dynahash.c File Reference

#include "postgres.h"
#include <limits.h>
#include "access/xact.h"
#include "common/hashfn.h"
#include "lib/ilist.h"
#include "port/pg_bitutils.h"
#include "storage/shmem.h"
#include "storage/spin.h"
#include "utils/memutils.h"

Include dependency graph for dynahash.c:

Go to the source code of this file.

Data Structures
struct	FreeListData
struct	HASHHDR
struct	HTAB

Macros
#define	DEF_SEGSIZE   256
#define	DEF_SEGSIZE_SHIFT   8 /* must be log2(DEF_SEGSIZE) */
#define	DEF_DIRSIZE   256
#define	NUM_FREELISTS   32
#define	IS_PARTITIONED(hctl)   ((hctl)->num_partitions != 0)
#define	FREELIST_IDX(hctl, hashcode)    (IS_PARTITIONED(hctl) ? (hashcode) % NUM_FREELISTS : 0)
#define	ELEMENTKEY(helem)   (((char *)(helem)) + MAXALIGN(sizeof(HASHELEMENT)))
#define	ELEMENT_FROM_KEY(key)    ((HASHELEMENT *) (((char *) (key)) - MAXALIGN(sizeof(HASHELEMENT))))
#define	MOD(x, y)   ((x) & ((y)-1))
#define	MAX_SEQ_SCANS   100

Typedefs
typedef HASHELEMENT *	HASHBUCKET
typedef HASHBUCKET *	HASHSEGMENT

Functions
static void *	DynaHashAlloc (Size size)
static HASHSEGMENT	seg_alloc (HTAB *hashp)
static bool	element_alloc (HTAB *hashp, int nelem, int freelist_idx)
static bool	dir_realloc (HTAB *hashp)
static bool	expand_table (HTAB *hashp)
static HASHBUCKET	get_hash_entry (HTAB *hashp, int freelist_idx)
static void	hdefault (HTAB *hashp)
static int	choose_nelem_alloc (Size entrysize)
static bool	init_htab (HTAB *hashp, int64 nelem)
static pg_noreturn void	hash_corrupted (HTAB *hashp)
static uint32	hash_initial_lookup (HTAB *hashp, uint32 hashvalue, HASHBUCKET **bucketptr)
static int	my_log2 (int64 num)
static int64	next_pow2_int64 (int64 num)
static int	next_pow2_int (int64 num)
static void	register_seq_scan (HTAB *hashp)
static void	deregister_seq_scan (HTAB *hashp)
static bool	has_seq_scans (HTAB *hashp)
static int	string_compare (const char *key1, const char *key2, Size keysize)
HTAB *	hash_create (const char *tabname, int64 nelem, const HASHCTL *info, int flags)
Size	hash_estimate_size (int64 num_entries, Size entrysize)
int64	hash_select_dirsize (int64 num_entries)
Size	hash_get_shared_size (HASHCTL *info, int flags)
void	hash_destroy (HTAB *hashp)
void	hash_stats (const char *caller, HTAB *hashp)
uint32	get_hash_value (HTAB *hashp, const void *keyPtr)
static uint32	calc_bucket (HASHHDR *hctl, uint32 hash_val)
void *	hash_search (HTAB *hashp, const void *keyPtr, HASHACTION action, bool *foundPtr)
void *	hash_search_with_hash_value (HTAB *hashp, const void *keyPtr, uint32 hashvalue, HASHACTION action, bool *foundPtr)
bool	hash_update_hash_key (HTAB *hashp, void *existingEntry, const void *newKeyPtr)
int64	hash_get_num_entries (HTAB *hashp)
void	hash_seq_init (HASH_SEQ_STATUS *status, HTAB *hashp)
void	hash_seq_init_with_hash_value (HASH_SEQ_STATUS *status, HTAB *hashp, uint32 hashvalue)
void *	hash_seq_search (HASH_SEQ_STATUS *status)
void	hash_seq_term (HASH_SEQ_STATUS *status)
void	hash_freeze (HTAB *hashp)
void	AtEOXact_HashTables (bool isCommit)
void	AtEOSubXact_HashTables (bool isCommit, int nestDepth)

Variables
static MemoryContext	CurrentDynaHashCxt = NULL
static HTAB *	seq_scan_tables [MAX_SEQ_SCANS]
static int	seq_scan_level [MAX_SEQ_SCANS]
static int	num_seq_scans = 0

Macro Definition Documentation

DEF_DIRSIZE

#define DEF_DIRSIZE   256

Definition at line 125 of file dynahash.c.

DEF_SEGSIZE

#define DEF_SEGSIZE   256

Definition at line 123 of file dynahash.c.

DEF_SEGSIZE_SHIFT

#define DEF_SEGSIZE_SHIFT   8 /* must be log2(DEF_SEGSIZE) */

Definition at line 124 of file dynahash.c.

ELEMENT_FROM_KEY

#define ELEMENT_FROM_KEY

(

key

)

((HASHELEMENT *) (((char *) (key)) - MAXALIGN(sizeof(HASHELEMENT))))

Definition at line 260 of file dynahash.c.

{
    Assert(MemoryContextIsValid(CurrentDynaHashCxt));
    return MemoryContextAllocExtended(CurrentDynaHashCxt, size,
                                      MCXT_ALLOC_NO_OOM);
}
 
 
/*
 * HashCompareFunc for string keys
 *
 * Because we copy keys with strlcpy(), they will be truncated at keysize-1
 * bytes, so we can only compare that many ... hence strncmp is almost but
 * not quite the right thing.
 */
static int
string_compare(const char *key1, const char *key2, Size keysize)
{
    return strncmp(key1, key2, keysize - 1);
}
 
 
/************************** CREATE ROUTINES **********************/
 
/*
 * hash_create -- create a new dynamic hash table
 *
 *  tabname: a name for the table (for debugging purposes)
 *  nelem: maximum number of elements expected
 *  *info: additional table parameters, as indicated by flags
 *  flags: bitmask indicating which parameters to take from *info
 *
 * The flags value *must* include HASH_ELEM.  (Formerly, this was nominally
 * optional, but the default keysize and entrysize values were useless.)
 * The flags value must also include exactly one of HASH_STRINGS, HASH_BLOBS,
 * or HASH_FUNCTION, to define the key hashing semantics (C strings,
 * binary blobs, or custom, respectively).  Callers specifying a custom
 * hash function will likely also want to use HASH_COMPARE, and perhaps
 * also HASH_KEYCOPY, to control key comparison and copying.
 * Another often-used flag is HASH_CONTEXT, to allocate the hash table
 * under info->hcxt rather than under TopMemoryContext; the default
 * behavior is only suitable for session-lifespan hash tables.
 * Other flags bits are special-purpose and seldom used, except for those
 * associated with shared-memory hash tables, for which see ShmemInitHash().
 *
 * Fields in *info are read only when the associated flags bit is set.
 * It is not necessary to initialize other fields of *info.
 * Neither tabname nor *info need persist after the hash_create() call.
 *
 * Note: It is deprecated for callers of hash_create() to explicitly specify
 * string_hash, tag_hash, uint32_hash, or oid_hash.  Just set HASH_STRINGS or
 * HASH_BLOBS.  Use HASH_FUNCTION only when you want something other than
 * one of these.
 *
 * Note: for a shared-memory hashtable, nelem needs to be a pretty good
 * estimate, since we can't expand the table on the fly.  But an unshared
 * hashtable can be expanded on-the-fly, so it's better for nelem to be
 * on the small side and let the table grow if it's exceeded.  An overly
 * large nelem will penalize hash_seq_search speed without buying much.
 */
HTAB *
hash_create(const char *tabname, int64 nelem, const HASHCTL *info, int flags)
{
    HTAB       *hashp;
    HASHHDR    *hctl;
 
    /*
     * Hash tables now allocate space for key and data, but you have to say
     * how much space to allocate.
     */
    Assert(flags & HASH_ELEM);
    Assert(info->keysize > 0);
    Assert(info->entrysize >= info->keysize);
 
    /*
     * For shared hash tables, we have a local hash header (HTAB struct) that
     * we allocate in TopMemoryContext; all else is in shared memory.
     *
     * For non-shared hash tables, everything including the hash header is in
     * a memory context created specially for the hash table --- this makes
     * hash_destroy very simple.  The memory context is made a child of either
     * a context specified by the caller, or TopMemoryContext if nothing is
     * specified.
     */
    if (flags & HASH_SHARED_MEM)
    {
        /* Set up to allocate the hash header */
        CurrentDynaHashCxt = TopMemoryContext;
    }
    else
    {
        /* Create the hash table's private memory context */
        if (flags & HASH_CONTEXT)
            CurrentDynaHashCxt = info->hcxt;
        else
            CurrentDynaHashCxt = TopMemoryContext;
        CurrentDynaHashCxt = AllocSetContextCreate(CurrentDynaHashCxt,
                                                   "dynahash",
                                                   ALLOCSET_DEFAULT_SIZES);
    }
 
    /* Initialize the hash header, plus a copy of the table name */
    hashp = (HTAB *) MemoryContextAlloc(CurrentDynaHashCxt,
                                        sizeof(HTAB) + strlen(tabname) + 1);
    MemSet(hashp, 0, sizeof(HTAB));
 
    hashp->tabname = (char *) (hashp + 1);
    strcpy(hashp->tabname, tabname);
 
    /* If we have a private context, label it with hashtable's name */
    if (!(flags & HASH_SHARED_MEM))
        MemoryContextSetIdentifier(CurrentDynaHashCxt, hashp->tabname);
 
    /*
     * Select the appropriate hash function (see comments at head of file).
     */
    if (flags & HASH_FUNCTION)
    {
        Assert(!(flags & (HASH_BLOBS | HASH_STRINGS)));
        hashp->hash = info->hash;
    }
    else if (flags & HASH_BLOBS)
    {
        Assert(!(flags & HASH_STRINGS));
        /* We can optimize hashing for common key sizes */
        if (info->keysize == sizeof(uint32))
            hashp->hash = uint32_hash;
        else
            hashp->hash = tag_hash;
    }
    else
    {
        /*
         * string_hash used to be considered the default hash method, and in a
         * non-assert build it effectively still is.  But we now consider it
         * an assertion error to not say HASH_STRINGS explicitly.  To help
         * catch mistaken usage of HASH_STRINGS, we also insist on a
         * reasonably long string length: if the keysize is only 4 or 8 bytes,
         * it's almost certainly an integer or pointer not a string.
         */
        Assert(flags & HASH_STRINGS);
        Assert(info->keysize > 8);
 
        hashp->hash = string_hash;
    }
 
    /*
     * If you don't specify a match function, it defaults to string_compare if
     * you used string_hash, and to memcmp otherwise.
     *
     * Note: explicitly specifying string_hash is deprecated, because this
     * might not work for callers in loadable modules on some platforms due to
     * referencing a trampoline instead of the string_hash function proper.
     * Specify HASH_STRINGS instead.
     */
    if (flags & HASH_COMPARE)
        hashp->match = info->match;
    else if (hashp->hash == string_hash)
        hashp->match = (HashCompareFunc) string_compare;
    else
        hashp->match = memcmp;
 
    /*
     * Similarly, the key-copying function defaults to strlcpy or memcpy.
     */
    if (flags & HASH_KEYCOPY)
        hashp->keycopy = info->keycopy;
    else if (hashp->hash == string_hash)
    {
        /*
         * The signature of keycopy is meant for memcpy(), which returns
         * void*, but strlcpy() returns size_t.  Since we never use the return
         * value of keycopy, and size_t is pretty much always the same size as
         * void *, this should be safe.  The extra cast in the middle is to
         * avoid warnings from -Wcast-function-type.
         */
        hashp->keycopy = (HashCopyFunc) (pg_funcptr_t) strlcpy;
    }
    else
        hashp->keycopy = memcpy;
 
    /* And select the entry allocation function, too. */
    if (flags & HASH_ALLOC)
        hashp->alloc = info->alloc;
    else
        hashp->alloc = DynaHashAlloc;
 
    if (flags & HASH_SHARED_MEM)
    {
        /*
         * ctl structure and directory are preallocated for shared memory
         * tables.  Note that HASH_DIRSIZE and HASH_ALLOC had better be set as
         * well.
         */
        hashp->hctl = info->hctl;
        hashp->dir = (HASHSEGMENT *) (((char *) info->hctl) + sizeof(HASHHDR));
        hashp->hcxt = NULL;
        hashp->isshared = true;
 
        /* hash table already exists, we're just attaching to it */
        if (flags & HASH_ATTACH)
        {
            /* make local copies of some heavily-used values */
            hctl = hashp->hctl;
            hashp->keysize = hctl->keysize;
            hashp->ssize = hctl->ssize;
            hashp->sshift = hctl->sshift;
 
            return hashp;
        }
    }
    else
    {
        /* setup hash table defaults */
        hashp->hctl = NULL;
        hashp->dir = NULL;
        hashp->hcxt = CurrentDynaHashCxt;
        hashp->isshared = false;
    }
 
    if (!hashp->hctl)
    {
        hashp->hctl = (HASHHDR *) hashp->alloc(sizeof(HASHHDR));
        if (!hashp->hctl)
            ereport(ERROR,
                    (errcode(ERRCODE_OUT_OF_MEMORY),
                     errmsg("out of memory")));
    }
 
    hashp->frozen = false;
 
    hdefault(hashp);
 
    hctl = hashp->hctl;
 
    if (flags & HASH_PARTITION)
    {
        /* Doesn't make sense to partition a local hash table */
        Assert(flags & HASH_SHARED_MEM);
 
        /*
         * The number of partitions had better be a power of 2. Also, it must
         * be less than INT_MAX (see init_htab()), so call the int version of
         * next_pow2.
         */
        Assert(info->num_partitions == next_pow2_int(info->num_partitions));
 
        hctl->num_partitions = info->num_partitions;
    }
 
    if (flags & HASH_SEGMENT)
    {
        hctl->ssize = info->ssize;
        hctl->sshift = my_log2(info->ssize);
        /* ssize had better be a power of 2 */
        Assert(hctl->ssize == (1L << hctl->sshift));
    }
 
    /*
     * SHM hash tables have fixed directory size passed by the caller.
     */
    if (flags & HASH_DIRSIZE)
    {
        hctl->max_dsize = info->max_dsize;
        hctl->dsize = info->dsize;
    }
 
    /* remember the entry sizes, too */
    hctl->keysize = info->keysize;
    hctl->entrysize = info->entrysize;
 
    /* make local copies of heavily-used constant fields */
    hashp->keysize = hctl->keysize;
    hashp->ssize = hctl->ssize;
    hashp->sshift = hctl->sshift;
 
    /* Build the hash directory structure */
    if (!init_htab(hashp, nelem))
        elog(ERROR, "failed to initialize hash table \"%s\"", hashp->tabname);
 
    /*
     * For a shared hash table, preallocate the requested number of elements.
     * This reduces problems with run-time out-of-shared-memory conditions.
     *
     * For a non-shared hash table, preallocate the requested number of
     * elements if it's less than our chosen nelem_alloc.  This avoids wasting
     * space if the caller correctly estimates a small table size.
     */
    if ((flags & HASH_SHARED_MEM) ||
        nelem < hctl->nelem_alloc)
    {
        int         i,
                    freelist_partitions,
                    nelem_alloc,
                    nelem_alloc_first;
 
        /*
         * If hash table is partitioned, give each freelist an equal share of
         * the initial allocation.  Otherwise only freeList[0] is used.
         */
        if (IS_PARTITIONED(hashp->hctl))
            freelist_partitions = NUM_FREELISTS;
        else
            freelist_partitions = 1;
 
        nelem_alloc = nelem / freelist_partitions;
        if (nelem_alloc <= 0)
            nelem_alloc = 1;
 
        /*
         * Make sure we'll allocate all the requested elements; freeList[0]
         * gets the excess if the request isn't divisible by NUM_FREELISTS.
         */
        if (nelem_alloc * freelist_partitions < nelem)
            nelem_alloc_first =
                nelem - nelem_alloc * (freelist_partitions - 1);
        else
            nelem_alloc_first = nelem_alloc;
 
        for (i = 0; i < freelist_partitions; i++)
        {
            int         temp = (i == 0) ? nelem_alloc_first : nelem_alloc;
 
            if (!element_alloc(hashp, temp, i))
                ereport(ERROR,
                        (errcode(ERRCODE_OUT_OF_MEMORY),
                         errmsg("out of memory")));
        }
    }
 
    /* Set isfixed if requested, but not till after we build initial entries */
    if (flags & HASH_FIXED_SIZE)
        hctl->isfixed = true;
 
    return hashp;
}
 
/*
 * Set default HASHHDR parameters.
 */
static void
hdefault(HTAB *hashp)
{
    HASHHDR    *hctl = hashp->hctl;
 
    MemSet(hctl, 0, sizeof(HASHHDR));
 
    hctl->dsize = DEF_DIRSIZE;
    hctl->nsegs = 0;
 
    hctl->num_partitions = 0;   /* not partitioned */
 
    /* table has no fixed maximum size */
    hctl->max_dsize = NO_MAX_DSIZE;
 
    hctl->ssize = DEF_SEGSIZE;
    hctl->sshift = DEF_SEGSIZE_SHIFT;
 
    hctl->isfixed = false;      /* can be enlarged */
 
#ifdef HASH_STATISTICS
    hctl->accesses = hctl->collisions = hctl->expansions = 0;
#endif
}
 
/*
 * Given the user-specified entry size, choose nelem_alloc, ie, how many
 * elements to add to the hash table when we need more.
 */
static int
choose_nelem_alloc(Size entrysize)
{
    int         nelem_alloc;
    Size        elementSize;
    Size        allocSize;
 
    /* Each element has a HASHELEMENT header plus user data. */
    /* NB: this had better match element_alloc() */
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(entrysize);
 
    /*
     * The idea here is to choose nelem_alloc at least 32, but round up so
     * that the allocation request will be a power of 2 or just less. This
     * makes little difference for hash tables in shared memory, but for hash
     * tables managed by palloc, the allocation request will be rounded up to
     * a power of 2 anyway.  If we fail to take this into account, we'll waste
     * as much as half the allocated space.
     */
    allocSize = 32 * 4;         /* assume elementSize at least 8 */
    do
    {
        allocSize <<= 1;
        nelem_alloc = allocSize / elementSize;
    } while (nelem_alloc < 32);
 
    return nelem_alloc;
}
 
/*
 * Compute derived fields of hctl and build the initial directory/segment
 * arrays
 */
static bool
init_htab(HTAB *hashp, int64 nelem)
{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT *segp;
    int         nbuckets;
    int         nsegs;
    int         i;
 
    /*
     * initialize mutexes if it's a partitioned table
     */
    if (IS_PARTITIONED(hctl))
        for (i = 0; i < NUM_FREELISTS; i++)
            SpinLockInit(&(hctl->freeList[i].mutex));
 
    /*
     * Allocate space for the next greater power of two number of buckets,
     * assuming a desired maximum load factor of 1.
     */
    nbuckets = next_pow2_int(nelem);
 
    /*
     * In a partitioned table, nbuckets must be at least equal to
     * num_partitions; were it less, keys with apparently different partition
     * numbers would map to the same bucket, breaking partition independence.
     * (Normally nbuckets will be much bigger; this is just a safety check.)
     */
    while (nbuckets < hctl->num_partitions)
        nbuckets <<= 1;
 
    hctl->max_bucket = hctl->low_mask = nbuckets - 1;
    hctl->high_mask = (nbuckets << 1) - 1;
 
    /*
     * Figure number of directory segments needed, round up to a power of 2
     */
    nsegs = (nbuckets - 1) / hctl->ssize + 1;
    nsegs = next_pow2_int(nsegs);
 
    /*
     * Make sure directory is big enough. If pre-allocated directory is too
     * small, choke (caller screwed up).
     */
    if (nsegs > hctl->dsize)
    {
        if (!(hashp->dir))
            hctl->dsize = nsegs;
        else
            return false;
    }
 
    /* Allocate a directory */
    if (!(hashp->dir))
    {
        CurrentDynaHashCxt = hashp->hcxt;
        hashp->dir = (HASHSEGMENT *)
            hashp->alloc(hctl->dsize * sizeof(HASHSEGMENT));
        if (!hashp->dir)
            return false;
    }
 
    /* Allocate initial segments */
    for (segp = hashp->dir; hctl->nsegs < nsegs; hctl->nsegs++, segp++)
    {
        *segp = seg_alloc(hashp);
        if (*segp == NULL)
            return false;
    }
 
    /* Choose number of entries to allocate at a time */
    hctl->nelem_alloc = choose_nelem_alloc(hctl->entrysize);
 
    return true;
}
 
/*
 * Estimate the space needed for a hashtable containing the given number
 * of entries of given size.
 * NOTE: this is used to estimate the footprint of hashtables in shared
 * memory; therefore it does not count HTAB which is in local memory.
 * NB: assumes that all hash structure parameters have default values!
 */
Size
hash_estimate_size(int64 num_entries, Size entrysize)
{
    Size        size;
    int64       nBuckets,
                nSegments,
                nDirEntries,
                nElementAllocs,
                elementSize,
                elementAllocCnt;
 
    /* estimate number of buckets wanted */
    nBuckets = next_pow2_int64(num_entries);
    /* # of segments needed for nBuckets */
    nSegments = next_pow2_int64((nBuckets - 1) / DEF_SEGSIZE + 1);
    /* directory entries */
    nDirEntries = DEF_DIRSIZE;
    while (nDirEntries < nSegments)
        nDirEntries <<= 1;      /* dir_alloc doubles dsize at each call */
 
    /* fixed control info */
    size = MAXALIGN(sizeof(HASHHDR));   /* but not HTAB, per above */
    /* directory */
    size = add_size(size, mul_size(nDirEntries, sizeof(HASHSEGMENT)));
    /* segments */
    size = add_size(size, mul_size(nSegments,
                                   MAXALIGN(DEF_SEGSIZE * sizeof(HASHBUCKET))));
    /* elements --- allocated in groups of choose_nelem_alloc() entries */
    elementAllocCnt = choose_nelem_alloc(entrysize);
    nElementAllocs = (num_entries - 1) / elementAllocCnt + 1;
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(entrysize);
    size = add_size(size,
                    mul_size(nElementAllocs,
                             mul_size(elementAllocCnt, elementSize)));
 
    return size;
}
 
/*
 * Select an appropriate directory size for a hashtable with the given
 * maximum number of entries.
 * This is only needed for hashtables in shared memory, whose directories
 * cannot be expanded dynamically.
 * NB: assumes that all hash structure parameters have default values!
 *
 * XXX this had better agree with the behavior of init_htab()...
 */
int64
hash_select_dirsize(int64 num_entries)
{
    int64       nBuckets,
                nSegments,
                nDirEntries;
 
    /* estimate number of buckets wanted */
    nBuckets = next_pow2_int64(num_entries);
    /* # of segments needed for nBuckets */
    nSegments = next_pow2_int64((nBuckets - 1) / DEF_SEGSIZE + 1);
    /* directory entries */
    nDirEntries = DEF_DIRSIZE;
    while (nDirEntries < nSegments)
        nDirEntries <<= 1;      /* dir_alloc doubles dsize at each call */
 
    return nDirEntries;
}
 
/*
 * Compute the required initial memory allocation for a shared-memory
 * hashtable with the given parameters.  We need space for the HASHHDR
 * and for the (non expansible) directory.
 */
Size
hash_get_shared_size(HASHCTL *info, int flags)
{
    Assert(flags & HASH_DIRSIZE);
    Assert(info->dsize == info->max_dsize);
    return sizeof(HASHHDR) + info->dsize * sizeof(HASHSEGMENT);
}
 
 
/********************** DESTROY ROUTINES ************************/
 
void
hash_destroy(HTAB *hashp)
{
    if (hashp != NULL)
    {
        /* allocation method must be one we know how to free, too */
        Assert(hashp->alloc == DynaHashAlloc);
        /* so this hashtable must have its own context */
        Assert(hashp->hcxt != NULL);
 
        hash_stats(__func__, hashp);
 
        /*
         * Free everything by destroying the hash table's memory context.
         */
        MemoryContextDelete(hashp->hcxt);
    }
}
 
void
hash_stats(const char *caller, HTAB *hashp)
{
#ifdef HASH_STATISTICS
    HASHHDR    *hctl = hashp->hctl;
 
    elog(DEBUG4,
         "hash_stats:  Caller: %s  Table Name: \"%s\"  Accesses: " UINT64_FORMAT "  Collisions: " UINT64_FORMAT "  Expansions: " UINT64_FORMAT "  Entries: " INT64_FORMAT "  Key Size: %zu  Max Bucket: %u  Segment Count: " INT64_FORMAT,
         caller != NULL ? caller : "(unknown)", hashp->tabname, hctl->accesses,
         hctl->collisions, hctl->expansions, hash_get_num_entries(hashp),
         hctl->keysize, hctl->max_bucket, hctl->nsegs);
#endif
}
 
/*******************************SEARCH ROUTINES *****************************/
 
 
/*
 * get_hash_value -- exported routine to calculate a key's hash value
 *
 * We export this because for partitioned tables, callers need to compute
 * the partition number (from the low-order bits of the hash value) before
 * searching.
 */
uint32
get_hash_value(HTAB *hashp, const void *keyPtr)
{
    return hashp->hash(keyPtr, hashp->keysize);
}
 
/* Convert a hash value to a bucket number */
static inline uint32
calc_bucket(HASHHDR *hctl, uint32 hash_val)
{
    uint32      bucket;
 
    bucket = hash_val & hctl->high_mask;
    if (bucket > hctl->max_bucket)
        bucket = bucket & hctl->low_mask;
 
    return bucket;
}
 
/*
 * hash_search -- look up key in table and perform action
 * hash_search_with_hash_value -- same, with key's hash value already computed
 *
 * action is one of:
 *      HASH_FIND: look up key in table
 *      HASH_ENTER: look up key in table, creating entry if not present
 *      HASH_ENTER_NULL: same, but return NULL if out of memory
 *      HASH_REMOVE: look up key in table, remove entry if present
 *
 * Return value is a pointer to the element found/entered/removed if any,
 * or NULL if no match was found.  (NB: in the case of the REMOVE action,
 * the result is a dangling pointer that shouldn't be dereferenced!)
 *
 * HASH_ENTER will normally ereport a generic "out of memory" error if
 * it is unable to create a new entry.  The HASH_ENTER_NULL operation is
 * the same except it will return NULL if out of memory.
 *
 * If foundPtr isn't NULL, then *foundPtr is set true if we found an
 * existing entry in the table, false otherwise.  This is needed in the
 * HASH_ENTER case, but is redundant with the return value otherwise.
 *
 * For hash_search_with_hash_value, the hashvalue parameter must have been
 * calculated with get_hash_value().
 */
void *
hash_search(HTAB *hashp,
            const void *keyPtr,
            HASHACTION action,
            bool *foundPtr)
{
    return hash_search_with_hash_value(hashp,
                                       keyPtr,
                                       hashp->hash(keyPtr, hashp->keysize),
                                       action,
                                       foundPtr);
}
 
void *
hash_search_with_hash_value(HTAB *hashp,
                            const void *keyPtr,
                            uint32 hashvalue,
                            HASHACTION action,
                            bool *foundPtr)
{
    HASHHDR    *hctl = hashp->hctl;
    int         freelist_idx = FREELIST_IDX(hctl, hashvalue);
    Size        keysize;
    HASHBUCKET  currBucket;
    HASHBUCKET *prevBucketPtr;
    HashCompareFunc match;
 
#ifdef HASH_STATISTICS
    hctl->accesses++;
#endif
 
    /*
     * If inserting, check if it is time to split a bucket.
     *
     * NOTE: failure to expand table is not a fatal error, it just means we
     * have to run at higher fill factor than we wanted.  However, if we're
     * using the palloc allocator then it will throw error anyway on
     * out-of-memory, so we must do this before modifying the table.
     */
    if (action == HASH_ENTER || action == HASH_ENTER_NULL)
    {
        /*
         * Can't split if running in partitioned mode, nor if frozen, nor if
         * table is the subject of any active hash_seq_search scans.
         */
        if (hctl->freeList[0].nentries > (int64) hctl->max_bucket &&
            !IS_PARTITIONED(hctl) && !hashp->frozen &&
            !has_seq_scans(hashp))
            (void) expand_table(hashp);
    }
 
    /*
     * Do the initial lookup
     */
    (void) hash_initial_lookup(hashp, hashvalue, &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    /*
     * Follow collision chain looking for matching key
     */
    match = hashp->match;       /* save one fetch in inner loop */
    keysize = hashp->keysize;   /* ditto */
 
    while (currBucket != NULL)
    {
        if (currBucket->hashvalue == hashvalue &&
            match(ELEMENTKEY(currBucket), keyPtr, keysize) == 0)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
#ifdef HASH_STATISTICS
        hctl->collisions++;
#endif
    }
 
    if (foundPtr)
        *foundPtr = (bool) (currBucket != NULL);
 
    /*
     * OK, now what?
     */
    switch (action)
    {
        case HASH_FIND:
            if (currBucket != NULL)
                return ELEMENTKEY(currBucket);
            return NULL;
 
        case HASH_REMOVE:
            if (currBucket != NULL)
            {
                /* if partitioned, must lock to touch nentries and freeList */
                if (IS_PARTITIONED(hctl))
                    SpinLockAcquire(&(hctl->freeList[freelist_idx].mutex));
 
                /* delete the record from the appropriate nentries counter. */
                Assert(hctl->freeList[freelist_idx].nentries > 0);
                hctl->freeList[freelist_idx].nentries--;
 
                /* remove record from hash bucket's chain. */
                *prevBucketPtr = currBucket->link;
 
                /* add the record to the appropriate freelist. */
                currBucket->link = hctl->freeList[freelist_idx].freeList;
                hctl->freeList[freelist_idx].freeList = currBucket;
 
                if (IS_PARTITIONED(hctl))
                    SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
                /*
                 * better hope the caller is synchronizing access to this
                 * element, because someone else is going to reuse it the next
                 * time something is added to the table
                 */
                return ELEMENTKEY(currBucket);
            }
            return NULL;
 
        case HASH_ENTER:
        case HASH_ENTER_NULL:
            /* Return existing element if found, else create one */
            if (currBucket != NULL)
                return ELEMENTKEY(currBucket);
 
            /* disallow inserts if frozen */
            if (hashp->frozen)
                elog(ERROR, "cannot insert into frozen hashtable \"%s\"",
                     hashp->tabname);
 
            currBucket = get_hash_entry(hashp, freelist_idx);
            if (currBucket == NULL)
            {
                /* out of memory */
                if (action == HASH_ENTER_NULL)
                    return NULL;
                /* report a generic message */
                if (hashp->isshared)
                    ereport(ERROR,
                            (errcode(ERRCODE_OUT_OF_MEMORY),
                             errmsg("out of shared memory")));
                else
                    ereport(ERROR,
                            (errcode(ERRCODE_OUT_OF_MEMORY),
                             errmsg("out of memory")));
            }
 
            /* link into hashbucket chain */
            *prevBucketPtr = currBucket;
            currBucket->link = NULL;
 
            /* copy key into record */
            currBucket->hashvalue = hashvalue;
            hashp->keycopy(ELEMENTKEY(currBucket), keyPtr, keysize);
 
            /*
             * Caller is expected to fill the data field on return.  DO NOT
             * insert any code that could possibly throw error here, as doing
             * so would leave the table entry incomplete and hence corrupt the
             * caller's data structure.
             */
 
            return ELEMENTKEY(currBucket);
    }
 
    elog(ERROR, "unrecognized hash action code: %d", (int) action);
 
    return NULL;                /* keep compiler quiet */
}
 
/*
 * hash_update_hash_key -- change the hash key of an existing table entry
 *
 * This is equivalent to removing the entry, making a new entry, and copying
 * over its data, except that the entry never goes to the table's freelist.
 * Therefore this cannot suffer an out-of-memory failure, even if there are
 * other processes operating in other partitions of the hashtable.
 *
 * Returns true if successful, false if the requested new hash key is already
 * present.  Throws error if the specified entry pointer isn't actually a
 * table member.
 *
 * NB: currently, there is no special case for old and new hash keys being
 * identical, which means we'll report false for that situation.  This is
 * preferable for existing uses.
 *
 * NB: for a partitioned hashtable, caller must hold lock on both relevant
 * partitions, if the new hash key would belong to a different partition.
 */
bool
hash_update_hash_key(HTAB *hashp,
                     void *existingEntry,
                     const void *newKeyPtr)
{
    HASHELEMENT *existingElement = ELEMENT_FROM_KEY(existingEntry);
    uint32      newhashvalue;
    Size        keysize;
    uint32      bucket;
    uint32      newbucket;
    HASHBUCKET  currBucket;
    HASHBUCKET *prevBucketPtr;
    HASHBUCKET *oldPrevPtr;
    HashCompareFunc match;
 
#ifdef HASH_STATISTICS
    HASHHDR    *hctl = hashp->hctl;
 
    hctl->accesses++;
#endif
 
    /* disallow updates if frozen */
    if (hashp->frozen)
        elog(ERROR, "cannot update in frozen hashtable \"%s\"",
             hashp->tabname);
 
    /*
     * Lookup the existing element using its saved hash value.  We need to do
     * this to be able to unlink it from its hash chain, but as a side benefit
     * we can verify the validity of the passed existingEntry pointer.
     */
    bucket = hash_initial_lookup(hashp, existingElement->hashvalue,
                                 &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    while (currBucket != NULL)
    {
        if (currBucket == existingElement)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
    }
 
    if (currBucket == NULL)
        elog(ERROR, "hash_update_hash_key argument is not in hashtable \"%s\"",
             hashp->tabname);
 
    oldPrevPtr = prevBucketPtr;
 
    /*
     * Now perform the equivalent of a HASH_ENTER operation to locate the hash
     * chain we want to put the entry into.
     */
    newhashvalue = hashp->hash(newKeyPtr, hashp->keysize);
    newbucket = hash_initial_lookup(hashp, newhashvalue, &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    /*
     * Follow collision chain looking for matching key
     */
    match = hashp->match;       /* save one fetch in inner loop */
    keysize = hashp->keysize;   /* ditto */
 
    while (currBucket != NULL)
    {
        if (currBucket->hashvalue == newhashvalue &&
            match(ELEMENTKEY(currBucket), newKeyPtr, keysize) == 0)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
#ifdef HASH_STATISTICS
        hctl->collisions++;
#endif
    }
 
    if (currBucket != NULL)
        return false;           /* collision with an existing entry */
 
    currBucket = existingElement;
 
    /*
     * If old and new hash values belong to the same bucket, we need not
     * change any chain links, and indeed should not since this simplistic
     * update will corrupt the list if currBucket is the last element.  (We
     * cannot fall out earlier, however, since we need to scan the bucket to
     * check for duplicate keys.)
     */
    if (bucket != newbucket)
    {
        /* OK to remove record from old hash bucket's chain. */
        *oldPrevPtr = currBucket->link;
 
        /* link into new hashbucket chain */
        *prevBucketPtr = currBucket;
        currBucket->link = NULL;
    }
 
    /* copy new key into record */
    currBucket->hashvalue = newhashvalue;
    hashp->keycopy(ELEMENTKEY(currBucket), newKeyPtr, keysize);
 
    /* rest of record is untouched */
 
    return true;
}
 
/*
 * Allocate a new hashtable entry if possible; return NULL if out of memory.
 * (Or, if the underlying space allocator throws error for out-of-memory,
 * we won't return at all.)
 */
static HASHBUCKET
get_hash_entry(HTAB *hashp, int freelist_idx)
{
    HASHHDR    *hctl = hashp->hctl;
    HASHBUCKET  newElement;
 
    for (;;)
    {
        /* if partitioned, must lock to touch nentries and freeList */
        if (IS_PARTITIONED(hctl))
            SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
 
        /* try to get an entry from the freelist */
        newElement = hctl->freeList[freelist_idx].freeList;
 
        if (newElement != NULL)
            break;
 
        if (IS_PARTITIONED(hctl))
            SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
        /*
         * No free elements in this freelist.  In a partitioned table, there
         * might be entries in other freelists, but to reduce contention we
         * prefer to first try to get another chunk of buckets from the main
         * shmem allocator.  If that fails, though, we *MUST* root through all
         * the other freelists before giving up.  There are multiple callers
         * that assume that they can allocate every element in the initially
         * requested table size, or that deleting an element guarantees they
         * can insert a new element, even if shared memory is entirely full.
         * Failing because the needed element is in a different freelist is
         * not acceptable.
         */
        if (!element_alloc(hashp, hctl->nelem_alloc, freelist_idx))
        {
            int         borrow_from_idx;
 
            if (!IS_PARTITIONED(hctl))
                return NULL;    /* out of memory */
 
            /* try to borrow element from another freelist */
            borrow_from_idx = freelist_idx;
            for (;;)
            {
                borrow_from_idx = (borrow_from_idx + 1) % NUM_FREELISTS;
                if (borrow_from_idx == freelist_idx)
                    break;      /* examined all freelists, fail */
 
                SpinLockAcquire(&(hctl->freeList[borrow_from_idx].mutex));
                newElement = hctl->freeList[borrow_from_idx].freeList;
 
                if (newElement != NULL)
                {
                    hctl->freeList[borrow_from_idx].freeList = newElement->link;
                    SpinLockRelease(&(hctl->freeList[borrow_from_idx].mutex));
 
                    /* careful: count the new element in its proper freelist */
                    SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
                    hctl->freeList[freelist_idx].nentries++;
                    SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
                    return newElement;
                }
 
                SpinLockRelease(&(hctl->freeList[borrow_from_idx].mutex));
            }
 
            /* no elements available to borrow either, so out of memory */
            return NULL;
        }
    }
 
    /* remove entry from freelist, bump nentries */
    hctl->freeList[freelist_idx].freeList = newElement->link;
    hctl->freeList[freelist_idx].nentries++;
 
    if (IS_PARTITIONED(hctl))
        SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
    return newElement;
}
 
/*
 * hash_get_num_entries -- get the number of entries in a hashtable
 */
int64
hash_get_num_entries(HTAB *hashp)
{
    int         i;
    int64       sum = hashp->hctl->freeList[0].nentries;
 
    /*
     * We currently don't bother with acquiring the mutexes; it's only
     * sensible to call this function if you've got lock on all partitions of
     * the table.
     */
    if (IS_PARTITIONED(hashp->hctl))
    {
        for (i = 1; i < NUM_FREELISTS; i++)
            sum += hashp->hctl->freeList[i].nentries;
    }
 
    return sum;
}
 
/*
 * hash_seq_init/_search/_term
 *          Sequentially search through hash table and return
 *          all the elements one by one, return NULL when no more.
 *
 * hash_seq_term should be called if and only if the scan is abandoned before
 * completion; if hash_seq_search returns NULL then it has already done the
 * end-of-scan cleanup.
 *
 * NOTE: caller may delete the returned element before continuing the scan.
 * However, deleting any other element while the scan is in progress is
 * UNDEFINED (it might be the one that curIndex is pointing at!).  Also,
 * if elements are added to the table while the scan is in progress, it is
 * unspecified whether they will be visited by the scan or not.
 *
 * NOTE: it is possible to use hash_seq_init/hash_seq_search without any
 * worry about hash_seq_term cleanup, if the hashtable is first locked against
 * further insertions by calling hash_freeze.
 *
 * NOTE: to use this with a partitioned hashtable, caller had better hold
 * at least shared lock on all partitions of the table throughout the scan!
 * We can cope with insertions or deletions by our own backend, but *not*
 * with concurrent insertions or deletions by another.
 */
void
hash_seq_init(HASH_SEQ_STATUS *status, HTAB *hashp)
{
    status->hashp = hashp;
    status->curBucket = 0;
    status->curEntry = NULL;
    status->hasHashvalue = false;
    if (!hashp->frozen)
        register_seq_scan(hashp);
}
 
/*
 * Same as above but scan by the given hash value.
 * See also hash_seq_search().
 *
 * NOTE: the default hash function doesn't match syscache hash function.
 * Thus, if you're going to use this function in syscache callback, make sure
 * you're using custom hash function.  See relatt_cache_syshash()
 * for example.
 */
void
hash_seq_init_with_hash_value(HASH_SEQ_STATUS *status, HTAB *hashp,
                              uint32 hashvalue)
{
    HASHBUCKET *bucketPtr;
 
    hash_seq_init(status, hashp);
 
    status->hasHashvalue = true;
    status->hashvalue = hashvalue;
 
    status->curBucket = hash_initial_lookup(hashp, hashvalue, &bucketPtr);
    status->curEntry = *bucketPtr;
}
 
void *
hash_seq_search(HASH_SEQ_STATUS *status)
{
    HTAB       *hashp;
    HASHHDR    *hctl;
    uint32      max_bucket;
    int64       ssize;
    int64       segment_num;
    int64       segment_ndx;
    HASHSEGMENT segp;
    uint32      curBucket;
    HASHELEMENT *curElem;
 
    if (status->hasHashvalue)
    {
        /*
         * Scan entries only in the current bucket because only this bucket
         * can contain entries with the given hash value.
         */
        while ((curElem = status->curEntry) != NULL)
        {
            status->curEntry = curElem->link;
            if (status->hashvalue != curElem->hashvalue)
                continue;
            return (void *) ELEMENTKEY(curElem);
        }
 
        hash_seq_term(status);
        return NULL;
    }
 
    if ((curElem = status->curEntry) != NULL)
    {
        /* Continuing scan of curBucket... */
        status->curEntry = curElem->link;
        if (status->curEntry == NULL)   /* end of this bucket */
            ++status->curBucket;
        return ELEMENTKEY(curElem);
    }
 
    /*
     * Search for next nonempty bucket starting at curBucket.
     */
    curBucket = status->curBucket;
    hashp = status->hashp;
    hctl = hashp->hctl;
    ssize = hashp->ssize;
    max_bucket = hctl->max_bucket;
 
    if (curBucket > max_bucket)
    {
        hash_seq_term(status);
        return NULL;            /* search is done */
    }
 
    /*
     * first find the right segment in the table directory.
     */
    segment_num = curBucket >> hashp->sshift;
    segment_ndx = MOD(curBucket, ssize);
 
    segp = hashp->dir[segment_num];
 
    /*
     * Pick up the first item in this bucket's chain.  If chain is not empty
     * we can begin searching it.  Otherwise we have to advance to find the
     * next nonempty bucket.  We try to optimize that case since searching a
     * near-empty hashtable has to iterate this loop a lot.
     */
    while ((curElem = segp[segment_ndx]) == NULL)
    {
        /* empty bucket, advance to next */
        if (++curBucket > max_bucket)
        {
            status->curBucket = curBucket;
            hash_seq_term(status);
            return NULL;        /* search is done */
        }
        if (++segment_ndx >= ssize)
        {
            segment_num++;
            segment_ndx = 0;
            segp = hashp->dir[segment_num];
        }
    }
 
    /* Begin scan of curBucket... */
    status->curEntry = curElem->link;
    if (status->curEntry == NULL)   /* end of this bucket */
        ++curBucket;
    status->curBucket = curBucket;
    return ELEMENTKEY(curElem);
}
 
void
hash_seq_term(HASH_SEQ_STATUS *status)
{
    if (!status->hashp->frozen)
        deregister_seq_scan(status->hashp);
}
 
/*
 * hash_freeze
 *          Freeze a hashtable against future insertions (deletions are
 *          still allowed)
 *
 * The reason for doing this is that by preventing any more bucket splits,
 * we no longer need to worry about registering hash_seq_search scans,
 * and thus caller need not be careful about ensuring hash_seq_term gets
 * called at the right times.
 *
 * Multiple calls to hash_freeze() are allowed, but you can't freeze a table
 * with active scans (since hash_seq_term would then do the wrong thing).
 */
void
hash_freeze(HTAB *hashp)
{
    if (hashp->isshared)
        elog(ERROR, "cannot freeze shared hashtable \"%s\"", hashp->tabname);
    if (!hashp->frozen && has_seq_scans(hashp))
        elog(ERROR, "cannot freeze hashtable \"%s\" because it has active scans",
             hashp->tabname);
    hashp->frozen = true;
}
 
 
/********************************* UTILITIES ************************/
 
/*
 * Expand the table by adding one more hash bucket.
 */
static bool
expand_table(HTAB *hashp)
{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT old_seg,
                new_seg;
    int64       old_bucket,
                new_bucket;
    int64       new_segnum,
                new_segndx;
    int64       old_segnum,
                old_segndx;
    HASHBUCKET *oldlink,
               *newlink;
    HASHBUCKET  currElement,
                nextElement;
 
    Assert(!IS_PARTITIONED(hctl));
 
#ifdef HASH_STATISTICS
    hctl->expansions++;
#endif
 
    new_bucket = hctl->max_bucket + 1;
    new_segnum = new_bucket >> hashp->sshift;
    new_segndx = MOD(new_bucket, hashp->ssize);
 
    if (new_segnum >= hctl->nsegs)
    {
        /* Allocate new segment if necessary -- could fail if dir full */
        if (new_segnum >= hctl->dsize)
            if (!dir_realloc(hashp))
                return false;
        if (!(hashp->dir[new_segnum] = seg_alloc(hashp)))
            return false;
        hctl->nsegs++;
    }
 
    /* OK, we created a new bucket */
    hctl->max_bucket++;
 
    /*
     * *Before* changing masks, find old bucket corresponding to same hash
     * values; values in that bucket may need to be relocated to new bucket.
     * Note that new_bucket is certainly larger than low_mask at this point,
     * so we can skip the first step of the regular hash mask calc.
     */
    old_bucket = (new_bucket & hctl->low_mask);
 
    /*
     * If we crossed a power of 2, readjust masks.
     */
    if ((uint32) new_bucket > hctl->high_mask)
    {
        hctl->low_mask = hctl->high_mask;
        hctl->high_mask = (uint32) new_bucket | hctl->low_mask;
    }
 
    /*
     * Relocate records to the new bucket.  NOTE: because of the way the hash
     * masking is done in calc_bucket, only one old bucket can need to be
     * split at this point.  With a different way of reducing the hash value,
     * that might not be true!
     */
    old_segnum = old_bucket >> hashp->sshift;
    old_segndx = MOD(old_bucket, hashp->ssize);
 
    old_seg = hashp->dir[old_segnum];
    new_seg = hashp->dir[new_segnum];
 
    oldlink = &old_seg[old_segndx];
    newlink = &new_seg[new_segndx];
 
    for (currElement = *oldlink;
         currElement != NULL;
         currElement = nextElement)
    {
        nextElement = currElement->link;
        if ((int64) calc_bucket(hctl, currElement->hashvalue) == old_bucket)
        {
            *oldlink = currElement;
            oldlink = &currElement->link;
        }
        else
        {
            *newlink = currElement;
            newlink = &currElement->link;
        }
    }
    /* don't forget to terminate the rebuilt hash chains... */
    *oldlink = NULL;
    *newlink = NULL;
 
    return true;
}
 
 
static bool
dir_realloc(HTAB *hashp)
{
    HASHSEGMENT *p;
    HASHSEGMENT *old_p;
    int64       new_dsize;
    int64       old_dirsize;
    int64       new_dirsize;
 
    if (hashp->hctl->max_dsize != NO_MAX_DSIZE)
        return false;
 
    /* Reallocate directory */
    new_dsize = hashp->hctl->dsize << 1;
    old_dirsize = hashp->hctl->dsize * sizeof(HASHSEGMENT);
    new_dirsize = new_dsize * sizeof(HASHSEGMENT);
 
    old_p = hashp->dir;
    CurrentDynaHashCxt = hashp->hcxt;
    p = (HASHSEGMENT *) hashp->alloc((Size) new_dirsize);
 
    if (p != NULL)
    {
        memcpy(p, old_p, old_dirsize);
        MemSet(((char *) p) + old_dirsize, 0, new_dirsize - old_dirsize);
        hashp->dir = p;
        hashp->hctl->dsize = new_dsize;
 
        /* XXX assume the allocator is palloc, so we know how to free */
        Assert(hashp->alloc == DynaHashAlloc);
        pfree(old_p);
 
        return true;
    }
 
    return false;
}
 
 
static HASHSEGMENT
seg_alloc(HTAB *hashp)
{
    HASHSEGMENT segp;
 
    CurrentDynaHashCxt = hashp->hcxt;
    segp = (HASHSEGMENT) hashp->alloc(sizeof(HASHBUCKET) * hashp->ssize);
 
    if (!segp)
        return NULL;
 
    MemSet(segp, 0, sizeof(HASHBUCKET) * hashp->ssize);
 
    return segp;
}
 
/*
 * allocate some new elements and link them into the indicated free list
 */
static bool
element_alloc(HTAB *hashp, int nelem, int freelist_idx)
{
    HASHHDR    *hctl = hashp->hctl;
    Size        elementSize;
    Size        requestSize;
    char       *allocedBlock;
    HASHELEMENT *firstElement;
    HASHELEMENT *tmpElement;
    HASHELEMENT *prevElement;
    int         i;
 
    if (hctl->isfixed)
        return false;
 
    /* Each element has a HASHELEMENT header plus user data. */
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(hctl->entrysize);
 
    requestSize = nelem * elementSize;
 
    /* Add space for slist_node list link if we need one. */
#ifdef USE_VALGRIND
    if (!hashp->isshared)
        requestSize += MAXALIGN(sizeof(slist_node));
#endif
 
    /* Allocate the memory. */
    CurrentDynaHashCxt = hashp->hcxt;
    allocedBlock = hashp->alloc(requestSize);
 
    if (!allocedBlock)
        return false;
 
    /*
     * If USE_VALGRIND, each allocated block of elements of a non-shared
     * hashtable is chained into a list, so that Valgrind won't think it's
     * been leaked.
     */
#ifdef USE_VALGRIND
    if (hashp->isshared)
        firstElement = (HASHELEMENT *) allocedBlock;
    else
    {
        slist_push_head(&hashp->element_blocks, (slist_node *) allocedBlock);
        firstElement = (HASHELEMENT *) (allocedBlock + MAXALIGN(sizeof(slist_node)));
    }
#else
    firstElement = (HASHELEMENT *) allocedBlock;
#endif
 
    /* prepare to link all the new entries into the freelist */
    prevElement = NULL;
    tmpElement = firstElement;
    for (i = 0; i < nelem; i++)
    {
        tmpElement->link = prevElement;
        prevElement = tmpElement;
        tmpElement = (HASHELEMENT *) (((char *) tmpElement) + elementSize);
    }
 
    /* if partitioned, must lock to touch freeList */
    if (IS_PARTITIONED(hctl))
        SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
 
    /* freelist could be nonempty if two backends did this concurrently */
    firstElement->link = hctl->freeList[freelist_idx].freeList;
    hctl->freeList[freelist_idx].freeList = prevElement;
 
    if (IS_PARTITIONED(hctl))
        SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
    return true;
}
 
/*
 * Do initial lookup of a bucket for the given hash value, retrieving its
 * bucket number and its hash bucket.
 */
static inline uint32
hash_initial_lookup(HTAB *hashp, uint32 hashvalue, HASHBUCKET **bucketptr)
{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT segp;
    int64       segment_num;
    int64       segment_ndx;
    uint32      bucket;
 
    bucket = calc_bucket(hctl, hashvalue);
 
    segment_num = bucket >> hashp->sshift;
    segment_ndx = MOD(bucket, hashp->ssize);
 
    segp = hashp->dir[segment_num];
 
    if (segp == NULL)
        hash_corrupted(hashp);
 
    *bucketptr = &segp[segment_ndx];
    return bucket;
}
 
/* complain when we have detected a corrupted hashtable */
static void
hash_corrupted(HTAB *hashp)
{
    /*
     * If the corruption is in a shared hashtable, we'd better force a
     * systemwide restart.  Otherwise, just shut down this one backend.
     */
    if (hashp->isshared)
        elog(PANIC, "hash table \"%s\" corrupted", hashp->tabname);
    else
        elog(FATAL, "hash table \"%s\" corrupted", hashp->tabname);
}
 
/* calculate ceil(log base 2) of num */
static int
my_log2(int64 num)
{
    /*
     * guard against too-large input, which would be invalid for
     * pg_ceil_log2_*()
     */
    if (num > PG_INT64_MAX / 2)
        num = PG_INT64_MAX / 2;
 
    return pg_ceil_log2_64(num);
}
 
/* calculate first power of 2 >= num, bounded to what will fit in a int64 */
static int64
next_pow2_int64(int64 num)
{
    /* my_log2's internal range check is sufficient */
    return 1L << my_log2(num);
}
 
/* calculate first power of 2 >= num, bounded to what will fit in an int */
static int
next_pow2_int(int64 num)
{
    if (num > INT_MAX / 2)
        num = INT_MAX / 2;
    return 1 << my_log2(num);
}
 
 
/************************* SEQ SCAN TRACKING ************************/
 
/*
 * We track active hash_seq_search scans here.  The need for this mechanism
 * comes from the fact that a scan will get confused if a bucket split occurs
 * while it's in progress: it might visit entries twice, or even miss some
 * entirely (if it's partway through the same bucket that splits).  Hence
 * we want to inhibit bucket splits if there are any active scans on the
 * table being inserted into.  This is a fairly rare case in current usage,
 * so just postponing the split until the next insertion seems sufficient.
 *
 * Given present usages of the function, only a few scans are likely to be
 * open concurrently; so a finite-size stack of open scans seems sufficient,
 * and we don't worry that linear search is too slow.  Note that we do
 * allow multiple scans of the same hashtable to be open concurrently.
 *
 * This mechanism can support concurrent scan and insertion in a shared
 * hashtable if it's the same backend doing both.  It would fail otherwise,
 * but locking reasons seem to preclude any such scenario anyway, so we don't
 * worry.
 *
 * This arrangement is reasonably robust if a transient hashtable is deleted
 * without notifying us.  The absolute worst case is we might inhibit splits
 * in another table created later at exactly the same address.  We will give
 * a warning at transaction end for reference leaks, so any bugs leading to
 * lack of notification should be easy to catch.
 */
 
#define MAX_SEQ_SCANS 100
 
static HTAB *seq_scan_tables[MAX_SEQ_SCANS];    /* tables being scanned */
static int  seq_scan_level[MAX_SEQ_SCANS];  /* subtransaction nest level */
static int  num_seq_scans = 0;
 
 
/* Register a table as having an active hash_seq_search scan */
static void
register_seq_scan(HTAB *hashp)
{
    if (num_seq_scans >= MAX_SEQ_SCANS)
        elog(ERROR, "too many active hash_seq_search scans, cannot start one on \"%s\"",
             hashp->tabname);
    seq_scan_tables[num_seq_scans] = hashp;
    seq_scan_level[num_seq_scans] = GetCurrentTransactionNestLevel();
    num_seq_scans++;
}
 
/* Deregister an active scan */
static void
deregister_seq_scan(HTAB *hashp)
{
    int         i;
 
    /* Search backward since it's most likely at the stack top */
    for (i = num_seq_scans - 1; i >= 0; i--)
    {
        if (seq_scan_tables[i] == hashp)
        {
            seq_scan_tables[i] = seq_scan_tables[num_seq_scans - 1];
            seq_scan_level[i] = seq_scan_level[num_seq_scans - 1];
            num_seq_scans--;
            return;
        }
    }
    elog(ERROR, "no hash_seq_search scan for hash table \"%s\"",
         hashp->tabname);
}
 
/* Check if a table has any active scan */
static bool
has_seq_scans(HTAB *hashp)
{
    int         i;
 
    for (i = 0; i < num_seq_scans; i++)
    {
        if (seq_scan_tables[i] == hashp)
            return true;
    }
    return false;
}
 
/* Clean up any open scans at end of transaction */
void
AtEOXact_HashTables(bool isCommit)
{
    /*
     * During abort cleanup, open scans are expected; just silently clean 'em
     * out.  An open scan at commit means someone forgot a hash_seq_term()
     * call, so complain.
     *
     * Note: it's tempting to try to print the tabname here, but refrain for
     * fear of touching deallocated memory.  This isn't a user-facing message
     * anyway, so it needn't be pretty.
     */
    if (isCommit)
    {
        int         i;
 
        for (i = 0; i < num_seq_scans; i++)
        {
            elog(WARNING, "leaked hash_seq_search scan for hash table %p",
                 seq_scan_tables[i]);
        }
    }
    num_seq_scans = 0;
}
 
/* Clean up any open scans at end of subtransaction */
void
AtEOSubXact_HashTables(bool isCommit, int nestDepth)
{
    int         i;
 
    /*
     * Search backward to make cleanup easy.  Note we must check all entries,
     * not only those at the end of the array, because deletion technique
     * doesn't keep them in order.
     */
    for (i = num_seq_scans - 1; i >= 0; i--)
    {
        if (seq_scan_level[i] >= nestDepth)
        {
            if (isCommit)
                elog(WARNING, "leaked hash_seq_search scan for hash table %p",
                     seq_scan_tables[i]);
            seq_scan_tables[i] = seq_scan_tables[num_seq_scans - 1];
            seq_scan_level[i] = seq_scan_level[num_seq_scans - 1];
            num_seq_scans--;
        }
    }
}

ELEMENTKEY

#define ELEMENTKEY

(

helem

)

(((char *)(helem)) + MAXALIGN(sizeof(HASHELEMENT)))

Definition at line 255 of file dynahash.c.

FREELIST_IDX

#define FREELIST_IDX	(		hctl,
			hashcode
	)		(IS_PARTITIONED(hctl) ? (hashcode) % NUM_FREELISTS : 0)

Definition at line 214 of file dynahash.c.

                                                       : 0)

IS_PARTITIONED

#define IS_PARTITIONED

(

hctl

)

((hctl)->num_partitions != 0)

Definition at line 212 of file dynahash.c.

MAX_SEQ_SCANS

#define MAX_SEQ_SCANS   100

Definition at line 1875 of file dynahash.c.

MOD

#define MOD	(		x,
			y
	)		((x) & ((y)-1))

Definition at line 266 of file dynahash.c.

NUM_FREELISTS

#define NUM_FREELISTS   32

Definition at line 128 of file dynahash.c.

Typedef Documentation

HASHBUCKET

typedef HASHELEMENT* HASHBUCKET

Definition at line 131 of file dynahash.c.

HASHSEGMENT

typedef HASHBUCKET* HASHSEGMENT

Definition at line 134 of file dynahash.c.

Function Documentation

AtEOSubXact_HashTables()

void AtEOSubXact_HashTables	(	bool	isCommit,
		int	nestDepth
	)

Definition at line 1957 of file dynahash.c.

{
    int         i;
 
    /*
     * Search backward to make cleanup easy.  Note we must check all entries,
     * not only those at the end of the array, because deletion technique
     * doesn't keep them in order.
     */
    for (i = num_seq_scans - 1; i >= 0; i--)
    {
        if (seq_scan_level[i] >= nestDepth)
        {
            if (isCommit)
                elog(WARNING, "leaked hash_seq_search scan for hash table %p",
                     seq_scan_tables[i]);
            seq_scan_tables[i] = seq_scan_tables[num_seq_scans - 1];
            seq_scan_level[i] = seq_scan_level[num_seq_scans - 1];
            num_seq_scans--;
        }
    }
}

References elog, fb(), i, num_seq_scans, seq_scan_level, seq_scan_tables, and WARNING.

Referenced by AbortSubTransaction(), and CommitSubTransaction().

AtEOXact_HashTables()

void AtEOXact_HashTables

(

bool

isCommit

)

Definition at line 1931 of file dynahash.c.

{
    /*
     * During abort cleanup, open scans are expected; just silently clean 'em
     * out.  An open scan at commit means someone forgot a hash_seq_term()
     * call, so complain.
     *
     * Note: it's tempting to try to print the tabname here, but refrain for
     * fear of touching deallocated memory.  This isn't a user-facing message
     * anyway, so it needn't be pretty.
     */
    if (isCommit)
    {
        int         i;
 
        for (i = 0; i < num_seq_scans; i++)
        {
            elog(WARNING, "leaked hash_seq_search scan for hash table %p",
                 seq_scan_tables[i]);
        }
    }
    num_seq_scans = 0;
}

References elog, fb(), i, num_seq_scans, seq_scan_tables, and WARNING.

Referenced by AbortTransaction(), AutoVacLauncherMain(), BackgroundWriterMain(), CheckpointerMain(), CommitTransaction(), pgarch_archiveXlog(), PrepareTransaction(), WalSummarizerMain(), and WalWriterMain().

calc_bucket()

static uint32 calc_bucket ( HASHHDR *  hctl, uint32  hash_val  )

inlinestatic

Definition at line 915 of file dynahash.c.

{
    uint32      bucket;
 
    bucket = hash_val & hctl->high_mask;
    if (bucket > hctl->max_bucket)
        bucket = bucket & hctl->low_mask;
 
    return bucket;
}

References fb(), HASHHDR::high_mask, HASHHDR::low_mask, and HASHHDR::max_bucket.

Referenced by expand_table(), and hash_initial_lookup().

choose_nelem_alloc()

static int choose_nelem_alloc ( Size  entrysize )

static

Definition at line 667 of file dynahash.c.

{
    int         nelem_alloc;
    Size        elementSize;
    Size        allocSize;
 
    /* Each element has a HASHELEMENT header plus user data. */
    /* NB: this had better match element_alloc() */
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(entrysize);
 
    /*
     * The idea here is to choose nelem_alloc at least 32, but round up so
     * that the allocation request will be a power of 2 or just less. This
     * makes little difference for hash tables in shared memory, but for hash
     * tables managed by palloc, the allocation request will be rounded up to
     * a power of 2 anyway.  If we fail to take this into account, we'll waste
     * as much as half the allocated space.
     */
    allocSize = 32 * 4;         /* assume elementSize at least 8 */
    do
    {
        allocSize <<= 1;
        nelem_alloc = allocSize / elementSize;
    } while (nelem_alloc < 32);
 
    return nelem_alloc;
}

References fb(), and MAXALIGN.

Referenced by hash_estimate_size(), and init_htab().

deregister_seq_scan()

static void deregister_seq_scan ( HTAB *  hashp )

static

Definition at line 1896 of file dynahash.c.

{
    int         i;
 
    /* Search backward since it's most likely at the stack top */
    for (i = num_seq_scans - 1; i >= 0; i--)
    {
        if (seq_scan_tables[i] == hashp)
        {
            seq_scan_tables[i] = seq_scan_tables[num_seq_scans - 1];
            seq_scan_level[i] = seq_scan_level[num_seq_scans - 1];
            num_seq_scans--;
            return;
        }
    }
    elog(ERROR, "no hash_seq_search scan for hash table \"%s\"",
         hashp->tabname);
}

References elog, ERROR, i, num_seq_scans, seq_scan_level, seq_scan_tables, and HTAB::tabname.

Referenced by hash_seq_term().

dir_realloc()

static bool dir_realloc ( HTAB *  hashp )

static

Definition at line 1643 of file dynahash.c.

{
    HASHSEGMENT *p;
    HASHSEGMENT *old_p;
    int64       new_dsize;
    int64       old_dirsize;
    int64       new_dirsize;
 
    if (hashp->hctl->max_dsize != NO_MAX_DSIZE)
        return false;
 
    /* Reallocate directory */
    new_dsize = hashp->hctl->dsize << 1;
    old_dirsize = hashp->hctl->dsize * sizeof(HASHSEGMENT);
    new_dirsize = new_dsize * sizeof(HASHSEGMENT);
 
    old_p = hashp->dir;
    CurrentDynaHashCxt = hashp->hcxt;
    p = (HASHSEGMENT *) hashp->alloc((Size) new_dirsize);
 
    if (p != NULL)
    {
        memcpy(p, old_p, old_dirsize);
        MemSet(((char *) p) + old_dirsize, 0, new_dirsize - old_dirsize);
        hashp->dir = p;
        hashp->hctl->dsize = new_dsize;
 
        /* XXX assume the allocator is palloc, so we know how to free */
        Assert(hashp->alloc == DynaHashAlloc);
        pfree(old_p);
 
        return true;
    }
 
    return false;
}

References HTAB::alloc, Assert, CurrentDynaHashCxt, HTAB::dir, HASHHDR::dsize, DynaHashAlloc(), fb(), HTAB::hctl, HTAB::hcxt, HASHHDR::max_dsize, MemSet, NO_MAX_DSIZE, and pfree().

Referenced by expand_table().

DynaHashAlloc()

static void * DynaHashAlloc ( Size  size )

static

Definition at line 297 of file dynahash.c.

{
    Assert(MemoryContextIsValid(CurrentDynaHashCxt));
    return MemoryContextAllocExtended(CurrentDynaHashCxt, size,
                                      MCXT_ALLOC_NO_OOM);
}

References Assert, CurrentDynaHashCxt, MCXT_ALLOC_NO_OOM, MemoryContextAllocExtended(), and MemoryContextIsValid.

Referenced by dir_realloc(), hash_create(), and hash_destroy().

element_alloc()

static bool element_alloc ( HTAB *  hashp, int  nelem, int  freelist_idx  )

static

Definition at line 1701 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    Size        elementSize;
    Size        requestSize;
    char       *allocedBlock;
    HASHELEMENT *firstElement;
    HASHELEMENT *tmpElement;
    HASHELEMENT *prevElement;
    int         i;
 
    if (hctl->isfixed)
        return false;
 
    /* Each element has a HASHELEMENT header plus user data. */
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(hctl->entrysize);
 
    requestSize = nelem * elementSize;
 
    /* Add space for slist_node list link if we need one. */
#ifdef USE_VALGRIND
    if (!hashp->isshared)
        requestSize += MAXALIGN(sizeof(slist_node));
#endif
 
    /* Allocate the memory. */
    CurrentDynaHashCxt = hashp->hcxt;
    allocedBlock = hashp->alloc(requestSize);
 
    if (!allocedBlock)
        return false;
 
    /*
     * If USE_VALGRIND, each allocated block of elements of a non-shared
     * hashtable is chained into a list, so that Valgrind won't think it's
     * been leaked.
     */
#ifdef USE_VALGRIND
    if (hashp->isshared)
        firstElement = (HASHELEMENT *) allocedBlock;
    else
    {
        slist_push_head(&hashp->element_blocks, (slist_node *) allocedBlock);
        firstElement = (HASHELEMENT *) (allocedBlock + MAXALIGN(sizeof(slist_node)));
    }
#else
    firstElement = (HASHELEMENT *) allocedBlock;
#endif
 
    /* prepare to link all the new entries into the freelist */
    prevElement = NULL;
    tmpElement = firstElement;
    for (i = 0; i < nelem; i++)
    {
        tmpElement->link = prevElement;
        prevElement = tmpElement;
        tmpElement = (HASHELEMENT *) (((char *) tmpElement) + elementSize);
    }
 
    /* if partitioned, must lock to touch freeList */
    if (IS_PARTITIONED(hctl))
        SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
 
    /* freelist could be nonempty if two backends did this concurrently */
    firstElement->link = hctl->freeList[freelist_idx].freeList;
    hctl->freeList[freelist_idx].freeList = prevElement;
 
    if (IS_PARTITIONED(hctl))
        SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
    return true;
}

References HTAB::alloc, CurrentDynaHashCxt, HASHHDR::entrysize, fb(), FreeListData::freeList, HASHHDR::freeList, HTAB::hctl, HTAB::hcxt, i, IS_PARTITIONED, HASHHDR::isfixed, HTAB::isshared, MAXALIGN, FreeListData::mutex, slist_push_head(), SpinLockAcquire, and SpinLockRelease.

Referenced by get_hash_entry(), and hash_create().

expand_table()

static bool expand_table ( HTAB *  hashp )

static

Definition at line 1546 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT old_seg,
                new_seg;
    int64       old_bucket,
                new_bucket;
    int64       new_segnum,
                new_segndx;
    int64       old_segnum,
                old_segndx;
    HASHBUCKET *oldlink,
               *newlink;
    HASHBUCKET  currElement,
                nextElement;
 
    Assert(!IS_PARTITIONED(hctl));
 
#ifdef HASH_STATISTICS
    hctl->expansions++;
#endif
 
    new_bucket = hctl->max_bucket + 1;
    new_segnum = new_bucket >> hashp->sshift;
    new_segndx = MOD(new_bucket, hashp->ssize);
 
    if (new_segnum >= hctl->nsegs)
    {
        /* Allocate new segment if necessary -- could fail if dir full */
        if (new_segnum >= hctl->dsize)
            if (!dir_realloc(hashp))
                return false;
        if (!(hashp->dir[new_segnum] = seg_alloc(hashp)))
            return false;
        hctl->nsegs++;
    }
 
    /* OK, we created a new bucket */
    hctl->max_bucket++;
 
    /*
     * *Before* changing masks, find old bucket corresponding to same hash
     * values; values in that bucket may need to be relocated to new bucket.
     * Note that new_bucket is certainly larger than low_mask at this point,
     * so we can skip the first step of the regular hash mask calc.
     */
    old_bucket = (new_bucket & hctl->low_mask);
 
    /*
     * If we crossed a power of 2, readjust masks.
     */
    if ((uint32) new_bucket > hctl->high_mask)
    {
        hctl->low_mask = hctl->high_mask;
        hctl->high_mask = (uint32) new_bucket | hctl->low_mask;
    }
 
    /*
     * Relocate records to the new bucket.  NOTE: because of the way the hash
     * masking is done in calc_bucket, only one old bucket can need to be
     * split at this point.  With a different way of reducing the hash value,
     * that might not be true!
     */
    old_segnum = old_bucket >> hashp->sshift;
    old_segndx = MOD(old_bucket, hashp->ssize);
 
    old_seg = hashp->dir[old_segnum];
    new_seg = hashp->dir[new_segnum];
 
    oldlink = &old_seg[old_segndx];
    newlink = &new_seg[new_segndx];
 
    for (currElement = *oldlink;
         currElement != NULL;
         currElement = nextElement)
    {
        nextElement = currElement->link;
        if ((int64) calc_bucket(hctl, currElement->hashvalue) == old_bucket)
        {
            *oldlink = currElement;
            oldlink = &currElement->link;
        }
        else
        {
            *newlink = currElement;
            newlink = &currElement->link;
        }
    }
    /* don't forget to terminate the rebuilt hash chains... */
    *oldlink = NULL;
    *newlink = NULL;
 
    return true;
}

References Assert, calc_bucket(), HTAB::dir, dir_realloc(), HASHHDR::dsize, fb(), HTAB::hctl, HASHHDR::high_mask, IS_PARTITIONED, HASHELEMENT::link, HASHHDR::low_mask, HASHHDR::max_bucket, MOD, HASHHDR::nsegs, seg_alloc(), HTAB::sshift, and HTAB::ssize.

Referenced by hash_search_with_hash_value().

get_hash_entry()

static HASHBUCKET get_hash_entry ( HTAB *  hashp, int  freelist_idx  )

static

Definition at line 1251 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    HASHBUCKET  newElement;
 
    for (;;)
    {
        /* if partitioned, must lock to touch nentries and freeList */
        if (IS_PARTITIONED(hctl))
            SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
 
        /* try to get an entry from the freelist */
        newElement = hctl->freeList[freelist_idx].freeList;
 
        if (newElement != NULL)
            break;
 
        if (IS_PARTITIONED(hctl))
            SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
        /*
         * No free elements in this freelist.  In a partitioned table, there
         * might be entries in other freelists, but to reduce contention we
         * prefer to first try to get another chunk of buckets from the main
         * shmem allocator.  If that fails, though, we *MUST* root through all
         * the other freelists before giving up.  There are multiple callers
         * that assume that they can allocate every element in the initially
         * requested table size, or that deleting an element guarantees they
         * can insert a new element, even if shared memory is entirely full.
         * Failing because the needed element is in a different freelist is
         * not acceptable.
         */
        if (!element_alloc(hashp, hctl->nelem_alloc, freelist_idx))
        {
            int         borrow_from_idx;
 
            if (!IS_PARTITIONED(hctl))
                return NULL;    /* out of memory */
 
            /* try to borrow element from another freelist */
            borrow_from_idx = freelist_idx;
            for (;;)
            {
                borrow_from_idx = (borrow_from_idx + 1) % NUM_FREELISTS;
                if (borrow_from_idx == freelist_idx)
                    break;      /* examined all freelists, fail */
 
                SpinLockAcquire(&(hctl->freeList[borrow_from_idx].mutex));
                newElement = hctl->freeList[borrow_from_idx].freeList;
 
                if (newElement != NULL)
                {
                    hctl->freeList[borrow_from_idx].freeList = newElement->link;
                    SpinLockRelease(&(hctl->freeList[borrow_from_idx].mutex));
 
                    /* careful: count the new element in its proper freelist */
                    SpinLockAcquire(&hctl->freeList[freelist_idx].mutex);
                    hctl->freeList[freelist_idx].nentries++;
                    SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
                    return newElement;
                }
 
                SpinLockRelease(&(hctl->freeList[borrow_from_idx].mutex));
            }
 
            /* no elements available to borrow either, so out of memory */
            return NULL;
        }
    }
 
    /* remove entry from freelist, bump nentries */
    hctl->freeList[freelist_idx].freeList = newElement->link;
    hctl->freeList[freelist_idx].nentries++;
 
    if (IS_PARTITIONED(hctl))
        SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
    return newElement;
}

References element_alloc(), fb(), FreeListData::freeList, HASHHDR::freeList, HTAB::hctl, IS_PARTITIONED, HASHELEMENT::link, FreeListData::mutex, HASHHDR::nelem_alloc, FreeListData::nentries, NUM_FREELISTS, SpinLockAcquire, and SpinLockRelease.

Referenced by hash_search_with_hash_value().

get_hash_value()

uint32 get_hash_value	(	HTAB *	hashp,
		const void *	keyPtr
	)

Definition at line 908 of file dynahash.c.

{
    return hashp->hash(keyPtr, hashp->keysize);
}

References fb(), HTAB::hash, and HTAB::keysize.

Referenced by BufTableHashCode(), LockTagHashCode(), and lookup_type_cache().

has_seq_scans()

static bool has_seq_scans ( HTAB *  hashp )

static

Definition at line 1917 of file dynahash.c.

{
    int         i;
 
    for (i = 0; i < num_seq_scans; i++)
    {
        if (seq_scan_tables[i] == hashp)
            return true;
    }
    return false;
}

References i, num_seq_scans, and seq_scan_tables.

Referenced by hash_freeze(), and hash_search_with_hash_value().

hash_corrupted()

static void hash_corrupted ( HTAB *  hashp )

static

Definition at line 1803 of file dynahash.c.

{
    /*
     * If the corruption is in a shared hashtable, we'd better force a
     * systemwide restart.  Otherwise, just shut down this one backend.
     */
    if (hashp->isshared)
        elog(PANIC, "hash table \"%s\" corrupted", hashp->tabname);
    else
        elog(FATAL, "hash table \"%s\" corrupted", hashp->tabname);
}

References elog, FATAL, HTAB::isshared, PANIC, and HTAB::tabname.

Referenced by hash_initial_lookup().

hash_create()

HTAB * hash_create	(	const char *	tabname,
		int64	nelem,
		const HASHCTL *	info,
		int	flags
	)

Definition at line 358 of file dynahash.c.

{
    HTAB       *hashp;
    HASHHDR    *hctl;
 
    /*
     * Hash tables now allocate space for key and data, but you have to say
     * how much space to allocate.
     */
    Assert(flags & HASH_ELEM);
    Assert(info->keysize > 0);
    Assert(info->entrysize >= info->keysize);
 
    /*
     * For shared hash tables, we have a local hash header (HTAB struct) that
     * we allocate in TopMemoryContext; all else is in shared memory.
     *
     * For non-shared hash tables, everything including the hash header is in
     * a memory context created specially for the hash table --- this makes
     * hash_destroy very simple.  The memory context is made a child of either
     * a context specified by the caller, or TopMemoryContext if nothing is
     * specified.
     */
    if (flags & HASH_SHARED_MEM)
    {
        /* Set up to allocate the hash header */
        CurrentDynaHashCxt = TopMemoryContext;
    }
    else
    {
        /* Create the hash table's private memory context */
        if (flags & HASH_CONTEXT)
            CurrentDynaHashCxt = info->hcxt;
        else
            CurrentDynaHashCxt = TopMemoryContext;
        CurrentDynaHashCxt = AllocSetContextCreate(CurrentDynaHashCxt,
                                                   "dynahash",
                                                   ALLOCSET_DEFAULT_SIZES);
    }
 
    /* Initialize the hash header, plus a copy of the table name */
    hashp = (HTAB *) MemoryContextAlloc(CurrentDynaHashCxt,
                                        sizeof(HTAB) + strlen(tabname) + 1);
    MemSet(hashp, 0, sizeof(HTAB));
 
    hashp->tabname = (char *) (hashp + 1);
    strcpy(hashp->tabname, tabname);
 
    /* If we have a private context, label it with hashtable's name */
    if (!(flags & HASH_SHARED_MEM))
        MemoryContextSetIdentifier(CurrentDynaHashCxt, hashp->tabname);
 
    /*
     * Select the appropriate hash function (see comments at head of file).
     */
    if (flags & HASH_FUNCTION)
    {
        Assert(!(flags & (HASH_BLOBS | HASH_STRINGS)));
        hashp->hash = info->hash;
    }
    else if (flags & HASH_BLOBS)
    {
        Assert(!(flags & HASH_STRINGS));
        /* We can optimize hashing for common key sizes */
        if (info->keysize == sizeof(uint32))
            hashp->hash = uint32_hash;
        else
            hashp->hash = tag_hash;
    }
    else
    {
        /*
         * string_hash used to be considered the default hash method, and in a
         * non-assert build it effectively still is.  But we now consider it
         * an assertion error to not say HASH_STRINGS explicitly.  To help
         * catch mistaken usage of HASH_STRINGS, we also insist on a
         * reasonably long string length: if the keysize is only 4 or 8 bytes,
         * it's almost certainly an integer or pointer not a string.
         */
        Assert(flags & HASH_STRINGS);
        Assert(info->keysize > 8);
 
        hashp->hash = string_hash;
    }
 
    /*
     * If you don't specify a match function, it defaults to string_compare if
     * you used string_hash, and to memcmp otherwise.
     *
     * Note: explicitly specifying string_hash is deprecated, because this
     * might not work for callers in loadable modules on some platforms due to
     * referencing a trampoline instead of the string_hash function proper.
     * Specify HASH_STRINGS instead.
     */
    if (flags & HASH_COMPARE)
        hashp->match = info->match;
    else if (hashp->hash == string_hash)
        hashp->match = (HashCompareFunc) string_compare;
    else
        hashp->match = memcmp;
 
    /*
     * Similarly, the key-copying function defaults to strlcpy or memcpy.
     */
    if (flags & HASH_KEYCOPY)
        hashp->keycopy = info->keycopy;
    else if (hashp->hash == string_hash)
    {
        /*
         * The signature of keycopy is meant for memcpy(), which returns
         * void*, but strlcpy() returns size_t.  Since we never use the return
         * value of keycopy, and size_t is pretty much always the same size as
         * void *, this should be safe.  The extra cast in the middle is to
         * avoid warnings from -Wcast-function-type.
         */
        hashp->keycopy = (HashCopyFunc) (pg_funcptr_t) strlcpy;
    }
    else
        hashp->keycopy = memcpy;
 
    /* And select the entry allocation function, too. */
    if (flags & HASH_ALLOC)
        hashp->alloc = info->alloc;
    else
        hashp->alloc = DynaHashAlloc;
 
    if (flags & HASH_SHARED_MEM)
    {
        /*
         * ctl structure and directory are preallocated for shared memory
         * tables.  Note that HASH_DIRSIZE and HASH_ALLOC had better be set as
         * well.
         */
        hashp->hctl = info->hctl;
        hashp->dir = (HASHSEGMENT *) (((char *) info->hctl) + sizeof(HASHHDR));
        hashp->hcxt = NULL;
        hashp->isshared = true;
 
        /* hash table already exists, we're just attaching to it */
        if (flags & HASH_ATTACH)
        {
            /* make local copies of some heavily-used values */
            hctl = hashp->hctl;
            hashp->keysize = hctl->keysize;
            hashp->ssize = hctl->ssize;
            hashp->sshift = hctl->sshift;
 
            return hashp;
        }
    }
    else
    {
        /* setup hash table defaults */
        hashp->hctl = NULL;
        hashp->dir = NULL;
        hashp->hcxt = CurrentDynaHashCxt;
        hashp->isshared = false;
    }
 
    if (!hashp->hctl)
    {
        hashp->hctl = (HASHHDR *) hashp->alloc(sizeof(HASHHDR));
        if (!hashp->hctl)
            ereport(ERROR,
                    (errcode(ERRCODE_OUT_OF_MEMORY),
                     errmsg("out of memory")));
    }
 
    hashp->frozen = false;
 
    hdefault(hashp);
 
    hctl = hashp->hctl;
 
    if (flags & HASH_PARTITION)
    {
        /* Doesn't make sense to partition a local hash table */
        Assert(flags & HASH_SHARED_MEM);
 
        /*
         * The number of partitions had better be a power of 2. Also, it must
         * be less than INT_MAX (see init_htab()), so call the int version of
         * next_pow2.
         */
        Assert(info->num_partitions == next_pow2_int(info->num_partitions));
 
        hctl->num_partitions = info->num_partitions;
    }
 
    if (flags & HASH_SEGMENT)
    {
        hctl->ssize = info->ssize;
        hctl->sshift = my_log2(info->ssize);
        /* ssize had better be a power of 2 */
        Assert(hctl->ssize == (1L << hctl->sshift));
    }
 
    /*
     * SHM hash tables have fixed directory size passed by the caller.
     */
    if (flags & HASH_DIRSIZE)
    {
        hctl->max_dsize = info->max_dsize;
        hctl->dsize = info->dsize;
    }
 
    /* remember the entry sizes, too */
    hctl->keysize = info->keysize;
    hctl->entrysize = info->entrysize;
 
    /* make local copies of heavily-used constant fields */
    hashp->keysize = hctl->keysize;
    hashp->ssize = hctl->ssize;
    hashp->sshift = hctl->sshift;
 
    /* Build the hash directory structure */
    if (!init_htab(hashp, nelem))
        elog(ERROR, "failed to initialize hash table \"%s\"", hashp->tabname);
 
    /*
     * For a shared hash table, preallocate the requested number of elements.
     * This reduces problems with run-time out-of-shared-memory conditions.
     *
     * For a non-shared hash table, preallocate the requested number of
     * elements if it's less than our chosen nelem_alloc.  This avoids wasting
     * space if the caller correctly estimates a small table size.
     */
    if ((flags & HASH_SHARED_MEM) ||
        nelem < hctl->nelem_alloc)
    {
        int         i,
                    freelist_partitions,
                    nelem_alloc,
                    nelem_alloc_first;
 
        /*
         * If hash table is partitioned, give each freelist an equal share of
         * the initial allocation.  Otherwise only freeList[0] is used.
         */
        if (IS_PARTITIONED(hashp->hctl))
            freelist_partitions = NUM_FREELISTS;
        else
            freelist_partitions = 1;
 
        nelem_alloc = nelem / freelist_partitions;
        if (nelem_alloc <= 0)
            nelem_alloc = 1;
 
        /*
         * Make sure we'll allocate all the requested elements; freeList[0]
         * gets the excess if the request isn't divisible by NUM_FREELISTS.
         */
        if (nelem_alloc * freelist_partitions < nelem)
            nelem_alloc_first =
                nelem - nelem_alloc * (freelist_partitions - 1);
        else
            nelem_alloc_first = nelem_alloc;
 
        for (i = 0; i < freelist_partitions; i++)
        {
            int         temp = (i == 0) ? nelem_alloc_first : nelem_alloc;
 
            if (!element_alloc(hashp, temp, i))
                ereport(ERROR,
                        (errcode(ERRCODE_OUT_OF_MEMORY),
                         errmsg("out of memory")));
        }
    }
 
    /* Set isfixed if requested, but not till after we build initial entries */
    if (flags & HASH_FIXED_SIZE)
        hctl->isfixed = true;
 
    return hashp;
}

References HTAB::alloc, HASHCTL::alloc, ALLOCSET_DEFAULT_SIZES, AllocSetContextCreate, Assert, CurrentDynaHashCxt, HTAB::dir, HASHHDR::dsize, HASHCTL::dsize, DynaHashAlloc(), element_alloc(), elog, HASHHDR::entrysize, HASHCTL::entrysize, ereport, errcode(), errmsg(), ERROR, fb(), HTAB::frozen, HTAB::hash, HASHCTL::hash, HASH_ALLOC, HASH_ATTACH, HASH_BLOBS, HASH_COMPARE, HASH_CONTEXT, HASH_DIRSIZE, HASH_ELEM, HASH_FIXED_SIZE, HASH_FUNCTION, HASH_KEYCOPY, HASH_PARTITION, HASH_SEGMENT, HASH_SHARED_MEM, HASH_STRINGS, HTAB::hctl, HASHCTL::hctl, HTAB::hcxt, HASHCTL::hcxt, hdefault(), i, init_htab(), IS_PARTITIONED, HASHHDR::isfixed, HTAB::isshared, HTAB::keycopy, HASHCTL::keycopy, HASHHDR::keysize, HTAB::keysize, HASHCTL::keysize, HTAB::match, HASHCTL::match, HASHHDR::max_dsize, HASHCTL::max_dsize, MemoryContextAlloc(), MemoryContextSetIdentifier(), MemSet, my_log2(), next_pow2_int(), NUM_FREELISTS, HASHHDR::num_partitions, HASHCTL::num_partitions, HASHHDR::sshift, HTAB::sshift, HASHHDR::ssize, HTAB::ssize, HASHCTL::ssize, string_compare(), string_hash(), strlcpy(), HTAB::tabname, tag_hash(), TopMemoryContext, and uint32_hash().

Referenced by _hash_finish_split(), _PG_init(), AddEventToPendingNotifies(), AddPendingSync(), assign_record_type_typmod(), begin_heap_rewrite(), build_colinfo_names_hash(), build_guc_variables(), build_join_rel_hash(), BuildEventTriggerCache(), cfunc_hashtable_init(), CheckForSessionAndXactLocks(), CompactCheckpointerRequestQueue(), compute_array_stats(), compute_tsvector_stats(), create_seq_hashtable(), createConnHash(), CreateLocalPredicateLockHash(), CreatePartitionDirectory(), do_autovacuum(), EnablePortalManager(), ExecInitModifyTable(), ExecuteTruncateGuts(), find_all_inheritors(), find_oper_cache_entry(), find_rendezvous_variable(), get_json_object_as_hash(), get_relation_notnullatts(), GetComboCommandId(), GetConnection(), gistInitBuildBuffers(), gistInitParentMap(), init_missing_cache(), init_procedure_caches(), init_rel_sync_cache(), init_timezone_hashtable(), init_ts_config_cache(), init_uncommitted_enum_types(), init_uncommitted_enum_values(), InitBufferManagerAccess(), InitializeAttoptCache(), InitializeRelfilenumberMap(), InitializeShippableCache(), InitializeTableSpaceCache(), InitLocalBuffers(), initLocalChannelTable(), InitLockManagerAccess(), initPendingListenActions(), InitQueryHashTable(), InitRecoveryTransactionEnvironment(), InitSync(), json_unique_check_init(), load_categories_hash(), log_invalid_page(), logical_begin_heap_rewrite(), logicalrep_partmap_init(), logicalrep_relmap_init(), lookup_proof_cache(), lookup_ts_dictionary_cache(), lookup_ts_parser_cache(), lookup_type_cache(), LookupOpclassInfo(), pa_allocate_worker(), pg_get_backend_memory_contexts(), plpgsql_estate_setup(), PLy_add_exceptions(), populate_recordset_object_start(), ProcessSyncingTablesForApply(), rebuild_database_list(), record_C_func(), RegisterExtensibleNodeEntry(), RelationCacheInitialize(), ReorderBufferAllocate(), ReorderBufferBuildTupleCidHash(), ReorderBufferToastInitHash(), ResetUnloggedRelationsInDbspaceDir(), ri_InitHashTables(), select_perl_context(), SerializePendingSyncs(), set_rtable_names(), ShmemInitHash(), smgropen(), transformGraph(), and XLogPrefetcherAllocate().

hash_destroy()

void hash_destroy

(

HTAB *

hashp

)

Definition at line 865 of file dynahash.c.

{
    if (hashp != NULL)
    {
        /* allocation method must be one we know how to free, too */
        Assert(hashp->alloc == DynaHashAlloc);
        /* so this hashtable must have its own context */
        Assert(hashp->hcxt != NULL);
 
        hash_stats(__func__, hashp);
 
        /*
         * Free everything by destroying the hash table's memory context.
         */
        MemoryContextDelete(hashp->hcxt);
    }
}

References HTAB::alloc, Assert, DynaHashAlloc(), fb(), hash_stats(), HTAB::hcxt, and MemoryContextDelete().

Referenced by _hash_finish_split(), CheckForSessionAndXactLocks(), CleanupListenersOnExit(), CompactCheckpointerRequestQueue(), destroy_colinfo_names_hash(), do_autovacuum(), ExecuteTruncateGuts(), find_all_inheritors(), get_json_object_as_hash(), pg_get_backend_memory_contexts(), pgoutput_memory_context_reset(), populate_recordset_object_end(), PostPrepare_PredicateLocks(), ProcessSyncingTablesForApply(), ReleasePredicateLocksLocal(), ReorderBufferFreeTXN(), ReorderBufferToastReset(), ReorderBufferTruncateTXN(), ResetSequenceCaches(), ResetUnloggedRelationsInDbspaceDir(), SerializePendingSyncs(), set_rtable_names(), ShutdownRecoveryTransactionEnvironment(), XLogCheckInvalidPages(), and XLogPrefetcherFree().

hash_estimate_size()

Size hash_estimate_size	(	int64	num_entries,
		Size	entrysize
	)

Definition at line 783 of file dynahash.c.

{
    Size        size;
    int64       nBuckets,
                nSegments,
                nDirEntries,
                nElementAllocs,
                elementSize,
                elementAllocCnt;
 
    /* estimate number of buckets wanted */
    nBuckets = next_pow2_int64(num_entries);
    /* # of segments needed for nBuckets */
    nSegments = next_pow2_int64((nBuckets - 1) / DEF_SEGSIZE + 1);
    /* directory entries */
    nDirEntries = DEF_DIRSIZE;
    while (nDirEntries < nSegments)
        nDirEntries <<= 1;      /* dir_alloc doubles dsize at each call */
 
    /* fixed control info */
    size = MAXALIGN(sizeof(HASHHDR));   /* but not HTAB, per above */
    /* directory */
    size = add_size(size, mul_size(nDirEntries, sizeof(HASHSEGMENT)));
    /* segments */
    size = add_size(size, mul_size(nSegments,
                                   MAXALIGN(DEF_SEGSIZE * sizeof(HASHBUCKET))));
    /* elements --- allocated in groups of choose_nelem_alloc() entries */
    elementAllocCnt = choose_nelem_alloc(entrysize);
    nElementAllocs = (num_entries - 1) / elementAllocCnt + 1;
    elementSize = MAXALIGN(sizeof(HASHELEMENT)) + MAXALIGN(entrysize);
    size = add_size(size,
                    mul_size(nElementAllocs,
                             mul_size(elementAllocCnt, elementSize)));
 
    return size;
}

References add_size(), choose_nelem_alloc(), DEF_DIRSIZE, DEF_SEGSIZE, fb(), MAXALIGN, mul_size(), and next_pow2_int64().

Referenced by BufTableShmemSize(), CalculateShmemSize(), LockManagerShmemSize(), pgss_memsize(), PredicateLockShmemSize(), and WaitEventCustomShmemSize().

hash_freeze()

void hash_freeze

(

HTAB *

hashp

)

Definition at line 1529 of file dynahash.c.

{
    if (hashp->isshared)
        elog(ERROR, "cannot freeze shared hashtable \"%s\"", hashp->tabname);
    if (!hashp->frozen && has_seq_scans(hashp))
        elog(ERROR, "cannot freeze hashtable \"%s\" because it has active scans",
             hashp->tabname);
    hashp->frozen = true;
}

References elog, ERROR, HTAB::frozen, has_seq_scans(), HTAB::isshared, and HTAB::tabname.

hash_get_num_entries()

int64 hash_get_num_entries

(

HTAB *

hashp

)

Definition at line 1336 of file dynahash.c.

{
    int         i;
    int64       sum = hashp->hctl->freeList[0].nentries;
 
    /*
     * We currently don't bother with acquiring the mutexes; it's only
     * sensible to call this function if you've got lock on all partitions of
     * the table.
     */
    if (IS_PARTITIONED(hashp->hctl))
    {
        for (i = 1; i < NUM_FREELISTS; i++)
            sum += hashp->hctl->freeList[i].nentries;
    }
 
    return sum;
}

References HASHHDR::freeList, HTAB::hctl, i, IS_PARTITIONED, FreeListData::nentries, and NUM_FREELISTS.

Referenced by build_guc_variables(), compute_array_stats(), compute_tsvector_stats(), entry_alloc(), entry_dealloc(), entry_reset(), EstimatePendingSyncsSpace(), EstimateUncommittedEnumsSpace(), get_crosstab_tuplestore(), get_explain_guc_options(), get_guc_variables(), GetLockStatusData(), GetPredicateLockStatusData(), GetRunningTransactionLocks(), GetWaitEventCustomNames(), hash_stats(), pgss_shmem_shutdown(), ResetUnloggedRelationsInDbspaceDir(), SerializePendingSyncs(), transformGraph(), and XLogHaveInvalidPages().

hash_get_shared_size()

Size hash_get_shared_size	(	HASHCTL *	info,
		int	flags
	)

Definition at line 854 of file dynahash.c.

{
    Assert(flags & HASH_DIRSIZE);
    Assert(info->dsize == info->max_dsize);
    return sizeof(HASHHDR) + info->dsize * sizeof(HASHSEGMENT);
}

References Assert, HASHCTL::dsize, HASH_DIRSIZE, and HASHCTL::max_dsize.

Referenced by ShmemInitHash().

hash_initial_lookup()

static uint32 hash_initial_lookup ( HTAB *  hashp, uint32  hashvalue, HASHBUCKET **  bucketptr  )

inlinestatic

Definition at line 1779 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT segp;
    int64       segment_num;
    int64       segment_ndx;
    uint32      bucket;
 
    bucket = calc_bucket(hctl, hashvalue);
 
    segment_num = bucket >> hashp->sshift;
    segment_ndx = MOD(bucket, hashp->ssize);
 
    segp = hashp->dir[segment_num];
 
    if (segp == NULL)
        hash_corrupted(hashp);
 
    *bucketptr = &segp[segment_ndx];
    return bucket;
}

References calc_bucket(), HTAB::dir, fb(), hash_corrupted(), HTAB::hctl, MOD, HTAB::sshift, and HTAB::ssize.

Referenced by hash_search_with_hash_value(), hash_seq_init_with_hash_value(), and hash_update_hash_key().

hash_search()

void * hash_search	(	HTAB *	hashp,
		const void *	keyPtr,
		HASHACTION	action,
		bool *	foundPtr
	)

Definition at line 952 of file dynahash.c.

{
    return hash_search_with_hash_value(hashp,
                                       keyPtr,
                                       hashp->hash(keyPtr, hashp->keysize),
                                       action,
                                       foundPtr);
}

References fb(), HTAB::hash, hash_search_with_hash_value(), and HTAB::keysize.

Referenced by _hash_finish_split(), _hash_splitbucket(), add_guc_variable(), add_join_rel(), add_to_names_hash(), AddEnumLabel(), AddEventToPendingNotifies(), AddPendingSync(), ApplyLogicalMappingFile(), ApplyPendingListenActions(), assign_record_type_typmod(), AsyncExistsPendingNotify(), AtEOSubXact_RelationCache(), AtEOXact_RelationCache(), build_guc_variables(), build_join_rel_hash(), BuildEventTriggerCache(), cfunc_hashtable_delete(), cfunc_hashtable_insert(), cfunc_hashtable_lookup(), CheckAndPromotePredicateLockRequest(), CheckForSerializableConflictOut(), CheckForSessionAndXactLocks(), colname_is_unique(), CompactCheckpointerRequestQueue(), compile_plperl_function(), compile_pltcl_function(), compute_array_stats(), compute_tsvector_stats(), createNewConnection(), define_custom_variable(), delete_rel_type_cache_if_needed(), deleteConnection(), do_autovacuum(), DropAllPredicateLocksFromTable(), DropAllPreparedStatements(), DropPreparedStatement(), entry_alloc(), entry_dealloc(), entry_reset(), EnumTypeUncommitted(), EnumUncommitted(), EnumValuesCreate(), EventCacheLookup(), ExecInitModifyTable(), ExecLookupResultRelByOid(), ExecuteTruncateGuts(), ExtendBufferedRelLocal(), FetchPreparedStatement(), finalize_in_progress_typentries(), find_all_inheritors(), find_join_rel(), find_oper_cache_entry(), find_option(), find_relation_notnullatts(), find_rendezvous_variable(), forget_invalid_pages(), forget_invalid_pages_db(), ForgetPrivateRefCountEntry(), get_attribute_options(), get_cast_hashentry(), get_rel_sync_entry(), get_relation_notnullatts(), get_tablespace(), GetComboCommandId(), GetConnection(), getConnectionByName(), GetExtensibleNodeEntry(), getmissingattr(), GetPrivateRefCountEntrySlow(), getState(), GetWaitEventCustomIdentifier(), gistGetNodeBuffer(), gistGetParent(), gistMemorizeParent(), gistRelocateBuildBuffersOnSplit(), hash_object_field_end(), init_sequence(), insert_rel_type_cache_if_needed(), InvalidateAttoptCacheCallback(), InvalidateLocalBuffer(), InvalidateOprCacheCallBack(), InvalidateShippableCacheCallback(), InvalidateTableSpaceCacheCallback(), is_shippable(), IsListeningOn(), JsObjectGetField(), json_unique_check_key(), LocalBufferAlloc(), LockAcquireExtended(), LockHasWaiters(), LockHeldByMe(), LockRelease(), log_invalid_page(), logical_rewrite_log_mapping(), logicalrep_partition_open(), logicalrep_rel_open(), logicalrep_relmap_update(), lookup_C_func(), lookup_proof_cache(), lookup_ts_config_cache(), lookup_ts_dictionary_cache(), lookup_ts_parser_cache(), lookup_type_cache(), LookupOpclassInfo(), make_oper_cache_entry(), MarkGUCPrefixReserved(), pa_allocate_worker(), pa_find_worker(), pa_free_worker(), PartitionDirectoryLookup(), pg_get_backend_memory_contexts(), pg_tzset(), pgss_store(), plperl_spi_exec_prepared(), plperl_spi_freeplan(), plperl_spi_prepare(), plperl_spi_query_prepared(), pltcl_fetch_interp(), PLy_commit(), PLy_generate_spi_exceptions(), PLy_procedure_get(), PLy_rollback(), PLy_spi_subtransaction_abort(), populate_recordset_object_field_end(), predicatelock_twophase_recover(), PredicateLockExists(), PredicateLockShmemInit(), PredicateLockTwoPhaseFinish(), PrefetchLocalBuffer(), PrepareTableEntriesForListen(), PrepareTableEntriesForUnlisten(), PrepareTableEntriesForUnlistenAll(), ProcessSyncingTablesForApply(), ProcessSyncRequests(), prune_element_hashtable(), prune_lexemes_hashtable(), PutMemoryContextsStatsTupleStore(), rebuild_database_list(), record_C_func(), RegisterExtensibleNodeEntry(), RegisterPredicateLockingXid(), rel_sync_cache_relation_cb(), RelationPreTruncate(), ReleaseOneSerializableXact(), RelFileLocatorSkippingWAL(), RelfilenumberMapInvalidateCallback(), RelidByRelfilenumber(), RememberSyncRequest(), RemoveLocalLock(), ReorderBufferBuildTupleCidHash(), ReorderBufferCleanupTXN(), ReorderBufferToastAppendChunk(), ReorderBufferToastReplace(), ReorderBufferTXNByXid(), ReservePrivateRefCountEntry(), ResetUnloggedRelationsInDbspaceDir(), ResolveCminCmaxDuringDecoding(), RestoreUncommittedEnums(), rewrite_heap_dead_tuple(), rewrite_heap_tuple(), ri_FetchPreparedPlan(), ri_HashCompareOp(), ri_HashPreparedPlan(), ri_LoadConstraintInfo(), select_perl_context(), SerializePendingSyncs(), set_rtable_names(), ShmemInitStruct(), smgrdestroy(), smgrDoPendingSyncs(), smgropen(), smgrreleaserellocator(), StandbyAcquireAccessExclusiveLock(), StandbyReleaseAllLocks(), StandbyReleaseLocks(), StandbyReleaseOldLocks(), StandbyReleaseXidEntryLocks(), StorePreparedStatement(), table_recheck_autovac(), TypeCacheRelCallback(), WaitEventCustomNew(), XLogPrefetcherAddFilter(), XLogPrefetcherCompleteFilters(), and XLogPrefetcherIsFiltered().

hash_search_with_hash_value()

void * hash_search_with_hash_value	(	HTAB *	hashp,
		const void *	keyPtr,
		uint32	hashvalue,
		HASHACTION	action,
		bool *	foundPtr
	)

Definition at line 965 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    int         freelist_idx = FREELIST_IDX(hctl, hashvalue);
    Size        keysize;
    HASHBUCKET  currBucket;
    HASHBUCKET *prevBucketPtr;
    HashCompareFunc match;
 
#ifdef HASH_STATISTICS
    hctl->accesses++;
#endif
 
    /*
     * If inserting, check if it is time to split a bucket.
     *
     * NOTE: failure to expand table is not a fatal error, it just means we
     * have to run at higher fill factor than we wanted.  However, if we're
     * using the palloc allocator then it will throw error anyway on
     * out-of-memory, so we must do this before modifying the table.
     */
    if (action == HASH_ENTER || action == HASH_ENTER_NULL)
    {
        /*
         * Can't split if running in partitioned mode, nor if frozen, nor if
         * table is the subject of any active hash_seq_search scans.
         */
        if (hctl->freeList[0].nentries > (int64) hctl->max_bucket &&
            !IS_PARTITIONED(hctl) && !hashp->frozen &&
            !has_seq_scans(hashp))
            (void) expand_table(hashp);
    }
 
    /*
     * Do the initial lookup
     */
    (void) hash_initial_lookup(hashp, hashvalue, &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    /*
     * Follow collision chain looking for matching key
     */
    match = hashp->match;       /* save one fetch in inner loop */
    keysize = hashp->keysize;   /* ditto */
 
    while (currBucket != NULL)
    {
        if (currBucket->hashvalue == hashvalue &&
            match(ELEMENTKEY(currBucket), keyPtr, keysize) == 0)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
#ifdef HASH_STATISTICS
        hctl->collisions++;
#endif
    }
 
    if (foundPtr)
        *foundPtr = (bool) (currBucket != NULL);
 
    /*
     * OK, now what?
     */
    switch (action)
    {
        case HASH_FIND:
            if (currBucket != NULL)
                return ELEMENTKEY(currBucket);
            return NULL;
 
        case HASH_REMOVE:
            if (currBucket != NULL)
            {
                /* if partitioned, must lock to touch nentries and freeList */
                if (IS_PARTITIONED(hctl))
                    SpinLockAcquire(&(hctl->freeList[freelist_idx].mutex));
 
                /* delete the record from the appropriate nentries counter. */
                Assert(hctl->freeList[freelist_idx].nentries > 0);
                hctl->freeList[freelist_idx].nentries--;
 
                /* remove record from hash bucket's chain. */
                *prevBucketPtr = currBucket->link;
 
                /* add the record to the appropriate freelist. */
                currBucket->link = hctl->freeList[freelist_idx].freeList;
                hctl->freeList[freelist_idx].freeList = currBucket;
 
                if (IS_PARTITIONED(hctl))
                    SpinLockRelease(&hctl->freeList[freelist_idx].mutex);
 
                /*
                 * better hope the caller is synchronizing access to this
                 * element, because someone else is going to reuse it the next
                 * time something is added to the table
                 */
                return ELEMENTKEY(currBucket);
            }
            return NULL;
 
        case HASH_ENTER:
        case HASH_ENTER_NULL:
            /* Return existing element if found, else create one */
            if (currBucket != NULL)
                return ELEMENTKEY(currBucket);
 
            /* disallow inserts if frozen */
            if (hashp->frozen)
                elog(ERROR, "cannot insert into frozen hashtable \"%s\"",
                     hashp->tabname);
 
            currBucket = get_hash_entry(hashp, freelist_idx);
            if (currBucket == NULL)
            {
                /* out of memory */
                if (action == HASH_ENTER_NULL)
                    return NULL;
                /* report a generic message */
                if (hashp->isshared)
                    ereport(ERROR,
                            (errcode(ERRCODE_OUT_OF_MEMORY),
                             errmsg("out of shared memory")));
                else
                    ereport(ERROR,
                            (errcode(ERRCODE_OUT_OF_MEMORY),
                             errmsg("out of memory")));
            }
 
            /* link into hashbucket chain */
            *prevBucketPtr = currBucket;
            currBucket->link = NULL;
 
            /* copy key into record */
            currBucket->hashvalue = hashvalue;
            hashp->keycopy(ELEMENTKEY(currBucket), keyPtr, keysize);
 
            /*
             * Caller is expected to fill the data field on return.  DO NOT
             * insert any code that could possibly throw error here, as doing
             * so would leave the table entry incomplete and hence corrupt the
             * caller's data structure.
             */
 
            return ELEMENTKEY(currBucket);
    }
 
    elog(ERROR, "unrecognized hash action code: %d", (int) action);
 
    return NULL;                /* keep compiler quiet */
}

References Assert, ELEMENTKEY, elog, ereport, errcode(), errmsg(), ERROR, expand_table(), fb(), FreeListData::freeList, HASHHDR::freeList, FREELIST_IDX, HTAB::frozen, get_hash_entry(), has_seq_scans(), HASH_ENTER, HASH_ENTER_NULL, HASH_FIND, hash_initial_lookup(), HASH_REMOVE, HTAB::hctl, IS_PARTITIONED, HTAB::isshared, HTAB::keycopy, HTAB::keysize, HTAB::match, HASHHDR::max_bucket, FreeListData::mutex, FreeListData::nentries, SpinLockAcquire, SpinLockRelease, and HTAB::tabname.

Referenced by BufTableDelete(), BufTableInsert(), BufTableLookup(), CheckTargetForConflictsIn(), CleanUpLock(), ClearOldPredicateLocks(), CreatePredicateLock(), DecrementParentLocks(), DeleteChildTargetLocks(), DeleteLockTarget(), DropAllPredicateLocksFromTable(), FastPathGetRelationLockEntry(), GetLockConflicts(), hash_search(), lock_twophase_recover(), LockAcquireExtended(), LockRefindAndRelease(), LockRelease(), LockWaiterCount(), PageIsPredicateLocked(), PredicateLockAcquire(), ReleaseOneSerializableXact(), RemoveScratchTarget(), RemoveTargetIfNoLongerUsed(), RestoreScratchTarget(), SetupLockInTable(), and TransferPredicateLocksToNewTarget().

hash_select_dirsize()

int64 hash_select_dirsize

(

int64

num_entries

)

Definition at line 830 of file dynahash.c.

{
    int64       nBuckets,
                nSegments,
                nDirEntries;
 
    /* estimate number of buckets wanted */
    nBuckets = next_pow2_int64(num_entries);
    /* # of segments needed for nBuckets */
    nSegments = next_pow2_int64((nBuckets - 1) / DEF_SEGSIZE + 1);
    /* directory entries */
    nDirEntries = DEF_DIRSIZE;
    while (nDirEntries < nSegments)
        nDirEntries <<= 1;      /* dir_alloc doubles dsize at each call */
 
    return nDirEntries;
}

References DEF_DIRSIZE, DEF_SEGSIZE, fb(), and next_pow2_int64().

Referenced by ShmemInitHash().

hash_seq_init()

void hash_seq_init	(	HASH_SEQ_STATUS *	status,
		HTAB *	hashp
	)

Definition at line 1380 of file dynahash.c.

{
    status->hashp = hashp;
    status->curBucket = 0;
    status->curEntry = NULL;
    status->hasHashvalue = false;
    if (!hashp->frozen)
        register_seq_scan(hashp);
}

References HASH_SEQ_STATUS::curBucket, HASH_SEQ_STATUS::curEntry, fb(), HTAB::frozen, HASH_SEQ_STATUS::hasHashvalue, HASH_SEQ_STATUS::hashp, and register_seq_scan().

Referenced by ApplyPendingListenActions(), AtAbort_Portals(), AtCleanup_Portals(), AtEOSubXact_RelationCache(), AtEOXact_RelationCache(), AtPrepare_Locks(), AtSubAbort_Portals(), AtSubCleanup_Portals(), AtSubCommit_Portals(), BeginReportingGUCOptions(), CheckForBufferLeaks(), CheckForSessionAndXactLocks(), CheckTableForSerializableConflictIn(), cleanup_rel_sync_cache(), compute_array_stats(), compute_tsvector_stats(), dblink_get_connections(), DestroyPartitionDirectory(), disconnect_cached_connections(), DropAllPredicateLocksFromTable(), DropAllPreparedStatements(), end_heap_rewrite(), entry_dealloc(), entry_reset(), ExecuteTruncateGuts(), forget_invalid_pages(), forget_invalid_pages_db(), ForgetPortalSnapshots(), gc_qtexts(), get_guc_variables(), GetLockStatusData(), GetPredicateLockStatusData(), GetRunningTransactionLocks(), GetWaitEventCustomNames(), hash_seq_init_with_hash_value(), HoldPinnedPortals(), InitializeGUCOptions(), InvalidateAttoptCacheCallback(), InvalidateOprCacheCallBack(), InvalidateOprProofCacheCallBack(), InvalidateShippableCacheCallback(), InvalidateTableSpaceCacheCallback(), InvalidateTSCacheCallBack(), LockReassignCurrentOwner(), LockReleaseAll(), LockReleaseCurrentOwner(), LockReleaseSession(), logical_end_heap_rewrite(), logical_heap_rewrite_flush_mappings(), logicalrep_partmap_invalidate_cb(), logicalrep_partmap_reset_relmap(), logicalrep_relmap_invalidate_cb(), MarkGUCPrefixReserved(), packGraph(), pg_cursor(), pg_get_shmem_allocations(), pg_get_shmem_allocations_numa(), pg_listening_channels(), pg_prepared_statement(), pg_stat_statements_internal(), pgfdw_inval_callback(), pgfdw_subxact_callback(), pgfdw_xact_callback(), pgss_shmem_shutdown(), plperl_fini(), PortalErrorCleanup(), PortalHashTableDeleteAll(), postgres_fdw_get_connections_internal(), PostPrepare_Locks(), PreCommit_Notify(), PreCommit_Portals(), PrepareTableEntriesForUnlistenAll(), ProcessConfigFileInternal(), ProcessSyncRequests(), prune_element_hashtable(), prune_lexemes_hashtable(), rebuild_database_list(), rel_sync_cache_publication_cb(), rel_sync_cache_relation_cb(), RelationCacheInitializePhase3(), RelationCacheInvalidate(), RelfilenumberMapInvalidateCallback(), RememberSyncRequest(), ReorderBufferToastReset(), selectColorTrigrams(), SerializePendingSyncs(), SerializeUncommittedEnums(), smgrDoPendingSyncs(), smgrreleaseall(), StandbyReleaseAllLocks(), StandbyReleaseOldLocks(), ThereAreNoReadyPortals(), TypeCacheOpcCallback(), TypeCacheRelCallback(), TypeCacheTypCallback(), write_relcache_init_file(), and XLogCheckInvalidPages().

hash_seq_init_with_hash_value()

void hash_seq_init_with_hash_value	(	HASH_SEQ_STATUS *	status,
		HTAB *	hashp,
		uint32	hashvalue
	)

Definition at line 1400 of file dynahash.c.

{
    HASHBUCKET *bucketPtr;
 
    hash_seq_init(status, hashp);
 
    status->hasHashvalue = true;
    status->hashvalue = hashvalue;
 
    status->curBucket = hash_initial_lookup(hashp, hashvalue, &bucketPtr);
    status->curEntry = *bucketPtr;
}

References HASH_SEQ_STATUS::curBucket, HASH_SEQ_STATUS::curEntry, fb(), hash_initial_lookup(), hash_seq_init(), HASH_SEQ_STATUS::hasHashvalue, and HASH_SEQ_STATUS::hashvalue.

Referenced by InvalidateAttoptCacheCallback(), and TypeCacheTypCallback().

hash_seq_search()

void * hash_seq_search

(

HASH_SEQ_STATUS *

status

)

Definition at line 1415 of file dynahash.c.

{
    HTAB       *hashp;
    HASHHDR    *hctl;
    uint32      max_bucket;
    int64       ssize;
    int64       segment_num;
    int64       segment_ndx;
    HASHSEGMENT segp;
    uint32      curBucket;
    HASHELEMENT *curElem;
 
    if (status->hasHashvalue)
    {
        /*
         * Scan entries only in the current bucket because only this bucket
         * can contain entries with the given hash value.
         */
        while ((curElem = status->curEntry) != NULL)
        {
            status->curEntry = curElem->link;
            if (status->hashvalue != curElem->hashvalue)
                continue;
            return (void *) ELEMENTKEY(curElem);
        }
 
        hash_seq_term(status);
        return NULL;
    }
 
    if ((curElem = status->curEntry) != NULL)
    {
        /* Continuing scan of curBucket... */
        status->curEntry = curElem->link;
        if (status->curEntry == NULL)   /* end of this bucket */
            ++status->curBucket;
        return ELEMENTKEY(curElem);
    }
 
    /*
     * Search for next nonempty bucket starting at curBucket.
     */
    curBucket = status->curBucket;
    hashp = status->hashp;
    hctl = hashp->hctl;
    ssize = hashp->ssize;
    max_bucket = hctl->max_bucket;
 
    if (curBucket > max_bucket)
    {
        hash_seq_term(status);
        return NULL;            /* search is done */
    }
 
    /*
     * first find the right segment in the table directory.
     */
    segment_num = curBucket >> hashp->sshift;
    segment_ndx = MOD(curBucket, ssize);
 
    segp = hashp->dir[segment_num];
 
    /*
     * Pick up the first item in this bucket's chain.  If chain is not empty
     * we can begin searching it.  Otherwise we have to advance to find the
     * next nonempty bucket.  We try to optimize that case since searching a
     * near-empty hashtable has to iterate this loop a lot.
     */
    while ((curElem = segp[segment_ndx]) == NULL)
    {
        /* empty bucket, advance to next */
        if (++curBucket > max_bucket)
        {
            status->curBucket = curBucket;
            hash_seq_term(status);
            return NULL;        /* search is done */
        }
        if (++segment_ndx >= ssize)
        {
            segment_num++;
            segment_ndx = 0;
            segp = hashp->dir[segment_num];
        }
    }
 
    /* Begin scan of curBucket... */
    status->curEntry = curElem->link;
    if (status->curEntry == NULL)   /* end of this bucket */
        ++curBucket;
    status->curBucket = curBucket;
    return ELEMENTKEY(curElem);
}

References HASH_SEQ_STATUS::curBucket, HASH_SEQ_STATUS::curEntry, HTAB::dir, ELEMENTKEY, fb(), hash_seq_term(), HASH_SEQ_STATUS::hasHashvalue, HASH_SEQ_STATUS::hashp, HASH_SEQ_STATUS::hashvalue, HTAB::hctl, HASHELEMENT::link, HASHHDR::max_bucket, MOD, HTAB::sshift, and HTAB::ssize.

Referenced by ApplyPendingListenActions(), AtAbort_Portals(), AtCleanup_Portals(), AtEOSubXact_RelationCache(), AtEOXact_RelationCache(), AtPrepare_Locks(), AtSubAbort_Portals(), AtSubCleanup_Portals(), AtSubCommit_Portals(), BeginReportingGUCOptions(), CheckForBufferLeaks(), CheckForSessionAndXactLocks(), CheckTableForSerializableConflictIn(), cleanup_rel_sync_cache(), compute_array_stats(), compute_tsvector_stats(), dblink_get_connections(), DestroyPartitionDirectory(), disconnect_cached_connections(), DropAllPredicateLocksFromTable(), DropAllPreparedStatements(), end_heap_rewrite(), entry_dealloc(), entry_reset(), ExecuteTruncateGuts(), forget_invalid_pages(), forget_invalid_pages_db(), ForgetPortalSnapshots(), gc_qtexts(), get_guc_variables(), GetLockStatusData(), GetPredicateLockStatusData(), GetRunningTransactionLocks(), GetWaitEventCustomNames(), HoldPinnedPortals(), InitializeGUCOptions(), InvalidateAttoptCacheCallback(), InvalidateOprCacheCallBack(), InvalidateOprProofCacheCallBack(), InvalidateShippableCacheCallback(), InvalidateTableSpaceCacheCallback(), InvalidateTSCacheCallBack(), LockReassignCurrentOwner(), LockReleaseAll(), LockReleaseCurrentOwner(), LockReleaseSession(), logical_end_heap_rewrite(), logical_heap_rewrite_flush_mappings(), logicalrep_partmap_invalidate_cb(), logicalrep_partmap_reset_relmap(), logicalrep_relmap_invalidate_cb(), MarkGUCPrefixReserved(), packGraph(), pg_cursor(), pg_get_shmem_allocations(), pg_get_shmem_allocations_numa(), pg_listening_channels(), pg_prepared_statement(), pg_stat_statements_internal(), pgfdw_inval_callback(), pgfdw_subxact_callback(), pgfdw_xact_callback(), pgss_shmem_shutdown(), plperl_fini(), PortalErrorCleanup(), PortalHashTableDeleteAll(), postgres_fdw_get_connections_internal(), PostPrepare_Locks(), PreCommit_Notify(), PreCommit_Portals(), PrepareTableEntriesForUnlistenAll(), ProcessConfigFileInternal(), ProcessSyncRequests(), prune_element_hashtable(), prune_lexemes_hashtable(), rebuild_database_list(), rel_sync_cache_publication_cb(), rel_sync_cache_relation_cb(), RelationCacheInitializePhase3(), RelationCacheInvalidate(), RelfilenumberMapInvalidateCallback(), RememberSyncRequest(), ReorderBufferToastReset(), selectColorTrigrams(), SerializePendingSyncs(), SerializeUncommittedEnums(), smgrDoPendingSyncs(), smgrreleaseall(), StandbyReleaseAllLocks(), StandbyReleaseOldLocks(), ThereAreNoReadyPortals(), TypeCacheOpcCallback(), TypeCacheRelCallback(), TypeCacheTypCallback(), write_relcache_init_file(), and XLogCheckInvalidPages().

hash_seq_term()

void hash_seq_term

(

HASH_SEQ_STATUS *

status

)

Definition at line 1509 of file dynahash.c.

{
    if (!status->hashp->frozen)
        deregister_seq_scan(status->hashp);
}

References deregister_seq_scan(), HTAB::frozen, and HASH_SEQ_STATUS::hashp.

Referenced by gc_qtexts(), hash_seq_search(), logicalrep_partmap_invalidate_cb(), logicalrep_relmap_invalidate_cb(), pgss_shmem_shutdown(), PortalHashTableDeleteAll(), PreCommit_Portals(), and RelationCacheInitializePhase3().

hash_stats()

void hash_stats	(	const char *	caller,
		HTAB *	hashp
	)

Definition at line 884 of file dynahash.c.

{
#ifdef HASH_STATISTICS
    HASHHDR    *hctl = hashp->hctl;
 
    elog(DEBUG4,
         "hash_stats:  Caller: %s  Table Name: \"%s\"  Accesses: " UINT64_FORMAT "  Collisions: " UINT64_FORMAT "  Expansions: " UINT64_FORMAT "  Entries: " INT64_FORMAT "  Key Size: %zu  Max Bucket: %u  Segment Count: " INT64_FORMAT,
         caller != NULL ? caller : "(unknown)", hashp->tabname, hctl->accesses,
         hctl->collisions, hctl->expansions, hash_get_num_entries(hashp),
         hctl->keysize, hctl->max_bucket, hctl->nsegs);
#endif
}

References DEBUG4, elog, fb(), hash_get_num_entries(), HTAB::hctl, INT64_FORMAT, HASHHDR::keysize, HASHHDR::max_bucket, HASHHDR::nsegs, HTAB::tabname, and UINT64_FORMAT.

Referenced by hash_destroy().

hash_update_hash_key()

bool hash_update_hash_key	(	HTAB *	hashp,
		void *	existingEntry,
		const void *	newKeyPtr
	)

Definition at line 1140 of file dynahash.c.

{
    HASHELEMENT *existingElement = ELEMENT_FROM_KEY(existingEntry);
    uint32      newhashvalue;
    Size        keysize;
    uint32      bucket;
    uint32      newbucket;
    HASHBUCKET  currBucket;
    HASHBUCKET *prevBucketPtr;
    HASHBUCKET *oldPrevPtr;
    HashCompareFunc match;
 
#ifdef HASH_STATISTICS
    HASHHDR    *hctl = hashp->hctl;
 
    hctl->accesses++;
#endif
 
    /* disallow updates if frozen */
    if (hashp->frozen)
        elog(ERROR, "cannot update in frozen hashtable \"%s\"",
             hashp->tabname);
 
    /*
     * Lookup the existing element using its saved hash value.  We need to do
     * this to be able to unlink it from its hash chain, but as a side benefit
     * we can verify the validity of the passed existingEntry pointer.
     */
    bucket = hash_initial_lookup(hashp, existingElement->hashvalue,
                                 &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    while (currBucket != NULL)
    {
        if (currBucket == existingElement)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
    }
 
    if (currBucket == NULL)
        elog(ERROR, "hash_update_hash_key argument is not in hashtable \"%s\"",
             hashp->tabname);
 
    oldPrevPtr = prevBucketPtr;
 
    /*
     * Now perform the equivalent of a HASH_ENTER operation to locate the hash
     * chain we want to put the entry into.
     */
    newhashvalue = hashp->hash(newKeyPtr, hashp->keysize);
    newbucket = hash_initial_lookup(hashp, newhashvalue, &prevBucketPtr);
    currBucket = *prevBucketPtr;
 
    /*
     * Follow collision chain looking for matching key
     */
    match = hashp->match;       /* save one fetch in inner loop */
    keysize = hashp->keysize;   /* ditto */
 
    while (currBucket != NULL)
    {
        if (currBucket->hashvalue == newhashvalue &&
            match(ELEMENTKEY(currBucket), newKeyPtr, keysize) == 0)
            break;
        prevBucketPtr = &(currBucket->link);
        currBucket = *prevBucketPtr;
#ifdef HASH_STATISTICS
        hctl->collisions++;
#endif
    }
 
    if (currBucket != NULL)
        return false;           /* collision with an existing entry */
 
    currBucket = existingElement;
 
    /*
     * If old and new hash values belong to the same bucket, we need not
     * change any chain links, and indeed should not since this simplistic
     * update will corrupt the list if currBucket is the last element.  (We
     * cannot fall out earlier, however, since we need to scan the bucket to
     * check for duplicate keys.)
     */
    if (bucket != newbucket)
    {
        /* OK to remove record from old hash bucket's chain. */
        *oldPrevPtr = currBucket->link;
 
        /* link into new hashbucket chain */
        *prevBucketPtr = currBucket;
        currBucket->link = NULL;
    }
 
    /* copy new key into record */
    currBucket->hashvalue = newhashvalue;
    hashp->keycopy(ELEMENTKEY(currBucket), newKeyPtr, keysize);
 
    /* rest of record is untouched */
 
    return true;
}

References ELEMENT_FROM_KEY, ELEMENTKEY, elog, ERROR, fb(), HTAB::frozen, HTAB::hash, hash_initial_lookup(), HTAB::hctl, HTAB::keycopy, HTAB::keysize, HTAB::match, and HTAB::tabname.

Referenced by PostPrepare_Locks().

hdefault()

static void hdefault ( HTAB *  hashp )

static

Definition at line 638 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
 
    MemSet(hctl, 0, sizeof(HASHHDR));
 
    hctl->dsize = DEF_DIRSIZE;
    hctl->nsegs = 0;
 
    hctl->num_partitions = 0;   /* not partitioned */
 
    /* table has no fixed maximum size */
    hctl->max_dsize = NO_MAX_DSIZE;
 
    hctl->ssize = DEF_SEGSIZE;
    hctl->sshift = DEF_SEGSIZE_SHIFT;
 
    hctl->isfixed = false;      /* can be enlarged */
 
#ifdef HASH_STATISTICS
    hctl->accesses = hctl->collisions = hctl->expansions = 0;
#endif
}

References DEF_DIRSIZE, DEF_SEGSIZE, DEF_SEGSIZE_SHIFT, HASHHDR::dsize, HTAB::hctl, HASHHDR::isfixed, HASHHDR::max_dsize, MemSet, NO_MAX_DSIZE, HASHHDR::nsegs, HASHHDR::num_partitions, HASHHDR::sshift, and HASHHDR::ssize.

Referenced by hash_create().

init_htab()

static bool init_htab ( HTAB *  hashp, int64  nelem  )

static

Definition at line 700 of file dynahash.c.

{
    HASHHDR    *hctl = hashp->hctl;
    HASHSEGMENT *segp;
    int         nbuckets;
    int         nsegs;
    int         i;
 
    /*
     * initialize mutexes if it's a partitioned table
     */
    if (IS_PARTITIONED(hctl))
        for (i = 0; i < NUM_FREELISTS; i++)
            SpinLockInit(&(hctl->freeList[i].mutex));
 
    /*
     * Allocate space for the next greater power of two number of buckets,
     * assuming a desired maximum load factor of 1.
     */
    nbuckets = next_pow2_int(nelem);
 
    /*
     * In a partitioned table, nbuckets must be at least equal to
     * num_partitions; were it less, keys with apparently different partition
     * numbers would map to the same bucket, breaking partition independence.
     * (Normally nbuckets will be much bigger; this is just a safety check.)
     */
    while (nbuckets < hctl->num_partitions)
        nbuckets <<= 1;
 
    hctl->max_bucket = hctl->low_mask = nbuckets - 1;
    hctl->high_mask = (nbuckets << 1) - 1;
 
    /*
     * Figure number of directory segments needed, round up to a power of 2
     */
    nsegs = (nbuckets - 1) / hctl->ssize + 1;
    nsegs = next_pow2_int(nsegs);
 
    /*
     * Make sure directory is big enough. If pre-allocated directory is too
     * small, choke (caller screwed up).
     */
    if (nsegs > hctl->dsize)
    {
        if (!(hashp->dir))
            hctl->dsize = nsegs;
        else
            return false;
    }
 
    /* Allocate a directory */
    if (!(hashp->dir))
    {
        CurrentDynaHashCxt = hashp->hcxt;
        hashp->dir = (HASHSEGMENT *)
            hashp->alloc(hctl->dsize * sizeof(HASHSEGMENT));
        if (!hashp->dir)
            return false;
    }
 
    /* Allocate initial segments */
    for (segp = hashp->dir; hctl->nsegs < nsegs; hctl->nsegs++, segp++)
    {
        *segp = seg_alloc(hashp);
        if (*segp == NULL)
            return false;
    }
 
    /* Choose number of entries to allocate at a time */
    hctl->nelem_alloc = choose_nelem_alloc(hctl->entrysize);
 
    return true;
}

References HTAB::alloc, choose_nelem_alloc(), CurrentDynaHashCxt, HTAB::dir, HASHHDR::dsize, HASHHDR::entrysize, fb(), HASHHDR::freeList, HTAB::hctl, HTAB::hcxt, HASHHDR::high_mask, i, IS_PARTITIONED, HASHHDR::low_mask, HASHHDR::max_bucket, FreeListData::mutex, HASHHDR::nelem_alloc, next_pow2_int(), HASHHDR::nsegs, NUM_FREELISTS, seg_alloc(), SpinLockInit, and HASHHDR::ssize.

Referenced by hash_create().

my_log2()

static int my_log2 ( int64  num )

static

Definition at line 1817 of file dynahash.c.

{
    /*
     * guard against too-large input, which would be invalid for
     * pg_ceil_log2_*()
     */
    if (num > PG_INT64_MAX / 2)
        num = PG_INT64_MAX / 2;
 
    return pg_ceil_log2_64(num);
}

References pg_ceil_log2_64(), and PG_INT64_MAX.

Referenced by hash_create(), next_pow2_int(), and next_pow2_int64().

next_pow2_int()

static int next_pow2_int ( int64  num )

static

Definition at line 1839 of file dynahash.c.

{
    if (num > INT_MAX / 2)
        num = INT_MAX / 2;
    return 1 << my_log2(num);
}

References fb(), and my_log2().

Referenced by hash_create(), and init_htab().

next_pow2_int64()

static int64 next_pow2_int64 ( int64  num )

static

Definition at line 1831 of file dynahash.c.

{
    /* my_log2's internal range check is sufficient */
    return 1L << my_log2(num);
}

References my_log2().

Referenced by hash_estimate_size(), and hash_select_dirsize().

register_seq_scan()

static void register_seq_scan ( HTAB *  hashp )

static

Definition at line 1884 of file dynahash.c.

{
    if (num_seq_scans >= MAX_SEQ_SCANS)
        elog(ERROR, "too many active hash_seq_search scans, cannot start one on \"%s\"",
             hashp->tabname);
    seq_scan_tables[num_seq_scans] = hashp;
    seq_scan_level[num_seq_scans] = GetCurrentTransactionNestLevel();
    num_seq_scans++;
}

References elog, ERROR, GetCurrentTransactionNestLevel(), MAX_SEQ_SCANS, num_seq_scans, seq_scan_level, seq_scan_tables, and HTAB::tabname.

Referenced by hash_seq_init().

seg_alloc()

static HASHSEGMENT seg_alloc ( HTAB *  hashp )

static

Definition at line 1682 of file dynahash.c.

{
    HASHSEGMENT segp;
 
    CurrentDynaHashCxt = hashp->hcxt;
    segp = (HASHSEGMENT) hashp->alloc(sizeof(HASHBUCKET) * hashp->ssize);
 
    if (!segp)
        return NULL;
 
    MemSet(segp, 0, sizeof(HASHBUCKET) * hashp->ssize);
 
    return segp;
}

References HTAB::alloc, CurrentDynaHashCxt, fb(), HTAB::hcxt, MemSet, and HTAB::ssize.

Referenced by expand_table(), and init_htab().

string_compare()

static int string_compare ( const char *  key1, const char *  key2, Size  keysize  )

static

Definition at line 313 of file dynahash.c.

{
    return strncmp(key1, key2, keysize - 1);
}

References fb().

Referenced by hash_create().

Variable Documentation

CurrentDynaHashCxt

MemoryContext CurrentDynaHashCxt = NULL

static

Definition at line 294 of file dynahash.c.

Referenced by dir_realloc(), DynaHashAlloc(), element_alloc(), hash_create(), init_htab(), and seg_alloc().

num_seq_scans

int num_seq_scans = 0

static

Definition at line 1879 of file dynahash.c.

Referenced by AtEOSubXact_HashTables(), AtEOXact_HashTables(), deregister_seq_scan(), has_seq_scans(), and register_seq_scan().

seq_scan_level

int seq_scan_level[MAX_SEQ_SCANS]

static

Definition at line 1878 of file dynahash.c.

Referenced by AtEOSubXact_HashTables(), deregister_seq_scan(), and register_seq_scan().

seq_scan_tables

HTAB* seq_scan_tables[MAX_SEQ_SCANS]

static

Definition at line 1877 of file dynahash.c.

Referenced by AtEOSubXact_HashTables(), AtEOXact_HashTables(), deregister_seq_scan(), has_seq_scans(), and register_seq_scan().