rewriteheap.c File Reference

#include "postgres.h"
#include <unistd.h>
#include "access/heapam.h"
#include "access/heapam_xlog.h"
#include "access/heaptoast.h"
#include "access/rewriteheap.h"
#include "access/transam.h"
#include "access/xact.h"
#include "access/xloginsert.h"
#include "catalog/catalog.h"
#include "common/file_utils.h"
#include "lib/ilist.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "replication/logical.h"
#include "replication/slot.h"
#include "storage/bufmgr.h"
#include "storage/fd.h"
#include "storage/procarray.h"
#include "storage/smgr.h"
#include "utils/memutils.h"
#include "utils/rel.h"

Include dependency graph for rewriteheap.c:

Go to the source code of this file.

Data Structures
struct	RewriteStateData
struct	TidHashKey
struct	UnresolvedTupData
struct	OldToNewMappingData
struct	RewriteMappingFile
struct	RewriteMappingDataEntry

Typedefs
typedef struct RewriteStateData	RewriteStateData
typedef UnresolvedTupData *	UnresolvedTup
typedef OldToNewMappingData *	OldToNewMapping
typedef struct RewriteMappingFile	RewriteMappingFile
typedef struct RewriteMappingDataEntry	RewriteMappingDataEntry

Functions
static void	raw_heap_insert (RewriteState state, HeapTuple tup)
static void	logical_begin_heap_rewrite (RewriteState state)
static void	logical_rewrite_heap_tuple (RewriteState state, ItemPointerData old_tid, HeapTuple new_tuple)
static void	logical_end_heap_rewrite (RewriteState state)
RewriteState	begin_heap_rewrite (Relation old_heap, Relation new_heap, TransactionId oldest_xmin, TransactionId freeze_xid, MultiXactId cutoff_multi)
void	end_heap_rewrite (RewriteState state)
void	rewrite_heap_tuple (RewriteState state, HeapTuple old_tuple, HeapTuple new_tuple)
bool	rewrite_heap_dead_tuple (RewriteState state, HeapTuple old_tuple)
static void	logical_heap_rewrite_flush_mappings (RewriteState state)
static void	logical_rewrite_log_mapping (RewriteState state, TransactionId xid, LogicalRewriteMappingData *map)
void	heap_xlog_logical_rewrite (XLogReaderState *r)
void	CheckPointLogicalRewriteHeap (void)

Typedef Documentation

OldToNewMapping

typedef OldToNewMappingData* OldToNewMapping

Definition at line 187 of file rewriteheap.c.

RewriteMappingDataEntry

typedef struct RewriteMappingDataEntry RewriteMappingDataEntry

RewriteMappingFile

typedef struct RewriteMappingFile RewriteMappingFile

RewriteStateData

typedef struct RewriteStateData RewriteStateData

UnresolvedTup

typedef UnresolvedTupData* UnresolvedTup

Definition at line 179 of file rewriteheap.c.

Function Documentation

begin_heap_rewrite()

RewriteState begin_heap_rewrite	(	Relation	old_heap,
		Relation	new_heap,
		TransactionId	oldest_xmin,
		TransactionId	freeze_xid,
		MultiXactId	cutoff_multi
	)

Definition at line 236 of file rewriteheap.c.

{
    RewriteState state;
    MemoryContext rw_cxt;
    MemoryContext old_cxt;
    HASHCTL     hash_ctl;
 
    /*
     * To ease cleanup, make a separate context that will contain the
     * RewriteState struct itself plus all subsidiary data.
     */
    rw_cxt = AllocSetContextCreate(CurrentMemoryContext,
                                   "Table rewrite",
                                   ALLOCSET_DEFAULT_SIZES);
    old_cxt = MemoryContextSwitchTo(rw_cxt);
 
    /* Create and fill in the state struct */
    state = palloc0(sizeof(RewriteStateData));
 
    state->rs_old_rel = old_heap;
    state->rs_new_rel = new_heap;
    state->rs_buffer = (Page) palloc_aligned(BLCKSZ, PG_IO_ALIGN_SIZE, 0);
    /* new_heap needn't be empty, just locked */
    state->rs_blockno = RelationGetNumberOfBlocks(new_heap);
    state->rs_buffer_valid = false;
    state->rs_oldest_xmin = oldest_xmin;
    state->rs_freeze_xid = freeze_xid;
    state->rs_cutoff_multi = cutoff_multi;
    state->rs_cxt = rw_cxt;
 
    /* Initialize hash tables used to track update chains */
    hash_ctl.keysize = sizeof(TidHashKey);
    hash_ctl.entrysize = sizeof(UnresolvedTupData);
    hash_ctl.hcxt = state->rs_cxt;
 
    state->rs_unresolved_tups =
        hash_create("Rewrite / Unresolved ctids",
                    128,        /* arbitrary initial size */
                    &hash_ctl,
                    HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
 
    hash_ctl.entrysize = sizeof(OldToNewMappingData);
 
    state->rs_old_new_tid_map =
        hash_create("Rewrite / Old to new tid map",
                    128,        /* arbitrary initial size */
                    &hash_ctl,
                    HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
 
    MemoryContextSwitchTo(old_cxt);
 
    logical_begin_heap_rewrite(state);
 
    return state;
}

References ALLOCSET_DEFAULT_SIZES, AllocSetContextCreate, CurrentMemoryContext, HASHCTL::entrysize, HASH_BLOBS, HASH_CONTEXT, hash_create(), HASH_ELEM, HASHCTL::hcxt, HASHCTL::keysize, logical_begin_heap_rewrite(), MemoryContextSwitchTo(), palloc0(), palloc_aligned(), PG_IO_ALIGN_SIZE, and RelationGetNumberOfBlocks.

Referenced by heapam_relation_copy_for_cluster().

CheckPointLogicalRewriteHeap()

void CheckPointLogicalRewriteHeap

(

void

)

Definition at line 1189 of file rewriteheap.c.

{
    XLogRecPtr  cutoff;
    XLogRecPtr  redo;
    DIR        *mappings_dir;
    struct dirent *mapping_de;
    char        path[MAXPGPATH + 20];
 
    /*
     * We start of with a minimum of the last redo pointer. No new decoding
     * slot will start before that, so that's a safe upper bound for removal.
     */
    redo = GetRedoRecPtr();
 
    /* now check for the restart ptrs from existing slots */
    cutoff = ReplicationSlotsComputeLogicalRestartLSN();
 
    /* don't start earlier than the restart lsn */
    if (cutoff != InvalidXLogRecPtr && redo < cutoff)
        cutoff = redo;
 
    mappings_dir = AllocateDir("pg_logical/mappings");
    while ((mapping_de = ReadDir(mappings_dir, "pg_logical/mappings")) != NULL)
    {
        Oid         dboid;
        Oid         relid;
        XLogRecPtr  lsn;
        TransactionId rewrite_xid;
        TransactionId create_xid;
        uint32      hi,
                    lo;
        PGFileType  de_type;
 
        if (strcmp(mapping_de->d_name, ".") == 0 ||
            strcmp(mapping_de->d_name, "..") == 0)
            continue;
 
        snprintf(path, sizeof(path), "pg_logical/mappings/%s", mapping_de->d_name);
        de_type = get_dirent_type(path, mapping_de, false, DEBUG1);
 
        if (de_type != PGFILETYPE_ERROR && de_type != PGFILETYPE_REG)
            continue;
 
        /* Skip over files that cannot be ours. */
        if (strncmp(mapping_de->d_name, "map-", 4) != 0)
            continue;
 
        if (sscanf(mapping_de->d_name, LOGICAL_REWRITE_FORMAT,
                   &dboid, &relid, &hi, &lo, &rewrite_xid, &create_xid) != 6)
            elog(ERROR, "could not parse filename \"%s\"", mapping_de->d_name);
 
        lsn = ((uint64) hi) << 32 | lo;
 
        if (lsn < cutoff || cutoff == InvalidXLogRecPtr)
        {
            elog(DEBUG1, "removing logical rewrite file \"%s\"", path);
            if (unlink(path) < 0)
                ereport(ERROR,
                        (errcode_for_file_access(),
                         errmsg("could not remove file \"%s\": %m", path)));
        }
        else
        {
            /* on some operating systems fsyncing a file requires O_RDWR */
            int         fd = OpenTransientFile(path, O_RDWR | PG_BINARY);
 
            /*
             * The file cannot vanish due to concurrency since this function
             * is the only one removing logical mappings and only one
             * checkpoint can be in progress at a time.
             */
            if (fd < 0)
                ereport(ERROR,
                        (errcode_for_file_access(),
                         errmsg("could not open file \"%s\": %m", path)));
 
            /*
             * We could try to avoid fsyncing files that either haven't
             * changed or have only been created since the checkpoint's start,
             * but it's currently not deemed worth the effort.
             */
            pgstat_report_wait_start(WAIT_EVENT_LOGICAL_REWRITE_CHECKPOINT_SYNC);
            if (pg_fsync(fd) != 0)
                ereport(data_sync_elevel(ERROR),
                        (errcode_for_file_access(),
                         errmsg("could not fsync file \"%s\": %m", path)));
            pgstat_report_wait_end();
 
            if (CloseTransientFile(fd) != 0)
                ereport(ERROR,
                        (errcode_for_file_access(),
                         errmsg("could not close file \"%s\": %m", path)));
        }
    }
    FreeDir(mappings_dir);
 
    /* persist directory entries to disk */
    fsync_fname("pg_logical/mappings", true);
}

References AllocateDir(), CloseTransientFile(), dirent::d_name, data_sync_elevel(), DEBUG1, elog(), ereport, errcode_for_file_access(), errmsg(), ERROR, fd(), FreeDir(), fsync_fname(), get_dirent_type(), GetRedoRecPtr(), InvalidXLogRecPtr, LOGICAL_REWRITE_FORMAT, MAXPGPATH, OpenTransientFile(), PG_BINARY, pg_fsync(), PGFILETYPE_ERROR, PGFILETYPE_REG, pgstat_report_wait_end(), pgstat_report_wait_start(), ReadDir(), ReplicationSlotsComputeLogicalRestartLSN(), and snprintf.

Referenced by CheckPointGuts().

end_heap_rewrite()

void end_heap_rewrite

(

RewriteState

state

)

Definition at line 299 of file rewriteheap.c.

{
    HASH_SEQ_STATUS seq_status;
    UnresolvedTup unresolved;
 
    /*
     * Write any remaining tuples in the UnresolvedTups table. If we have any
     * left, they should in fact be dead, but let's err on the safe side.
     */
    hash_seq_init(&seq_status, state->rs_unresolved_tups);
 
    while ((unresolved = hash_seq_search(&seq_status)) != NULL)
    {
        ItemPointerSetInvalid(&unresolved->tuple->t_data->t_ctid);
        raw_heap_insert(state, unresolved->tuple);
    }
 
    /* Write the last page, if any */
    if (state->rs_buffer_valid)
    {
        if (RelationNeedsWAL(state->rs_new_rel))
            log_newpage(&state->rs_new_rel->rd_locator,
                        MAIN_FORKNUM,
                        state->rs_blockno,
                        state->rs_buffer,
                        true);
 
        PageSetChecksumInplace(state->rs_buffer, state->rs_blockno);
 
        smgrextend(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM,
                   state->rs_blockno, state->rs_buffer, true);
    }
 
    /*
     * When we WAL-logged rel pages, we must nonetheless fsync them.  The
     * reason is the same as in storage.c's RelationCopyStorage(): we're
     * writing data that's not in shared buffers, and so a CHECKPOINT
     * occurring during the rewriteheap operation won't have fsync'd data we
     * wrote before the checkpoint.
     */
    if (RelationNeedsWAL(state->rs_new_rel))
        smgrimmedsync(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM);
 
    logical_end_heap_rewrite(state);
 
    /* Deleting the context frees everything */
    MemoryContextDelete(state->rs_cxt);
}

References hash_seq_init(), hash_seq_search(), ItemPointerSetInvalid(), log_newpage(), logical_end_heap_rewrite(), MAIN_FORKNUM, MemoryContextDelete(), PageSetChecksumInplace(), raw_heap_insert(), RelationGetSmgr(), RelationNeedsWAL, smgrextend(), smgrimmedsync(), HeapTupleHeaderData::t_ctid, HeapTupleData::t_data, and UnresolvedTupData::tuple.

Referenced by heapam_relation_copy_for_cluster().

heap_xlog_logical_rewrite()

void heap_xlog_logical_rewrite

(

XLogReaderState *

r

)

Definition at line 1107 of file rewriteheap.c.

{
    char        path[MAXPGPATH];
    int         fd;
    xl_heap_rewrite_mapping *xlrec;
    uint32      len;
    char       *data;
 
    xlrec = (xl_heap_rewrite_mapping *) XLogRecGetData(r);
 
    snprintf(path, MAXPGPATH,
             "pg_logical/mappings/" LOGICAL_REWRITE_FORMAT,
             xlrec->mapped_db, xlrec->mapped_rel,
             LSN_FORMAT_ARGS(xlrec->start_lsn),
             xlrec->mapped_xid, XLogRecGetXid(r));
 
    fd = OpenTransientFile(path,
                           O_CREAT | O_WRONLY | PG_BINARY);
    if (fd < 0)
        ereport(ERROR,
                (errcode_for_file_access(),
                 errmsg("could not create file \"%s\": %m", path)));
 
    /*
     * Truncate all data that's not guaranteed to have been safely fsynced (by
     * previous record or by the last checkpoint).
     */
    pgstat_report_wait_start(WAIT_EVENT_LOGICAL_REWRITE_TRUNCATE);
    if (ftruncate(fd, xlrec->offset) != 0)
        ereport(ERROR,
                (errcode_for_file_access(),
                 errmsg("could not truncate file \"%s\" to %u: %m",
                        path, (uint32) xlrec->offset)));
    pgstat_report_wait_end();
 
    data = XLogRecGetData(r) + sizeof(*xlrec);
 
    len = xlrec->num_mappings * sizeof(LogicalRewriteMappingData);
 
    /* write out tail end of mapping file (again) */
    errno = 0;
    pgstat_report_wait_start(WAIT_EVENT_LOGICAL_REWRITE_MAPPING_WRITE);
    if (pg_pwrite(fd, data, len, xlrec->offset) != len)
    {
        /* if write didn't set errno, assume problem is no disk space */
        if (errno == 0)
            errno = ENOSPC;
        ereport(ERROR,
                (errcode_for_file_access(),
                 errmsg("could not write to file \"%s\": %m", path)));
    }
    pgstat_report_wait_end();
 
    /*
     * Now fsync all previously written data. We could improve things and only
     * do this for the last write to a file, but the required bookkeeping
     * doesn't seem worth the trouble.
     */
    pgstat_report_wait_start(WAIT_EVENT_LOGICAL_REWRITE_MAPPING_SYNC);
    if (pg_fsync(fd) != 0)
        ereport(data_sync_elevel(ERROR),
                (errcode_for_file_access(),
                 errmsg("could not fsync file \"%s\": %m", path)));
    pgstat_report_wait_end();
 
    if (CloseTransientFile(fd) != 0)
        ereport(ERROR,
                (errcode_for_file_access(),
                 errmsg("could not close file \"%s\": %m", path)));
}

References CloseTransientFile(), data, data_sync_elevel(), ereport, errcode_for_file_access(), errmsg(), ERROR, fd(), ftruncate, len, LOGICAL_REWRITE_FORMAT, LSN_FORMAT_ARGS, xl_heap_rewrite_mapping::mapped_db, xl_heap_rewrite_mapping::mapped_rel, xl_heap_rewrite_mapping::mapped_xid, MAXPGPATH, xl_heap_rewrite_mapping::num_mappings, xl_heap_rewrite_mapping::offset, OpenTransientFile(), PG_BINARY, pg_fsync(), pg_pwrite, pgstat_report_wait_end(), pgstat_report_wait_start(), snprintf, xl_heap_rewrite_mapping::start_lsn, XLogRecGetData, and XLogRecGetXid.

Referenced by heap2_redo().

logical_begin_heap_rewrite()

static void logical_begin_heap_rewrite ( RewriteState  state )

static

Definition at line 793 of file rewriteheap.c.

{
    HASHCTL     hash_ctl;
    TransactionId logical_xmin;
 
    /*
     * We only need to persist these mappings if the rewritten table can be
     * accessed during logical decoding, if not, we can skip doing any
     * additional work.
     */
    state->rs_logical_rewrite =
        RelationIsAccessibleInLogicalDecoding(state->rs_old_rel);
 
    if (!state->rs_logical_rewrite)
        return;
 
    ProcArrayGetReplicationSlotXmin(NULL, &logical_xmin);
 
    /*
     * If there are no logical slots in progress we don't need to do anything,
     * there cannot be any remappings for relevant rows yet. The relation's
     * lock protects us against races.
     */
    if (logical_xmin == InvalidTransactionId)
    {
        state->rs_logical_rewrite = false;
        return;
    }
 
    state->rs_logical_xmin = logical_xmin;
    state->rs_begin_lsn = GetXLogInsertRecPtr();
    state->rs_num_rewrite_mappings = 0;
 
    hash_ctl.keysize = sizeof(TransactionId);
    hash_ctl.entrysize = sizeof(RewriteMappingFile);
    hash_ctl.hcxt = state->rs_cxt;
 
    state->rs_logical_mappings =
        hash_create("Logical rewrite mapping",
                    128,        /* arbitrary initial size */
                    &hash_ctl,
                    HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
}

References HASHCTL::entrysize, GetXLogInsertRecPtr(), HASH_BLOBS, HASH_CONTEXT, hash_create(), HASH_ELEM, HASHCTL::hcxt, InvalidTransactionId, HASHCTL::keysize, ProcArrayGetReplicationSlotXmin(), and RelationIsAccessibleInLogicalDecoding.

Referenced by begin_heap_rewrite().

logical_end_heap_rewrite()

static void logical_end_heap_rewrite ( RewriteState  state )

static

Definition at line 939 of file rewriteheap.c.

{
    HASH_SEQ_STATUS seq_status;
    RewriteMappingFile *src;
 
    /* done, no logical rewrite in progress */
    if (!state->rs_logical_rewrite)
        return;
 
    /* writeout remaining in-memory entries */
    if (state->rs_num_rewrite_mappings > 0)
        logical_heap_rewrite_flush_mappings(state);
 
    /* Iterate over all mappings we have written and fsync the files. */
    hash_seq_init(&seq_status, state->rs_logical_mappings);
    while ((src = (RewriteMappingFile *) hash_seq_search(&seq_status)) != NULL)
    {
        if (FileSync(src->vfd, WAIT_EVENT_LOGICAL_REWRITE_SYNC) != 0)
            ereport(data_sync_elevel(ERROR),
                    (errcode_for_file_access(),
                     errmsg("could not fsync file \"%s\": %m", src->path)));
        FileClose(src->vfd);
    }
    /* memory context cleanup will deal with the rest */
}

References data_sync_elevel(), ereport, errcode_for_file_access(), errmsg(), ERROR, FileClose(), FileSync(), hash_seq_init(), hash_seq_search(), logical_heap_rewrite_flush_mappings(), RewriteMappingFile::path, and RewriteMappingFile::vfd.

Referenced by end_heap_rewrite().

logical_heap_rewrite_flush_mappings()

static void logical_heap_rewrite_flush_mappings ( RewriteState  state )

static

Definition at line 841 of file rewriteheap.c.

{
    HASH_SEQ_STATUS seq_status;
    RewriteMappingFile *src;
    dlist_mutable_iter iter;
 
    Assert(state->rs_logical_rewrite);
 
    /* no logical rewrite in progress, no need to iterate over mappings */
    if (state->rs_num_rewrite_mappings == 0)
        return;
 
    elog(DEBUG1, "flushing %u logical rewrite mapping entries",
         state->rs_num_rewrite_mappings);
 
    hash_seq_init(&seq_status, state->rs_logical_mappings);
    while ((src = (RewriteMappingFile *) hash_seq_search(&seq_status)) != NULL)
    {
        char       *waldata;
        char       *waldata_start;
        xl_heap_rewrite_mapping xlrec;
        Oid         dboid;
        uint32      len;
        int         written;
        uint32      num_mappings = dclist_count(&src->mappings);
 
        /* this file hasn't got any new mappings */
        if (num_mappings == 0)
            continue;
 
        if (state->rs_old_rel->rd_rel->relisshared)
            dboid = InvalidOid;
        else
            dboid = MyDatabaseId;
 
        xlrec.num_mappings = num_mappings;
        xlrec.mapped_rel = RelationGetRelid(state->rs_old_rel);
        xlrec.mapped_xid = src->xid;
        xlrec.mapped_db = dboid;
        xlrec.offset = src->off;
        xlrec.start_lsn = state->rs_begin_lsn;
 
        /* write all mappings consecutively */
        len = num_mappings * sizeof(LogicalRewriteMappingData);
        waldata_start = waldata = palloc(len);
 
        /*
         * collect data we need to write out, but don't modify ondisk data yet
         */
        dclist_foreach_modify(iter, &src->mappings)
        {
            RewriteMappingDataEntry *pmap;
 
            pmap = dclist_container(RewriteMappingDataEntry, node, iter.cur);
 
            memcpy(waldata, &pmap->map, sizeof(pmap->map));
            waldata += sizeof(pmap->map);
 
            /* remove from the list and free */
            dclist_delete_from(&src->mappings, &pmap->node);
            pfree(pmap);
 
            /* update bookkeeping */
            state->rs_num_rewrite_mappings--;
        }
 
        Assert(dclist_count(&src->mappings) == 0);
        Assert(waldata == waldata_start + len);
 
        /*
         * Note that we deviate from the usual WAL coding practices here,
         * check the above "Logical rewrite support" comment for reasoning.
         */
        written = FileWrite(src->vfd, waldata_start, len, src->off,
                            WAIT_EVENT_LOGICAL_REWRITE_WRITE);
        if (written != len)
            ereport(ERROR,
                    (errcode_for_file_access(),
                     errmsg("could not write to file \"%s\", wrote %d of %d: %m", src->path,
                            written, len)));
        src->off += len;
 
        XLogBeginInsert();
        XLogRegisterData((char *) (&xlrec), sizeof(xlrec));
        XLogRegisterData(waldata_start, len);
 
        /* write xlog record */
        XLogInsert(RM_HEAP2_ID, XLOG_HEAP2_REWRITE);
 
        pfree(waldata_start);
    }
    Assert(state->rs_num_rewrite_mappings == 0);
}

References Assert(), dlist_mutable_iter::cur, dclist_container, dclist_count(), dclist_delete_from(), dclist_foreach_modify, DEBUG1, elog(), ereport, errcode_for_file_access(), errmsg(), ERROR, FileWrite(), hash_seq_init(), hash_seq_search(), InvalidOid, len, RewriteMappingDataEntry::map, xl_heap_rewrite_mapping::mapped_db, xl_heap_rewrite_mapping::mapped_rel, xl_heap_rewrite_mapping::mapped_xid, RewriteMappingFile::mappings, MyDatabaseId, RewriteMappingDataEntry::node, xl_heap_rewrite_mapping::num_mappings, RewriteMappingFile::off, xl_heap_rewrite_mapping::offset, palloc(), RewriteMappingFile::path, pfree(), RelationGetRelid, xl_heap_rewrite_mapping::start_lsn, RewriteMappingFile::vfd, RewriteMappingFile::xid, XLOG_HEAP2_REWRITE, XLogBeginInsert(), XLogInsert(), and XLogRegisterData().

Referenced by logical_end_heap_rewrite(), and logical_rewrite_log_mapping().

logical_rewrite_heap_tuple()

static void logical_rewrite_heap_tuple ( RewriteState  state, ItemPointerData  old_tid, HeapTuple  new_tuple  )

static

Definition at line 1033 of file rewriteheap.c.

{
    ItemPointerData new_tid = new_tuple->t_self;
    TransactionId cutoff = state->rs_logical_xmin;
    TransactionId xmin;
    TransactionId xmax;
    bool        do_log_xmin = false;
    bool        do_log_xmax = false;
    LogicalRewriteMappingData map;
 
    /* no logical rewrite in progress, we don't need to log anything */
    if (!state->rs_logical_rewrite)
        return;
 
    xmin = HeapTupleHeaderGetXmin(new_tuple->t_data);
    /* use *GetUpdateXid to correctly deal with multixacts */
    xmax = HeapTupleHeaderGetUpdateXid(new_tuple->t_data);
 
    /*
     * Log the mapping iff the tuple has been created recently.
     */
    if (TransactionIdIsNormal(xmin) && !TransactionIdPrecedes(xmin, cutoff))
        do_log_xmin = true;
 
    if (!TransactionIdIsNormal(xmax))
    {
        /*
         * no xmax is set, can't have any permanent ones, so this check is
         * sufficient
         */
    }
    else if (HEAP_XMAX_IS_LOCKED_ONLY(new_tuple->t_data->t_infomask))
    {
        /* only locked, we don't care */
    }
    else if (!TransactionIdPrecedes(xmax, cutoff))
    {
        /* tuple has been deleted recently, log */
        do_log_xmax = true;
    }
 
    /* if neither needs to be logged, we're done */
    if (!do_log_xmin && !do_log_xmax)
        return;
 
    /* fill out mapping information */
    map.old_locator = state->rs_old_rel->rd_locator;
    map.old_tid = old_tid;
    map.new_locator = state->rs_new_rel->rd_locator;
    map.new_tid = new_tid;
 
    /* ---
     * Now persist the mapping for the individual xids that are affected. We
     * need to log for both xmin and xmax if they aren't the same transaction
     * since the mapping files are per "affected" xid.
     * We don't muster all that much effort detecting whether xmin and xmax
     * are actually the same transaction, we just check whether the xid is the
     * same disregarding subtransactions. Logging too much is relatively
     * harmless and we could never do the check fully since subtransaction
     * data is thrown away during restarts.
     * ---
     */
    if (do_log_xmin)
        logical_rewrite_log_mapping(state, xmin, &map);
    /* separately log mapping for xmax unless it'd be redundant */
    if (do_log_xmax && !TransactionIdEquals(xmin, xmax))
        logical_rewrite_log_mapping(state, xmax, &map);
}

References HEAP_XMAX_IS_LOCKED_ONLY, HeapTupleHeaderGetUpdateXid, HeapTupleHeaderGetXmin, logical_rewrite_log_mapping(), LogicalRewriteMappingData::new_locator, LogicalRewriteMappingData::new_tid, LogicalRewriteMappingData::old_locator, LogicalRewriteMappingData::old_tid, HeapTupleData::t_data, HeapTupleHeaderData::t_infomask, HeapTupleData::t_self, TransactionIdEquals, TransactionIdIsNormal, and TransactionIdPrecedes().

Referenced by rewrite_heap_tuple().

logical_rewrite_log_mapping()

static void logical_rewrite_log_mapping ( RewriteState  state, TransactionId  xid, LogicalRewriteMappingData *  map  )

static

Definition at line 969 of file rewriteheap.c.

{
    RewriteMappingFile *src;
    RewriteMappingDataEntry *pmap;
    Oid         relid;
    bool        found;
 
    relid = RelationGetRelid(state->rs_old_rel);
 
    /* look for existing mappings for this 'mapped' xid */
    src = hash_search(state->rs_logical_mappings, &xid,
                      HASH_ENTER, &found);
 
    /*
     * We haven't yet had the need to map anything for this xid, create
     * per-xid data structures.
     */
    if (!found)
    {
        char        path[MAXPGPATH];
        Oid         dboid;
 
        if (state->rs_old_rel->rd_rel->relisshared)
            dboid = InvalidOid;
        else
            dboid = MyDatabaseId;
 
        snprintf(path, MAXPGPATH,
                 "pg_logical/mappings/" LOGICAL_REWRITE_FORMAT,
                 dboid, relid,
                 LSN_FORMAT_ARGS(state->rs_begin_lsn),
                 xid, GetCurrentTransactionId());
 
        dclist_init(&src->mappings);
        src->off = 0;
        memcpy(src->path, path, sizeof(path));
        src->vfd = PathNameOpenFile(path,
                                    O_CREAT | O_EXCL | O_WRONLY | PG_BINARY);
        if (src->vfd < 0)
            ereport(ERROR,
                    (errcode_for_file_access(),
                     errmsg("could not create file \"%s\": %m", path)));
    }
 
    pmap = MemoryContextAlloc(state->rs_cxt,
                              sizeof(RewriteMappingDataEntry));
    memcpy(&pmap->map, map, sizeof(LogicalRewriteMappingData));
    dclist_push_tail(&src->mappings, &pmap->node);
    state->rs_num_rewrite_mappings++;
 
    /*
     * Write out buffer every time we've too many in-memory entries across all
     * mapping files.
     */
    if (state->rs_num_rewrite_mappings >= 1000 /* arbitrary number */ )
        logical_heap_rewrite_flush_mappings(state);
}

References dclist_init(), dclist_push_tail(), ereport, errcode_for_file_access(), errmsg(), ERROR, GetCurrentTransactionId(), HASH_ENTER, hash_search(), InvalidOid, logical_heap_rewrite_flush_mappings(), LOGICAL_REWRITE_FORMAT, LSN_FORMAT_ARGS, RewriteMappingDataEntry::map, RewriteMappingFile::mappings, MAXPGPATH, MemoryContextAlloc(), MyDatabaseId, RewriteMappingDataEntry::node, RewriteMappingFile::off, RewriteMappingFile::path, PathNameOpenFile(), PG_BINARY, RelationGetRelid, snprintf, and RewriteMappingFile::vfd.

Referenced by logical_rewrite_heap_tuple().

raw_heap_insert()

static void raw_heap_insert ( RewriteState  state, HeapTuple  tup  )

static

Definition at line 612 of file rewriteheap.c.

{
    Page        page = state->rs_buffer;
    Size        pageFreeSpace,
                saveFreeSpace;
    Size        len;
    OffsetNumber newoff;
    HeapTuple   heaptup;
 
    /*
     * If the new tuple is too big for storage or contains already toasted
     * out-of-line attributes from some other relation, invoke the toaster.
     *
     * Note: below this point, heaptup is the data we actually intend to store
     * into the relation; tup is the caller's original untoasted data.
     */
    if (state->rs_new_rel->rd_rel->relkind == RELKIND_TOASTVALUE)
    {
        /* toast table entries should never be recursively toasted */
        Assert(!HeapTupleHasExternal(tup));
        heaptup = tup;
    }
    else if (HeapTupleHasExternal(tup) || tup->t_len > TOAST_TUPLE_THRESHOLD)
    {
        int         options = HEAP_INSERT_SKIP_FSM;
 
        /*
         * While rewriting the heap for VACUUM FULL / CLUSTER, make sure data
         * for the TOAST table are not logically decoded.  The main heap is
         * WAL-logged as XLOG FPI records, which are not logically decoded.
         */
        options |= HEAP_INSERT_NO_LOGICAL;
 
        heaptup = heap_toast_insert_or_update(state->rs_new_rel, tup, NULL,
                                              options);
    }
    else
        heaptup = tup;
 
    len = MAXALIGN(heaptup->t_len); /* be conservative */
 
    /*
     * If we're gonna fail for oversize tuple, do it right away
     */
    if (len > MaxHeapTupleSize)
        ereport(ERROR,
                (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
                 errmsg("row is too big: size %zu, maximum size %zu",
                        len, MaxHeapTupleSize)));
 
    /* Compute desired extra freespace due to fillfactor option */
    saveFreeSpace = RelationGetTargetPageFreeSpace(state->rs_new_rel,
                                                   HEAP_DEFAULT_FILLFACTOR);
 
    /* Now we can check to see if there's enough free space already. */
    if (state->rs_buffer_valid)
    {
        pageFreeSpace = PageGetHeapFreeSpace(page);
 
        if (len + saveFreeSpace > pageFreeSpace)
        {
            /*
             * Doesn't fit, so write out the existing page.  It always
             * contains a tuple.  Hence, unlike RelationGetBufferForTuple(),
             * enforce saveFreeSpace unconditionally.
             */
 
            /* XLOG stuff */
            if (RelationNeedsWAL(state->rs_new_rel))
                log_newpage(&state->rs_new_rel->rd_locator,
                            MAIN_FORKNUM,
                            state->rs_blockno,
                            page,
                            true);
 
            /*
             * Now write the page. We say skipFsync = true because there's no
             * need for smgr to schedule an fsync for this write; we'll do it
             * ourselves in end_heap_rewrite.
             */
            PageSetChecksumInplace(page, state->rs_blockno);
 
            smgrextend(RelationGetSmgr(state->rs_new_rel), MAIN_FORKNUM,
                       state->rs_blockno, page, true);
 
            state->rs_blockno++;
            state->rs_buffer_valid = false;
        }
    }
 
    if (!state->rs_buffer_valid)
    {
        /* Initialize a new empty page */
        PageInit(page, BLCKSZ, 0);
        state->rs_buffer_valid = true;
    }
 
    /* And now we can insert the tuple into the page */
    newoff = PageAddItem(page, (Item) heaptup->t_data, heaptup->t_len,
                         InvalidOffsetNumber, false, true);
    if (newoff == InvalidOffsetNumber)
        elog(ERROR, "failed to add tuple");
 
    /* Update caller's t_self to the actual position where it was stored */
    ItemPointerSet(&(tup->t_self), state->rs_blockno, newoff);
 
    /*
     * Insert the correct position into CTID of the stored tuple, too, if the
     * caller didn't supply a valid CTID.
     */
    if (!ItemPointerIsValid(&tup->t_data->t_ctid))
    {
        ItemId      newitemid;
        HeapTupleHeader onpage_tup;
 
        newitemid = PageGetItemId(page, newoff);
        onpage_tup = (HeapTupleHeader) PageGetItem(page, newitemid);
 
        onpage_tup->t_ctid = tup->t_self;
    }
 
    /* If heaptup is a private copy, release it. */
    if (heaptup != tup)
        heap_freetuple(heaptup);
}

References Assert(), elog(), ereport, errcode(), errmsg(), ERROR, HEAP_DEFAULT_FILLFACTOR, heap_freetuple(), HEAP_INSERT_NO_LOGICAL, HEAP_INSERT_SKIP_FSM, heap_toast_insert_or_update(), HeapTupleHasExternal, InvalidOffsetNumber, ItemPointerIsValid(), ItemPointerSet(), len, log_newpage(), MAIN_FORKNUM, MAXALIGN, MaxHeapTupleSize, PageAddItem, PageGetHeapFreeSpace(), PageGetItem(), PageGetItemId(), PageInit(), PageSetChecksumInplace(), RelationGetSmgr(), RelationGetTargetPageFreeSpace, RelationNeedsWAL, smgrextend(), HeapTupleHeaderData::t_ctid, HeapTupleData::t_data, HeapTupleData::t_len, HeapTupleData::t_self, and TOAST_TUPLE_THRESHOLD.

Referenced by end_heap_rewrite(), and rewrite_heap_tuple().

rewrite_heap_dead_tuple()

bool rewrite_heap_dead_tuple	(	RewriteState	state,
		HeapTuple	old_tuple
	)

Definition at line 562 of file rewriteheap.c.

{
    /*
     * If we have already seen an earlier tuple in the update chain that
     * points to this tuple, let's forget about that earlier tuple. It's in
     * fact dead as well, our simple xmax < OldestXmin test in
     * HeapTupleSatisfiesVacuum just wasn't enough to detect it. It happens
     * when xmin of a tuple is greater than xmax, which sounds
     * counter-intuitive but is perfectly valid.
     *
     * We don't bother to try to detect the situation the other way round,
     * when we encounter the dead tuple first and then the recently dead one
     * that points to it. If that happens, we'll have some unmatched entries
     * in the UnresolvedTups hash table at the end. That can happen anyway,
     * because a vacuum might have removed the dead tuple in the chain before
     * us.
     */
    UnresolvedTup unresolved;
    TidHashKey  hashkey;
    bool        found;
 
    memset(&hashkey, 0, sizeof(hashkey));
    hashkey.xmin = HeapTupleHeaderGetXmin(old_tuple->t_data);
    hashkey.tid = old_tuple->t_self;
 
    unresolved = hash_search(state->rs_unresolved_tups, &hashkey,
                             HASH_FIND, NULL);
 
    if (unresolved != NULL)
    {
        /* Need to free the contained tuple as well as the hashtable entry */
        heap_freetuple(unresolved->tuple);
        hash_search(state->rs_unresolved_tups, &hashkey,
                    HASH_REMOVE, &found);
        Assert(found);
        return true;
    }
 
    return false;
}

References Assert(), HASH_FIND, HASH_REMOVE, hash_search(), heap_freetuple(), HeapTupleHeaderGetXmin, HeapTupleData::t_data, HeapTupleData::t_self, TidHashKey::tid, UnresolvedTupData::tuple, and TidHashKey::xmin.

Referenced by heapam_relation_copy_for_cluster().

rewrite_heap_tuple()

void rewrite_heap_tuple	(	RewriteState	state,
		HeapTuple	old_tuple,
		HeapTuple	new_tuple
	)

Definition at line 360 of file rewriteheap.c.

{
    MemoryContext old_cxt;
    ItemPointerData old_tid;
    TidHashKey  hashkey;
    bool        found;
    bool        free_new;
 
    old_cxt = MemoryContextSwitchTo(state->rs_cxt);
 
    /*
     * Copy the original tuple's visibility information into new_tuple.
     *
     * XXX we might later need to copy some t_infomask2 bits, too? Right now,
     * we intentionally clear the HOT status bits.
     */
    memcpy(&new_tuple->t_data->t_choice.t_heap,
           &old_tuple->t_data->t_choice.t_heap,
           sizeof(HeapTupleFields));
 
    new_tuple->t_data->t_infomask &= ~HEAP_XACT_MASK;
    new_tuple->t_data->t_infomask2 &= ~HEAP2_XACT_MASK;
    new_tuple->t_data->t_infomask |=
        old_tuple->t_data->t_infomask & HEAP_XACT_MASK;
 
    /*
     * While we have our hands on the tuple, we may as well freeze any
     * eligible xmin or xmax, so that future VACUUM effort can be saved.
     */
    heap_freeze_tuple(new_tuple->t_data,
                      state->rs_old_rel->rd_rel->relfrozenxid,
                      state->rs_old_rel->rd_rel->relminmxid,
                      state->rs_freeze_xid,
                      state->rs_cutoff_multi);
 
    /*
     * Invalid ctid means that ctid should point to the tuple itself. We'll
     * override it later if the tuple is part of an update chain.
     */
    ItemPointerSetInvalid(&new_tuple->t_data->t_ctid);
 
    /*
     * If the tuple has been updated, check the old-to-new mapping hash table.
     */
    if (!((old_tuple->t_data->t_infomask & HEAP_XMAX_INVALID) ||
          HeapTupleHeaderIsOnlyLocked(old_tuple->t_data)) &&
        !HeapTupleHeaderIndicatesMovedPartitions(old_tuple->t_data) &&
        !(ItemPointerEquals(&(old_tuple->t_self),
                            &(old_tuple->t_data->t_ctid))))
    {
        OldToNewMapping mapping;
 
        memset(&hashkey, 0, sizeof(hashkey));
        hashkey.xmin = HeapTupleHeaderGetUpdateXid(old_tuple->t_data);
        hashkey.tid = old_tuple->t_data->t_ctid;
 
        mapping = (OldToNewMapping)
            hash_search(state->rs_old_new_tid_map, &hashkey,
                        HASH_FIND, NULL);
 
        if (mapping != NULL)
        {
            /*
             * We've already copied the tuple that t_ctid points to, so we can
             * set the ctid of this tuple to point to the new location, and
             * insert it right away.
             */
            new_tuple->t_data->t_ctid = mapping->new_tid;
 
            /* We don't need the mapping entry anymore */
            hash_search(state->rs_old_new_tid_map, &hashkey,
                        HASH_REMOVE, &found);
            Assert(found);
        }
        else
        {
            /*
             * We haven't seen the tuple t_ctid points to yet. Stash this
             * tuple into unresolved_tups to be written later.
             */
            UnresolvedTup unresolved;
 
            unresolved = hash_search(state->rs_unresolved_tups, &hashkey,
                                     HASH_ENTER, &found);
            Assert(!found);
 
            unresolved->old_tid = old_tuple->t_self;
            unresolved->tuple = heap_copytuple(new_tuple);
 
            /*
             * We can't do anything more now, since we don't know where the
             * tuple will be written.
             */
            MemoryContextSwitchTo(old_cxt);
            return;
        }
    }
 
    /*
     * Now we will write the tuple, and then check to see if it is the B tuple
     * in any new or known pair.  When we resolve a known pair, we will be
     * able to write that pair's A tuple, and then we have to check if it
     * resolves some other pair.  Hence, we need a loop here.
     */
    old_tid = old_tuple->t_self;
    free_new = false;
 
    for (;;)
    {
        ItemPointerData new_tid;
 
        /* Insert the tuple and find out where it's put in new_heap */
        raw_heap_insert(state, new_tuple);
        new_tid = new_tuple->t_self;
 
        logical_rewrite_heap_tuple(state, old_tid, new_tuple);
 
        /*
         * If the tuple is the updated version of a row, and the prior version
         * wouldn't be DEAD yet, then we need to either resolve the prior
         * version (if it's waiting in rs_unresolved_tups), or make an entry
         * in rs_old_new_tid_map (so we can resolve it when we do see it). The
         * previous tuple's xmax would equal this one's xmin, so it's
         * RECENTLY_DEAD if and only if the xmin is not before OldestXmin.
         */
        if ((new_tuple->t_data->t_infomask & HEAP_UPDATED) &&
            !TransactionIdPrecedes(HeapTupleHeaderGetXmin(new_tuple->t_data),
                                   state->rs_oldest_xmin))
        {
            /*
             * Okay, this is B in an update pair.  See if we've seen A.
             */
            UnresolvedTup unresolved;
 
            memset(&hashkey, 0, sizeof(hashkey));
            hashkey.xmin = HeapTupleHeaderGetXmin(new_tuple->t_data);
            hashkey.tid = old_tid;
 
            unresolved = hash_search(state->rs_unresolved_tups, &hashkey,
                                     HASH_FIND, NULL);
 
            if (unresolved != NULL)
            {
                /*
                 * We have seen and memorized the previous tuple already. Now
                 * that we know where we inserted the tuple its t_ctid points
                 * to, fix its t_ctid and insert it to the new heap.
                 */
                if (free_new)
                    heap_freetuple(new_tuple);
                new_tuple = unresolved->tuple;
                free_new = true;
                old_tid = unresolved->old_tid;
                new_tuple->t_data->t_ctid = new_tid;
 
                /*
                 * We don't need the hash entry anymore, but don't free its
                 * tuple just yet.
                 */
                hash_search(state->rs_unresolved_tups, &hashkey,
                            HASH_REMOVE, &found);
                Assert(found);
 
                /* loop back to insert the previous tuple in the chain */
                continue;
            }
            else
            {
                /*
                 * Remember the new tid of this tuple. We'll use it to set the
                 * ctid when we find the previous tuple in the chain.
                 */
                OldToNewMapping mapping;
 
                mapping = hash_search(state->rs_old_new_tid_map, &hashkey,
                                      HASH_ENTER, &found);
                Assert(!found);
 
                mapping->new_tid = new_tid;
            }
        }
 
        /* Done with this (chain of) tuples, for now */
        if (free_new)
            heap_freetuple(new_tuple);
        break;
    }
 
    MemoryContextSwitchTo(old_cxt);
}

References Assert(), HASH_ENTER, HASH_FIND, HASH_REMOVE, hash_search(), HEAP2_XACT_MASK, heap_copytuple(), heap_freetuple(), heap_freeze_tuple(), HEAP_UPDATED, HEAP_XACT_MASK, HEAP_XMAX_INVALID, HeapTupleHeaderGetUpdateXid, HeapTupleHeaderGetXmin, HeapTupleHeaderIndicatesMovedPartitions, HeapTupleHeaderIsOnlyLocked(), ItemPointerEquals(), ItemPointerSetInvalid(), logical_rewrite_heap_tuple(), MemoryContextSwitchTo(), OldToNewMappingData::new_tid, UnresolvedTupData::old_tid, raw_heap_insert(), HeapTupleHeaderData::t_choice, HeapTupleHeaderData::t_ctid, HeapTupleData::t_data, HeapTupleHeaderData::t_heap, HeapTupleHeaderData::t_infomask, HeapTupleHeaderData::t_infomask2, HeapTupleData::t_self, TidHashKey::tid, TransactionIdPrecedes(), UnresolvedTupData::tuple, and TidHashKey::xmin.

Referenced by reform_and_rewrite_tuple().