp,_ctrlb,_ctrl,_buf,_len,_off) \
do { \
    pglz_out_ctrl(_ctrlp,_ctrlb,_ctrl,_buf);                                \
    _ctrlb |= _ctrl;                                                        \
    _ctrl <<= 1;                                                            \
    if (_len > 17)                                                          \
    {                                                                       \
        (_buf)[0] = (unsigned char)((((_off) & 0xf00) >> 4) | 0x0f);        \
        (_buf)[1] = (unsigned char)(((_off) & 0xff));                       \
        (_buf)[2] = (unsigned char)((_len) - 18);                           \
        (_buf) += 3;                                                        \
    } else {                                                                \
        (_buf)[0] = (unsigned char)((((_off) & 0xf00) >> 4) | ((_len) - 3)); \
        (_buf)[1] = (unsigned char)((_off) & 0xff);                         \
        (_buf) += 2;                                                        \
    }                                                                       \
} while (0)


/* ----------
 * pglz_find_match -
 *
 *      Lookup the history table if the actual input stream matches
 *      another sequence of characters, starting somewhere earlier
 *      in the input buffer.
 * ----------
 */
static inline int
pglz_find_match(int16 *hstart, const char *input, const char *end,
                int *lenp, int *offp, int good_match, int good_drop, int mask)
{
    PGLZ_HistEntry *hent;
    int16       hentno;
    int32       len = 0;
    int32       off = 0;

    /*
     * Traverse the linked history list until a good enough match is found.
     */
    hentno = hstart[pglz_hist_idx(input, end, mask)];
    hent = &hist_entries[hentno];
    while (hent != INVALID_ENTRY_PTR)
    {
        const char *ip = input;
        const char *hp = hent->pos;
        int32       thisoff;
        int32       thislen;

        /*
         * Stop if the offset does not fit into our tag anymore.
         */
        thisoff = ip - hp;
        if (thisoff >= 0x0fff)
            break;

        /*
         * Determine length of match. A better match must be larger than the
         * best so far. And if we already have a match of 16 or more bytes,
         * it's worth the call overhead to use memcmp() to check if this match
         * is equal for the same size. After that we must fallback to
         * character by character comparison to know the exact position where
         * the diff occurred.
         */
        thislen = 0;
        if (len >= 16)
        {
            if (memcmp(ip, hp, len) == 0)
            {
                thislen = len;
                ip += len;
                hp += len;
                while (ip < end && *ip == *hp && thislen < PGLZ_MAX_MATCH)
                {
                    thislen++;
                    ip++;
                    hp++;
                }
            }
        }
        else
        {
            while (ip < end && *ip == *hp && thislen < PGLZ_MAX_MATCH)
            {
                thislen++;
                ip++;
                hp++;
            }
        }

        /*
         * Remember this match as the best (if it is)
         */
        if (thislen > len)
        {
            len = thislen;
            off = thisoff;
        }

        /*
         * Advance to the next history entry
         */
        hent = hent->next;

        /*
         * Be happy with lesser good matches the more entries we visited. But
         * no point in doing calculation if we're at end of list.
         */
        if (hent != INVALID_ENTRY_PTR)
        {
            if (len >= good_match)
                break;
            good_match -= (good_match * good_drop) / 100;
        }
    }

    /*
     * Return match information only if it results at least in one byte
     * reduction.
     */
    if (len > 2)
    {
        *lenp = len;
        *offp = off;
        return 1;
    }

    return 0;
}


/* ----------
 * pglz_compress -
 *
 *      Compresses source into dest using strategy. Returns the number of
 *      bytes written in buffer dest, or -1 if compression fails.
 * ----------
 */
int32
pglz_compress(const char *source, int32 slen, char *dest,
              const PGLZ_Strategy *strategy)
{
    unsigned char *bp = (unsigned char *) dest;
    unsigned char *bstart = bp;
    int         hist_next = 1;
    bool        hist_recycle = false;
    const char *dp = source;
    const char *dend = source + slen;
    unsigned char ctrl_dummy = 0;
    unsigned char *ctrlp = &ctrl_dummy;
    unsigned char ctrlb = 0;
    unsigned char ctrl = 0;
    bool        found_match = false;
    int32       match_len;
    int32       match_off;
    int32       good_match;
    int32       good_drop;
    int32       result_size;
    int32       result_max;
    int32       need_rate;
    int         hashsz;
    int         mask;

    /*
     * Our fallback strategy is the default.
     */
    if (strategy == NULL)
        strategy = PGLZ_strategy_default;

    /*
     * If the strategy forbids compression (at all or if source chunk size out
     * of range), fail.
     */
    if (strategy->match_size_good <= 0 ||
        slen < strategy->min_input_size ||
        slen > strategy->max_input_size)
        return -1;

    /*
     * Limit the match parameters to the supported range.
     */
    good_match = strategy->match_size_good;
    if (good_match > PGLZ_MAX_MATCH)
        good_match = PGLZ_MAX_MATCH;
    else if (good_match < 17)
        good_match = 17;

    good_drop = strategy->match_size_drop;
    if (good_drop < 0)
        good_drop = 0;
    else if (good_drop > 100)
        good_drop = 100;

    need_rate = strategy->min_comp_rate;
    if (need_rate < 0)
        need_rate = 0;
    else if (need_rate > 99)
        need_rate = 99;

    /*
     * Compute the maximum result size allowed by the strategy, namely the
     * input size minus the minimum wanted compression rate.  This had better
     * be <= slen, else we might overrun the provided output buffer.
     */
    if (slen > (INT_MAX / 100))
    {
        /* Approximate to avoid overflow */
        result_max = (slen / 100) * (100 - need_rate);
    }
    else
        result_max = (slen * (100 - need_rate)) / 100;

    /*
     * Experiments suggest that these hash sizes work pretty well. A large
     * hash table minimizes collision, but has a higher startup cost. For a
     * small input, the startup cost dominates. The table size must be a power
     * of two.
     */
    if (slen < 128)
        hashsz = 512;
    else if (slen < 256)
        hashsz = 1024;
    else if (slen < 512)
        hashsz = 2048;
    else if (slen < 1024)
        hashsz = 4096;
    else
        hashsz = 8192;
    mask = hashsz - 1;

    /*
     * Initialize the history lists to empty.  We do not need to zero the
     * hist_entries[] array; its entries are initialized as they are used.
     */
    memset(hist_start, 0, hashsz * sizeof(int16));

    /*
     * Compress the source directly into the output buffer.
     */
    while (dp < dend)
    {
        /*
         * If we already exceeded the maximum result size, fail.
         *
         * We check once per loop; since the loop body could emit as many as 4
         * bytes (a control byte and 3-byte tag), PGLZ_MAX_OUTPUT() had better
         * allow 4 slop bytes.
         */
        if (bp - bstart >= result_max)
            return -1;

        /*
         * If we've emitted more than first_success_by bytes without finding
         * anything compressible at all, fail.  This lets us fall out
         * reasonably quickly when looking at incompressible input (such as
         * pre-compressed data).
         */
        if (!found_match && bp - bstart >= strategy->first_success_by)
            return -1;

        /*
         * Try to find a match in the history
         */
        if (pglz_find_match(hist_start, dp, dend, &match_len,
                            &match_off, good_match, good_drop, mask))
        {
            /*
             * Create the tag and add history entries for all matched
             * characters.
             */
            pglz_out_tag(ctrlp, ctrlb, ctrl, bp, match_len, match_off);
            while (match_len--)
            {
                pglz_hist_add(hist_start, hist_entries,
                              hist_next, hist_recycle,
                              dp, dend, mask);
                dp++;           /* Do not do this ++ in the line above! */
                /* The macro would do it four times - Jan.  */
            }
            found_match = true;
        }
        else
        {
            /*
             * No match found. Copy one literal byte.
             */
            pglz_out_literal(ctrlp, ctrlb, ctrl, bp, *dp);
            pglz_hist_add(hist_start, hist_entries,
                          hist_next, hist_recycle,
                          dp, dend, mask);
            dp++;               /* Do not do this ++ in the line above! */
            /* The macro would do it four times - Jan.  */
        }
    }

    /*
     * Write out the last control byte and check that we haven't overrun the
     * output size allowed by the strategy.
     */
    *ctrlp = ctrlb;
    result_size = bp - bstart;
    if (result_size >= result_max)
        return -1;

    /* success */
    return result_size;
}


/* ----------
 * pglz_decompress -
 *
 *      Decompresses source into dest. Returns the number of bytes
 *      decompressed into the destination buffer, or -1 if the
 *      compressed data is corrupted.
 *
 *      If check_complete is true, the data is considered corrupted
 *      if we don't exactly fill the destination buffer.  Callers that
 *      are extracting a slice typically can't apply this check.
 * ----------
 */
int32
pglz_decompress(const char *source, int32 slen, char *dest,
                int32 rawsize, bool check_complete)
{
    const unsigned char *sp;
    const unsigned char *srcend;
    unsigned char *dp;
    unsigned char *destend;

    sp = (const unsigned char *) source;
    srcend = ((const unsigned char *) source) + slen;
    dp = (unsigned char *) dest;
    destend = dp + rawsize;

    while (sp < srcend && dp < destend)
    {
        /*
         * Read one control byte and process the next 8 items (or as many as
         * remain in the compressed input).
         */
        unsigned char ctrl = *sp++;
        int         ctrlc;

        for (ctrlc = 0; ctrlc < 8 && sp < srcend && dp < destend; ctrlc++)
        {
            if (ctrl & 1)
            {
                /*
                 * Set control bit means we must read a match tag. The match
                 * is coded with two bytes. First byte uses lower nibble to
                 * code length - 3. Higher nibble contains upper 4 bits of the
                 * offset. The next following byte contains the lower 8 bits
                 * of the offset. If the length is coded as 18, another
                 * extension tag byte tells how much longer the match really
                 * was (0-255).
                 */
                int32       len;
                int32       off;

                len = (sp[0] & 0x0f) + 3;
                off = ((sp[0] & 0xf0) << 4) | sp[1];
                sp += 2;
                if (len == 18)
                    len += *sp++;

                /*
                 * Check for corrupt data: if we fell off the end of the
                 * source, or if we obtained off = 0, we have problems.  (We
                 * must check this, else we risk an infinite loop below in the
                 * face of corrupt data.)
                 */
                if (unlikely(sp > srcend || off == 0))
                    return -1;

                /*
                 * Don't emit more data than requested.
                 */
                len = Min(len, destend - dp);

                /*
                 * Now we copy the bytes specified by the tag from OUTPUT to
                 * OUTPUT (copy len bytes from dp - off to dp).  The copied
                 * areas could overlap, so to avoid undefined behavior in
                 * memcpy(), be careful to copy only non-overlapping regions.
                 *
                 * Note that we cannot use memmove() instead, since while its
                 * behavior is well-defined, it's also not what we want.
                 */
                while (off < len)
                {
                    /*
                     * We can safely copy "off" bytes since that clearly
                     * results in non-overlapping source and destination.
                     */
                    memcpy(dp, dp - off, off);
                    len -= off;
                    dp += off;

                    /*----------
                     * This bit is less obvious: we can double "off" after
                     * each such step.  Consider this raw input:
                     *      112341234123412341234
                     * This will be encoded as 5 literal bytes "11234" and
                     * then a match tag with length 16 and offset 4.  After
                     * memcpy'ing the first 4 bytes, we will have emitted
                     *      112341234
                     * so we can double "off" to 8, then after the next step
                     * we have emitted
                     *      11234123412341234
                     * Then we can double "off" again, after which it is more
                     * than the remaining "len" so we fall out of this loop
                     * and finish with a non-overlapping copy of the
                     * remainder.  In general, a match tag with off < len
                     * implies that the decoded data has a repeat length of
                     * "off".  We can handle 1, 2, 4, etc repetitions of the
                     * repeated string per memcpy until we get to a situation
                     * where the final copy step is non-overlapping.
                     *
                     * (Another way to understand this is that we are keeping
                     * the copy source point dp - off the same throughout.)
                     *----------
                     */
                    off += off;
                }
                memcpy(dp, dp - off, len);
                dp += len;
            }
            else
            {
                /*
                 * An unset control bit means LITERAL BYTE. So we just copy
                 * one from INPUT to OUTPUT.
                 */
                *dp++ = *sp++;
            }

            /*
             * Advance the control bit
             */
            ctrl >>= 1;
        }
    }

    /*
     * If requested, check we decompressed the right amount.
     */
    if (check_complete && (dp != destend || sp != srcend))
        return -1;

    /*
     * That's it.
     */
    return (char *) dp - dest;
}


/* ----------
 * pglz_max_compressed_size -
 *
 *      Calculate the maximum compressed size for a given amount of raw data.
 *      Return the maximum size, or total compressed size if maximum size is
 *      larger than total compressed size.
 *
 * We can't use PGLZ_MAX_OUTPUT for this purpose, because that's used to size
 * the compression buffer (and abort the compression). It does not really say
 * what's the maximum compressed size for an input of a given length, and it
 * may happen that while the whole value is compressible (and thus fits into
 * PGLZ_MAX_OUTPUT nicely), the prefix is not compressible at all.
 * ----------
 */
int32
pglz_maximum_compressed_size(int32 rawsize, int32 total_compressed_size)
{
    int64       compressed_size;

    /*
     * pglz uses one control bit per byte, so if the entire desired prefix is
     * represented as literal bytes, we'll need (rawsize * 9) bits.  We care
     * about bytes though, so be sure to round up not down.
     *
     * Use int64 here to prevent overflow during calculation.
     */
    compressed_size = ((int64) rawsize * 9 + 7) / 8;

    /*
     * The above fails to account for a corner case: we could have compressed
     * data that starts with N-1 or N-2 literal bytes and then has a match tag
     * of 2 or 3 bytes.  It's therefore possible that we need to fetch 1 or 2
     * more bytes in order to have the whole match tag.  (Match tags earlier
     * in the compressed data don't cause a problem, since they should
     * represent more decompressed bytes than they occupy themselves.)
     */
    compressed_size += 2;

    /*
     * Maximum compressed size can't be larger than total compressed size.
     * (This also ensures that our result fits in int32.)
     */
    compressed_size = Min(compressed_size, total_compressed_size);

    return (int32) compressed_size;
}