#include "port/pg_iovec.h"
#include "storage/aio_types.h"
#include "storage/block.h"
#include "storage/buf.h"
#include "storage/bufpage.h"
#include "storage/relfilelocator.h"
#include "utils/relcache.h"
#include "utils/snapmgr.h"

Include dependency graph for bufmgr.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Data Structures
struct	PrefetchBufferResult

struct	BufferManagerRelation

struct	ReadBuffersOperation

Macros
#define	BMR_REL(p_rel) ((BufferManagerRelation){.rel = p_rel})

#define	BMR_SMGR(p_smgr, p_relpersistence) ((BufferManagerRelation){.smgr = p_smgr, .relpersistence = p_relpersistence})

#define	BMR_GET_SMGR(bmr) (RelationIsValid((bmr).rel) ? RelationGetSmgr((bmr).rel) : (bmr).smgr)

#define	READ_BUFFERS_ZERO_ON_ERROR (1 << 0)

#define	READ_BUFFERS_ISSUE_ADVICE (1 << 1)

#define	READ_BUFFERS_IGNORE_CHECKSUM_FAILURES (1 << 2)

#define	READ_BUFFERS_SYNCHRONOUSLY (1 << 3)

#define	DEFAULT_EFFECTIVE_IO_CONCURRENCY 16

#define	DEFAULT_MAINTENANCE_IO_CONCURRENCY 16

#define	MAX_IO_COMBINE_LIMIT PG_IOV_MAX

#define	DEFAULT_IO_COMBINE_LIMIT Min(MAX_IO_COMBINE_LIMIT, (128 * 1024) / BLCKSZ)

#define	MAX_IO_CONCURRENCY 1000

#define	P_NEW InvalidBlockNumber /* grow the file to get a new page */

#define	RelationGetNumberOfBlocks(reln) RelationGetNumberOfBlocksInFork(reln, MAIN_FORKNUM)

Typedefs
typedef void *	Block

typedef enum BufferAccessStrategyType	BufferAccessStrategyType

typedef struct PrefetchBufferResult	PrefetchBufferResult

typedef enum ExtendBufferedFlags	ExtendBufferedFlags

typedef struct SMgrRelationData *	SMgrRelation

typedef struct BufferManagerRelation	BufferManagerRelation

typedef struct ReadBuffersOperation	ReadBuffersOperation

typedef struct WritebackContext	WritebackContext

typedef enum BufferLockMode	BufferLockMode

Enumerations
enum	BufferAccessStrategyType { BAS_NORMAL , BAS_BULKREAD , BAS_BULKWRITE , BAS_VACUUM }

enum	ReadBufferMode { RBM_NORMAL , RBM_ZERO_AND_LOCK , RBM_ZERO_AND_CLEANUP_LOCK , RBM_ZERO_ON_ERROR , RBM_NORMAL_NO_LOG }

enum	ExtendBufferedFlags { EB_SKIP_EXTENSION_LOCK = (1 << 0) , EB_PERFORMING_RECOVERY = (1 << 1) , EB_CREATE_FORK_IF_NEEDED = (1 << 2) , EB_LOCK_FIRST = (1 << 3) , EB_CLEAR_SIZE_CACHE = (1 << 4) , EB_LOCK_TARGET = (1 << 5) }

enum	BufferLockMode { BUFFER_LOCK_UNLOCK , BUFFER_LOCK_SHARE , BUFFER_LOCK_EXCLUSIVE }

Functions
PrefetchBufferResult	PrefetchSharedBuffer (SMgrRelation smgr_reln, ForkNumber forkNum, BlockNumber blockNum)

PrefetchBufferResult	PrefetchBuffer (Relation reln, ForkNumber forkNum, BlockNumber blockNum)

bool	ReadRecentBuffer (RelFileLocator rlocator, ForkNumber forkNum, BlockNumber blockNum, Buffer recent_buffer)

Buffer	ReadBuffer (Relation reln, BlockNumber blockNum)

Buffer	ReadBufferExtended (Relation reln, ForkNumber forkNum, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy)

Buffer	ReadBufferWithoutRelcache (RelFileLocator rlocator, ForkNumber forkNum, BlockNumber blockNum, ReadBufferMode mode, BufferAccessStrategy strategy, bool permanent)

bool	StartReadBuffer (ReadBuffersOperation operation, Buffer buffer, BlockNumber blocknum, int flags)

bool	StartReadBuffers (ReadBuffersOperation operation, Buffer buffers, BlockNumber blockNum, int *nblocks, int flags)

void	WaitReadBuffers (ReadBuffersOperation *operation)

void	ReleaseBuffer (Buffer buffer)

void	UnlockReleaseBuffer (Buffer buffer)

bool	BufferIsLockedByMe (Buffer buffer)

bool	BufferIsLockedByMeInMode (Buffer buffer, BufferLockMode mode)

bool	BufferIsDirty (Buffer buffer)

void	MarkBufferDirty (Buffer buffer)

void	IncrBufferRefCount (Buffer buffer)

void	CheckBufferIsPinnedOnce (Buffer buffer)

Buffer	ReleaseAndReadBuffer (Buffer buffer, Relation relation, BlockNumber blockNum)

Buffer	ExtendBufferedRel (BufferManagerRelation bmr, ForkNumber forkNum, BufferAccessStrategy strategy, uint32 flags)

BlockNumber	ExtendBufferedRelBy (BufferManagerRelation bmr, ForkNumber fork, BufferAccessStrategy strategy, uint32 flags, uint32 extend_by, Buffer buffers, uint32 extended_by)

Buffer	ExtendBufferedRelTo (BufferManagerRelation bmr, ForkNumber fork, BufferAccessStrategy strategy, uint32 flags, BlockNumber extend_to, ReadBufferMode mode)

void	InitBufferManagerAccess (void)

void	AtEOXact_Buffers (bool isCommit)

char *	DebugPrintBufferRefcount (Buffer buffer)

void	CheckPointBuffers (int flags)

BlockNumber	BufferGetBlockNumber (Buffer buffer)

BlockNumber	RelationGetNumberOfBlocksInFork (Relation relation, ForkNumber forkNum)

void	FlushOneBuffer (Buffer buffer)

void	FlushRelationBuffers (Relation rel)

void	FlushRelationsAllBuffers (SMgrRelation *smgrs, int nrels)

void	CreateAndCopyRelationData (RelFileLocator src_rlocator, RelFileLocator dst_rlocator, bool permanent)

void	FlushDatabaseBuffers (Oid dbid)

void	DropRelationBuffers (SMgrRelation smgr_reln, ForkNumber forkNum, int nforks, BlockNumber firstDelBlock)

void	DropRelationsAllBuffers (SMgrRelation *smgr_reln, int nlocators)

void	DropDatabaseBuffers (Oid dbid)

bool	BufferIsPermanent (Buffer buffer)

XLogRecPtr	BufferGetLSNAtomic (Buffer buffer)

void	BufferGetTag (Buffer buffer, RelFileLocator rlocator, ForkNumber forknum, BlockNumber *blknum)

void	MarkBufferDirtyHint (Buffer buffer, bool buffer_std)

void	UnlockBuffers (void)

void	LockBuffer (Buffer buffer, BufferLockMode mode)

bool	ConditionalLockBuffer (Buffer buffer)

void	LockBufferForCleanup (Buffer buffer)

bool	ConditionalLockBufferForCleanup (Buffer buffer)

bool	IsBufferCleanupOK (Buffer buffer)

bool	HoldingBufferPinThatDelaysRecovery (void)

bool	BgBufferSync (WritebackContext *wb_context)

uint32	GetPinLimit (void)

uint32	GetLocalPinLimit (void)

uint32	GetAdditionalPinLimit (void)

uint32	GetAdditionalLocalPinLimit (void)

void	LimitAdditionalPins (uint32 *additional_pins)

void	LimitAdditionalLocalPins (uint32 *additional_pins)

bool	EvictUnpinnedBuffer (Buffer buf, bool *buffer_flushed)

void	EvictAllUnpinnedBuffers (int32 buffers_evicted, int32 buffers_flushed, int32 *buffers_skipped)

void	EvictRelUnpinnedBuffers (Relation rel, int32 buffers_evicted, int32 buffers_flushed, int32 *buffers_skipped)

bool	MarkDirtyUnpinnedBuffer (Buffer buf, bool *buffer_already_dirty)

void	MarkDirtyRelUnpinnedBuffers (Relation rel, int32 buffers_dirtied, int32 buffers_already_dirty, int32 *buffers_skipped)

void	MarkDirtyAllUnpinnedBuffers (int32 buffers_dirtied, int32 buffers_already_dirty, int32 *buffers_skipped)

void	BufferManagerShmemInit (void)

Size	BufferManagerShmemSize (void)

void	AtProcExit_LocalBuffers (void)

BufferAccessStrategy	GetAccessStrategy (BufferAccessStrategyType btype)

BufferAccessStrategy	GetAccessStrategyWithSize (BufferAccessStrategyType btype, int ring_size_kb)

int	GetAccessStrategyBufferCount (BufferAccessStrategy strategy)

int	GetAccessStrategyPinLimit (BufferAccessStrategy strategy)

void	FreeAccessStrategy (BufferAccessStrategy strategy)

static bool	BufferIsValid (Buffer bufnum)

static Block	BufferGetBlock (Buffer buffer)

static Size	BufferGetPageSize (Buffer buffer)

static Page	BufferGetPage (Buffer buffer)

Macro Definition Documentation

◆ BMR_GET_SMGR

#define BMR_GET_SMGR ( bmr ) (RelationIsValid((bmr).rel) ? RelationGetSmgr((bmr).rel) : (bmr).smgr)

Definition at line 118 of file bufmgr.h.

◆ BMR_REL

#define BMR_REL ( p_rel ) ((BufferManagerRelation){.rel = p_rel})

Definition at line 114 of file bufmgr.h.

◆ BMR_SMGR

#define BMR_SMGR	(	p_smgr,
		p_relpersistence
	)	((BufferManagerRelation){.smgr = p_smgr, .relpersistence = p_relpersistence})

Definition at line 116 of file bufmgr.h.

◆ DEFAULT_EFFECTIVE_IO_CONCURRENCY

#define DEFAULT_EFFECTIVE_IO_CONCURRENCY 16

Definition at line 168 of file bufmgr.h.

◆ DEFAULT_IO_COMBINE_LIMIT

#define DEFAULT_IO_COMBINE_LIMIT Min(MAX_IO_COMBINE_LIMIT, (128 * 1024) / BLCKSZ)

Definition at line 174 of file bufmgr.h.

◆ DEFAULT_MAINTENANCE_IO_CONCURRENCY

#define DEFAULT_MAINTENANCE_IO_CONCURRENCY 16

Definition at line 169 of file bufmgr.h.

◆ MAX_IO_COMBINE_LIMIT

#define MAX_IO_COMBINE_LIMIT PG_IOV_MAX

Definition at line 173 of file bufmgr.h.

◆ MAX_IO_CONCURRENCY

#define MAX_IO_CONCURRENCY 1000

Definition at line 195 of file bufmgr.h.

◆ P_NEW

#define P_NEW InvalidBlockNumber /* grow the file to get a new page */

Definition at line 198 of file bufmgr.h.

◆ READ_BUFFERS_IGNORE_CHECKSUM_FAILURES

#define READ_BUFFERS_IGNORE_CHECKSUM_FAILURES (1 << 2)

Definition at line 126 of file bufmgr.h.

◆ READ_BUFFERS_ISSUE_ADVICE

#define READ_BUFFERS_ISSUE_ADVICE (1 << 1)

Definition at line 124 of file bufmgr.h.

◆ READ_BUFFERS_SYNCHRONOUSLY

#define READ_BUFFERS_SYNCHRONOUSLY (1 << 3)

Definition at line 128 of file bufmgr.h.

◆ READ_BUFFERS_ZERO_ON_ERROR

#define READ_BUFFERS_ZERO_ON_ERROR (1 << 0)

Definition at line 122 of file bufmgr.h.

◆ RelationGetNumberOfBlocks

#define RelationGetNumberOfBlocks ( reln ) RelationGetNumberOfBlocksInFork(reln, MAIN_FORKNUM)

Definition at line 294 of file bufmgr.h.

Typedef Documentation

◆ Block

typedef void* Block

Definition at line 26 of file bufmgr.h.

◆ BufferAccessStrategyType

typedef enum BufferAccessStrategyType BufferAccessStrategyType

◆ BufferLockMode

typedef enum BufferLockMode BufferLockMode

◆ BufferManagerRelation

typedef struct BufferManagerRelation BufferManagerRelation

◆ ExtendBufferedFlags

typedef enum ExtendBufferedFlags ExtendBufferedFlags

◆ PrefetchBufferResult

typedef struct PrefetchBufferResult PrefetchBufferResult

◆ ReadBuffersOperation

typedef struct ReadBuffersOperation ReadBuffersOperation

Definition at line 154 of file bufmgr.h.

◆ SMgrRelation

typedef struct SMgrRelationData* SMgrRelation

Definition at line 97 of file bufmgr.h.

◆ WritebackContext

typedef struct WritebackContext WritebackContext

Definition at line 157 of file bufmgr.h.

Enumeration Type Documentation

◆ BufferAccessStrategyType

enum BufferAccessStrategyType

Enumerator
BAS_NORMAL
BAS_BULKREAD
BAS_BULKWRITE
BAS_VACUUM

Definition at line 34 of file bufmgr.h.

{
    BAS_NORMAL,                 /* Normal random access */
    BAS_BULKREAD,               /* Large read-only scan (hint bit updates are
                                 * ok) */
    BAS_BULKWRITE,              /* Large multi-block write (e.g. COPY IN) */
    BAS_VACUUM,                 /* VACUUM */
} BufferAccessStrategyType;

◆ BufferLockMode

enum BufferLockMode

Enumerator
BUFFER_LOCK_UNLOCK
BUFFER_LOCK_SHARE
BUFFER_LOCK_EXCLUSIVE

Definition at line 203 of file bufmgr.h.

{
    BUFFER_LOCK_UNLOCK,
    BUFFER_LOCK_SHARE,
    BUFFER_LOCK_EXCLUSIVE,
} BufferLockMode;

◆ ExtendBufferedFlags

enum ExtendBufferedFlags

Enumerator
EB_SKIP_EXTENSION_LOCK
EB_PERFORMING_RECOVERY
EB_CREATE_FORK_IF_NEEDED
EB_LOCK_FIRST
EB_CLEAR_SIZE_CACHE
EB_LOCK_TARGET

Definition at line 68 of file bufmgr.h.

{
    /*
     * Don't acquire extension lock. This is safe only if the relation isn't
     * shared, an access exclusive lock is held or if this is the startup
     * process.
     */
    EB_SKIP_EXTENSION_LOCK = (1 << 0),
 
    /* Is this extension part of recovery? */
    EB_PERFORMING_RECOVERY = (1 << 1),
 
    /*
     * Should the fork be created if it does not currently exist? This likely
     * only ever makes sense for relation forks.
     */
    EB_CREATE_FORK_IF_NEEDED = (1 << 2),
 
    /* Should the first (possibly only) return buffer be returned locked? */
    EB_LOCK_FIRST = (1 << 3),
 
    /* Should the smgr size cache be cleared? */
    EB_CLEAR_SIZE_CACHE = (1 << 4),
 
    /* internal flags follow */
    EB_LOCK_TARGET = (1 << 5),
}           ExtendBufferedFlags;

◆ ReadBufferMode

enum ReadBufferMode

Enumerator
RBM_NORMAL
RBM_ZERO_AND_LOCK
RBM_ZERO_AND_CLEANUP_LOCK
RBM_ZERO_ON_ERROR
RBM_NORMAL_NO_LOG

Definition at line 44 of file bufmgr.h.

{
    RBM_NORMAL,                 /* Normal read */
    RBM_ZERO_AND_LOCK,          /* Don't read from disk, caller will
                                 * initialize. Also locks the page. */
    RBM_ZERO_AND_CLEANUP_LOCK,  /* Like RBM_ZERO_AND_LOCK, but locks the page
                                 * in "cleanup" mode */
    RBM_ZERO_ON_ERROR,          /* Read, but return an all-zeros page on error */
    RBM_NORMAL_NO_LOG,          /* Don't log page as invalid during WAL
                                 * replay; otherwise same as RBM_NORMAL */
} ReadBufferMode;

Function Documentation

◆ AtEOXact_Buffers()

void AtEOXact_Buffers ( bool isCommit )

Definition at line 3990 of file bufmgr.c.

{
    CheckForBufferLeaks();
 
    AtEOXact_LocalBuffers(isCommit);
 
    Assert(PrivateRefCountOverflowed == 0);
}

References Assert(), AtEOXact_LocalBuffers(), CheckForBufferLeaks(), and PrivateRefCountOverflowed.

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

◆ AtProcExit_LocalBuffers()

void AtProcExit_LocalBuffers ( void )

Definition at line 1014 of file localbuf.c.

{
    /*
     * We shouldn't be holding any remaining pins; if we are, and assertions
     * aren't enabled, we'll fail later in DropRelationBuffers while trying to
     * drop the temp rels.
     */
    CheckForLocalBufferLeaks();
}

References CheckForLocalBufferLeaks().

Referenced by AtProcExit_Buffers().

◆ BgBufferSync()

bool BgBufferSync ( WritebackContext * wb_context )

Definition at line 3622 of file bufmgr.c.

{
    /* info obtained from freelist.c */
    int         strategy_buf_id;
    uint32      strategy_passes;
    uint32      recent_alloc;
 
    /*
     * Information saved between calls so we can determine the strategy
     * point's advance rate and avoid scanning already-cleaned buffers.
     */
    static bool saved_info_valid = false;
    static int  prev_strategy_buf_id;
    static uint32 prev_strategy_passes;
    static int  next_to_clean;
    static uint32 next_passes;
 
    /* Moving averages of allocation rate and clean-buffer density */
    static float smoothed_alloc = 0;
    static float smoothed_density = 10.0;
 
    /* Potentially these could be tunables, but for now, not */
    float       smoothing_samples = 16;
    float       scan_whole_pool_milliseconds = 120000.0;
 
    /* Used to compute how far we scan ahead */
    long        strategy_delta;
    int         bufs_to_lap;
    int         bufs_ahead;
    float       scans_per_alloc;
    int         reusable_buffers_est;
    int         upcoming_alloc_est;
    int         min_scan_buffers;
 
    /* Variables for the scanning loop proper */
    int         num_to_scan;
    int         num_written;
    int         reusable_buffers;
 
    /* Variables for final smoothed_density update */
    long        new_strategy_delta;
    uint32      new_recent_alloc;
 
    /*
     * Find out where the clock-sweep currently is, and how many buffer
     * allocations have happened since our last call.
     */
    strategy_buf_id = StrategySyncStart(&strategy_passes, &recent_alloc);
 
    /* Report buffer alloc counts to pgstat */
    PendingBgWriterStats.buf_alloc += recent_alloc;
 
    /*
     * If we're not running the LRU scan, just stop after doing the stats
     * stuff.  We mark the saved state invalid so that we can recover sanely
     * if LRU scan is turned back on later.
     */
    if (bgwriter_lru_maxpages <= 0)
    {
        saved_info_valid = false;
        return true;
    }
 
    /*
     * Compute strategy_delta = how many buffers have been scanned by the
     * clock-sweep since last time.  If first time through, assume none. Then
     * see if we are still ahead of the clock-sweep, and if so, how many
     * buffers we could scan before we'd catch up with it and "lap" it. Note:
     * weird-looking coding of xxx_passes comparisons are to avoid bogus
     * behavior when the passes counts wrap around.
     */
    if (saved_info_valid)
    {
        int32       passes_delta = strategy_passes - prev_strategy_passes;
 
        strategy_delta = strategy_buf_id - prev_strategy_buf_id;
        strategy_delta += (long) passes_delta * NBuffers;
 
        Assert(strategy_delta >= 0);
 
        if ((int32) (next_passes - strategy_passes) > 0)
        {
            /* we're one pass ahead of the strategy point */
            bufs_to_lap = strategy_buf_id - next_to_clean;
#ifdef BGW_DEBUG
            elog(DEBUG2, "bgwriter ahead: bgw %u-%u strategy %u-%u delta=%ld lap=%d",
                 next_passes, next_to_clean,
                 strategy_passes, strategy_buf_id,
                 strategy_delta, bufs_to_lap);
#endif
        }
        else if (next_passes == strategy_passes &&
                 next_to_clean >= strategy_buf_id)
        {
            /* on same pass, but ahead or at least not behind */
            bufs_to_lap = NBuffers - (next_to_clean - strategy_buf_id);
#ifdef BGW_DEBUG
            elog(DEBUG2, "bgwriter ahead: bgw %u-%u strategy %u-%u delta=%ld lap=%d",
                 next_passes, next_to_clean,
                 strategy_passes, strategy_buf_id,
                 strategy_delta, bufs_to_lap);
#endif
        }
        else
        {
            /*
             * We're behind, so skip forward to the strategy point and start
             * cleaning from there.
             */
#ifdef BGW_DEBUG
            elog(DEBUG2, "bgwriter behind: bgw %u-%u strategy %u-%u delta=%ld",
                 next_passes, next_to_clean,
                 strategy_passes, strategy_buf_id,
                 strategy_delta);
#endif
            next_to_clean = strategy_buf_id;
            next_passes = strategy_passes;
            bufs_to_lap = NBuffers;
        }
    }
    else
    {
        /*
         * Initializing at startup or after LRU scanning had been off. Always
         * start at the strategy point.
         */
#ifdef BGW_DEBUG
        elog(DEBUG2, "bgwriter initializing: strategy %u-%u",
             strategy_passes, strategy_buf_id);
#endif
        strategy_delta = 0;
        next_to_clean = strategy_buf_id;
        next_passes = strategy_passes;
        bufs_to_lap = NBuffers;
    }
 
    /* Update saved info for next time */
    prev_strategy_buf_id = strategy_buf_id;
    prev_strategy_passes = strategy_passes;
    saved_info_valid = true;
 
    /*
     * Compute how many buffers had to be scanned for each new allocation, ie,
     * 1/density of reusable buffers, and track a moving average of that.
     *
     * If the strategy point didn't move, we don't update the density estimate
     */
    if (strategy_delta > 0 && recent_alloc > 0)
    {
        scans_per_alloc = (float) strategy_delta / (float) recent_alloc;
        smoothed_density += (scans_per_alloc - smoothed_density) /
            smoothing_samples;
    }
 
    /*
     * Estimate how many reusable buffers there are between the current
     * strategy point and where we've scanned ahead to, based on the smoothed
     * density estimate.
     */
    bufs_ahead = NBuffers - bufs_to_lap;
    reusable_buffers_est = (float) bufs_ahead / smoothed_density;
 
    /*
     * Track a moving average of recent buffer allocations.  Here, rather than
     * a true average we want a fast-attack, slow-decline behavior: we
     * immediately follow any increase.
     */
    if (smoothed_alloc <= (float) recent_alloc)
        smoothed_alloc = recent_alloc;
    else
        smoothed_alloc += ((float) recent_alloc - smoothed_alloc) /
            smoothing_samples;
 
    /* Scale the estimate by a GUC to allow more aggressive tuning. */
    upcoming_alloc_est = (int) (smoothed_alloc * bgwriter_lru_multiplier);
 
    /*
     * If recent_alloc remains at zero for many cycles, smoothed_alloc will
     * eventually underflow to zero, and the underflows produce annoying
     * kernel warnings on some platforms.  Once upcoming_alloc_est has gone to
     * zero, there's no point in tracking smaller and smaller values of
     * smoothed_alloc, so just reset it to exactly zero to avoid this
     * syndrome.  It will pop back up as soon as recent_alloc increases.
     */
    if (upcoming_alloc_est == 0)
        smoothed_alloc = 0;
 
    /*
     * Even in cases where there's been little or no buffer allocation
     * activity, we want to make a small amount of progress through the buffer
     * cache so that as many reusable buffers as possible are clean after an
     * idle period.
     *
     * (scan_whole_pool_milliseconds / BgWriterDelay) computes how many times
     * the BGW will be called during the scan_whole_pool time; slice the
     * buffer pool into that many sections.
     */
    min_scan_buffers = (int) (NBuffers / (scan_whole_pool_milliseconds / BgWriterDelay));
 
    if (upcoming_alloc_est < (min_scan_buffers + reusable_buffers_est))
    {
#ifdef BGW_DEBUG
        elog(DEBUG2, "bgwriter: alloc_est=%d too small, using min=%d + reusable_est=%d",
             upcoming_alloc_est, min_scan_buffers, reusable_buffers_est);
#endif
        upcoming_alloc_est = min_scan_buffers + reusable_buffers_est;
    }
 
    /*
     * Now write out dirty reusable buffers, working forward from the
     * next_to_clean point, until we have lapped the strategy scan, or cleaned
     * enough buffers to match our estimate of the next cycle's allocation
     * requirements, or hit the bgwriter_lru_maxpages limit.
     */
 
    num_to_scan = bufs_to_lap;
    num_written = 0;
    reusable_buffers = reusable_buffers_est;
 
    /* Execute the LRU scan */
    while (num_to_scan > 0 && reusable_buffers < upcoming_alloc_est)
    {
        int         sync_state = SyncOneBuffer(next_to_clean, true,
                                               wb_context);
 
        if (++next_to_clean >= NBuffers)
        {
            next_to_clean = 0;
            next_passes++;
        }
        num_to_scan--;
 
        if (sync_state & BUF_WRITTEN)
        {
            reusable_buffers++;
            if (++num_written >= bgwriter_lru_maxpages)
            {
                PendingBgWriterStats.maxwritten_clean++;
                break;
            }
        }
        else if (sync_state & BUF_REUSABLE)
            reusable_buffers++;
    }
 
    PendingBgWriterStats.buf_written_clean += num_written;
 
#ifdef BGW_DEBUG
    elog(DEBUG1, "bgwriter: recent_alloc=%u smoothed=%.2f delta=%ld ahead=%d density=%.2f reusable_est=%d upcoming_est=%d scanned=%d wrote=%d reusable=%d",
         recent_alloc, smoothed_alloc, strategy_delta, bufs_ahead,
         smoothed_density, reusable_buffers_est, upcoming_alloc_est,
         bufs_to_lap - num_to_scan,
         num_written,
         reusable_buffers - reusable_buffers_est);
#endif
 
    /*
     * Consider the above scan as being like a new allocation scan.
     * Characterize its density and update the smoothed one based on it. This
     * effectively halves the moving average period in cases where both the
     * strategy and the background writer are doing some useful scanning,
     * which is helpful because a long memory isn't as desirable on the
     * density estimates.
     */
    new_strategy_delta = bufs_to_lap - num_to_scan;
    new_recent_alloc = reusable_buffers - reusable_buffers_est;
    if (new_strategy_delta > 0 && new_recent_alloc > 0)
    {
        scans_per_alloc = (float) new_strategy_delta / (float) new_recent_alloc;
        smoothed_density += (scans_per_alloc - smoothed_density) /
            smoothing_samples;
 
#ifdef BGW_DEBUG
        elog(DEBUG2, "bgwriter: cleaner density alloc=%u scan=%ld density=%.2f new smoothed=%.2f",
             new_recent_alloc, new_strategy_delta,
             scans_per_alloc, smoothed_density);
#endif
    }
 
    /* Return true if OK to hibernate */
    return (bufs_to_lap == 0 && recent_alloc == 0);
}

References Assert(), bgwriter_lru_maxpages, bgwriter_lru_multiplier, BgWriterDelay, PgStat_BgWriterStats::buf_alloc, BUF_REUSABLE, BUF_WRITTEN, PgStat_BgWriterStats::buf_written_clean, DEBUG1, DEBUG2, elog, PgStat_BgWriterStats::maxwritten_clean, NBuffers, PendingBgWriterStats, StrategySyncStart(), and SyncOneBuffer().

Referenced by BackgroundWriterMain().

◆ BufferGetBlock()

static Block BufferGetBlock ( Buffer buffer )

inlinestatic

Definition at line 403 of file bufmgr.h.

{
    Assert(BufferIsValid(buffer));
 
    if (BufferIsLocal(buffer))
        return LocalBufferBlockPointers[-buffer - 1];
    else
        return (Block) (BufferBlocks + ((Size) (buffer - 1)) * BLCKSZ);
}

References Assert(), BufferBlocks, BufferIsLocal, BufferIsValid(), and LocalBufferBlockPointers.

Referenced by AsyncReadBuffers(), buffer_readv_complete_one(), BufferGetPage(), heap_inplace_update_and_unlock(), pg_buffercache_os_pages_internal(), read_rel_block_ll(), and XLogSaveBufferForHint().

◆ BufferGetBlockNumber()

BlockNumber BufferGetBlockNumber ( Buffer buffer )

Definition at line 4223 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    Assert(BufferIsPinned(buffer));
 
    if (BufferIsLocal(buffer))
        bufHdr = GetLocalBufferDescriptor(-buffer - 1);
    else
        bufHdr = GetBufferDescriptor(buffer - 1);
 
    /* pinned, so OK to read tag without spinlock */
    return bufHdr->tag.blockNum;
}

References Assert(), buftag::blockNum, PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsPinned, GetBufferDescriptor(), GetLocalBufferDescriptor(), and BufferDesc::tag.

Referenced by _bt_binsrch_insert(), _bt_bottomupdel_pass(), _bt_check_unique(), _bt_checkpage(), _bt_delitems_delete(), _bt_delitems_vacuum(), _bt_doinsert(), _bt_finish_split(), _bt_getroot(), _bt_insert_parent(), _bt_insertonpg(), _bt_mark_page_halfdead(), _bt_moveright(), _bt_newlevel(), _bt_pagedel(), _bt_readpage(), _bt_restore_meta(), _bt_search(), _bt_simpledel_pass(), _bt_split(), _bt_unlink_halfdead_page(), _hash_addovflpage(), _hash_checkpage(), _hash_doinsert(), _hash_first(), _hash_freeovflpage(), _hash_getnewbuf(), _hash_readnext(), _hash_readpage(), _hash_splitbucket(), allocNewBuffer(), AsyncReadBuffers(), BitmapHeapScanNextBlock(), blinsert(), BloomInitMetapage(), brin_doinsert(), brin_doupdate(), brin_getinsertbuffer(), brin_initialize_empty_new_buffer(), brin_page_cleanup(), brin_xlog_insert_update(), brinbuild(), brinGetTupleForHeapBlock(), btvacuumpage(), check_index_page(), CheckReadBuffersOperation(), collect_corrupt_items(), collectMatchBitmap(), createPostingTree(), dataBeginPlaceToPageLeaf(), dataPrepareDownlink(), doPickSplit(), entryPrepareDownlink(), fill_seq_fork_with_data(), ginEntryInsert(), ginFindParents(), ginFinishSplit(), ginPlaceToPage(), ginRedoDeleteListPages(), ginRedoUpdateMetapage(), ginScanToDelete(), gistbufferinginserttuples(), gistbuild(), gistcheckpage(), gistdeletepage(), gistformdownlink(), gistinserttuples(), gistMemorizeAllDownlinks(), gistplacetopage(), gistRelocateBuildBuffersOnSplit(), gistScanPage(), gistvacuumpage(), hash_xlog_add_ovfl_page(), heap_delete(), heap_fetch_next_buffer(), heap_hot_search_buffer(), heap_insert(), heap_multi_insert(), heap_page_would_be_all_visible(), heap_prepare_pagescan(), heap_update(), heap_xlog_confirm(), heap_xlog_lock(), heapam_scan_analyze_next_block(), heapgettup(), heapgettup_pagemode(), index_compute_xid_horizon_for_tuples(), lazy_scan_heap(), lazy_scan_noprune(), lazy_scan_prune(), lazy_vacuum_heap_rel(), makeSublist(), moveLeafs(), moveRightIfItNeeded(), pgstathashindex(), prune_freeze_plan(), read_stream_start_pending_read(), ReadBufferBI(), RelationAddBlocks(), RelationCopyStorageUsingBuffer(), RelationGetBufferForTuple(), RelationPutHeapTuple(), revmap_get_buffer(), revmap_physical_extend(), ScanSourceDatabasePgClassPage(), spgAddNodeAction(), spgbuild(), spgdoinsert(), SpGistSetLastUsedPage(), spgSplitNodeAction(), spgvacuumpage(), spgWalk(), StartReadBuffersImpl(), startScanEntry(), terminate_brin_buildstate(), vacuumLeafPage(), verify_heapam(), visibilitymap_clear(), visibilitymap_get_status(), visibilitymap_pin(), visibilitymap_pin_ok(), visibilitymap_set(), and visibilitymap_set_vmbits().

◆ BufferGetLSNAtomic()

XLogRecPtr BufferGetLSNAtomic ( Buffer buffer )

Definition at line 4499 of file bufmgr.c.

{
    char       *page = BufferGetPage(buffer);
    BufferDesc *bufHdr;
    XLogRecPtr  lsn;
 
    /*
     * If we don't need locking for correctness, fastpath out.
     */
    if (!XLogHintBitIsNeeded() || BufferIsLocal(buffer))
        return PageGetLSN(page);
 
    /* Make sure we've got a real buffer, and that we hold a pin on it. */
    Assert(BufferIsValid(buffer));
    Assert(BufferIsPinned(buffer));
 
    bufHdr = GetBufferDescriptor(buffer - 1);
    LockBufHdr(bufHdr);
    lsn = PageGetLSN(page);
    UnlockBufHdr(bufHdr);
 
    return lsn;
}

References Assert(), PrivateRefCountEntry::buffer, BufferGetPage(), BufferIsLocal, BufferIsPinned, BufferIsValid(), GetBufferDescriptor(), LockBufHdr(), PageGetLSN(), UnlockBufHdr(), and XLogHintBitIsNeeded.

Referenced by _bt_drop_lock_and_maybe_pin(), _bt_killitems(), gistdoinsert(), gistFindPath(), gistkillitems(), gistScanPage(), SetHintBits(), and XLogSaveBufferForHint().

◆ BufferGetPage()

static Page BufferGetPage ( Buffer buffer )

inlinestatic

Definition at line 436 of file bufmgr.h.

{
    return (Page) BufferGetBlock(buffer);
}

References BufferGetBlock().

Referenced by _bt_allocbuf(), _bt_binsrch(), _bt_binsrch_insert(), _bt_bottomupdel_pass(), _bt_check_unique(), _bt_checkpage(), _bt_clear_incomplete_split(), _bt_conditionallockbuf(), _bt_dedup_pass(), _bt_delete_or_dedup_one_page(), _bt_delitems_delete(), _bt_delitems_delete_check(), _bt_delitems_vacuum(), _bt_endpoint(), _bt_findinsertloc(), _bt_finish_split(), _bt_get_endpoint(), _bt_getmeta(), _bt_getroot(), _bt_getstackbuf(), _bt_gettrueroot(), _bt_insert_parent(), _bt_insertonpg(), _bt_killitems(), _bt_leftsib_splitflag(), _bt_lock_and_validate_left(), _bt_lock_subtree_parent(), _bt_lockbuf(), _bt_mark_page_halfdead(), _bt_moveright(), _bt_newlevel(), _bt_pagedel(), _bt_readnextpage(), _bt_readpage(), _bt_restore_meta(), _bt_rightsib_halfdeadflag(), _bt_search(), _bt_search_insert(), _bt_set_cleanup_info(), _bt_simpledel_pass(), _bt_split(), _bt_stepright(), _bt_unlink_halfdead_page(), _bt_unlockbuf(), _bt_upgradelockbufcleanup(), _bt_vacuum_needs_cleanup(), _hash_addovflpage(), _hash_checkpage(), _hash_doinsert(), _hash_expandtable(), _hash_finish_split(), _hash_first(), _hash_freeovflpage(), _hash_get_newblock_from_oldbucket(), _hash_get_oldblock_from_newbucket(), _hash_getbucketbuf_from_hashkey(), _hash_getcachedmetap(), _hash_getinitbuf(), _hash_getnewbuf(), _hash_init(), _hash_init_metabuffer(), _hash_initbitmapbuffer(), _hash_initbuf(), _hash_kill_items(), _hash_pgaddmultitup(), _hash_pgaddtup(), _hash_readnext(), _hash_readpage(), _hash_readprev(), _hash_splitbucket(), _hash_squeezebucket(), _hash_vacuum_one_page(), allocNewBuffer(), BitmapHeapScanNextBlock(), blgetbitmap(), blinsert(), BloomNewBuffer(), blvacuumcleanup(), brin_can_do_samepage_update(), brin_doinsert(), brin_doupdate(), brin_evacuate_page(), brin_getinsertbuffer(), brin_initialize_empty_new_buffer(), brin_page_cleanup(), brin_start_evacuating_page(), brin_xlog_createidx(), brin_xlog_desummarize_page(), brin_xlog_insert_update(), brin_xlog_revmap_extend(), brin_xlog_samepage_update(), brin_xlog_update(), brinbuild(), brinbuildempty(), brinGetStats(), brinGetTupleForHeapBlock(), brininsert(), brinRevmapDesummarizeRange(), brinRevmapInitialize(), brinSetHeapBlockItemptr(), bt_metap(), bt_page_items_internal(), bt_recheck_sibling_links(), bt_rootdescend(), btree_xlog_dedup(), btree_xlog_delete(), btree_xlog_insert(), btree_xlog_mark_page_halfdead(), btree_xlog_newroot(), btree_xlog_split(), btree_xlog_unlink_page(), btree_xlog_vacuum(), btvacuumpage(), BufferGetLSNAtomic(), check_index_page(), collect_corrupt_items(), collect_visibility_data(), collectMatchBitmap(), collectMatchesForHeapRow(), count_nondeletable_pages(), createPostingTree(), dataBeginPlaceToPage(), dataBeginPlaceToPageInternal(), dataBeginPlaceToPageLeaf(), dataExecPlaceToPage(), dataExecPlaceToPageInternal(), dataLocateItem(), dataPlaceToPageLeafRecompress(), dataPrepareDownlink(), dataSplitPageInternal(), doPickSplit(), entryExecPlaceToPage(), entryIsEnoughSpace(), entryLoadMoreItems(), entryLocateEntry(), entryLocateLeafEntry(), entryPrepareDownlink(), entrySplitPage(), fill_seq_fork_with_data(), FreeSpaceMapPrepareTruncateRel(), fsm_readbuf(), fsm_search(), fsm_search_avail(), fsm_set_and_search(), fsm_vacuum_page(), generic_redo(), GenericXLogFinish(), GenericXLogRegisterBuffer(), get_raw_page_internal(), GetBTPageStatistics(), GetRecordedFreeSpace(), GetVisibilityMapPins(), gin_check_parent_keys_consistency(), gin_check_posting_tree_parent_keys_consistency(), gin_refind_parent(), ginbulkdelete(), ginDeletePage(), ginEntryInsert(), ginFindLeafPage(), ginFindParents(), ginFinishOldSplit(), ginFinishSplit(), ginGetStats(), ginHeapTupleFastInsert(), GinInitBuffer(), GinInitMetabuffer(), ginInsertCleanup(), ginInsertValue(), GinNewBuffer(), ginPlaceToPage(), ginRedoClearIncompleteSplit(), ginRedoCreatePTree(), ginRedoDeleteListPages(), ginRedoDeletePage(), ginRedoInsert(), ginRedoInsertData(), ginRedoInsertEntry(), ginRedoInsertListPage(), ginRedoUpdateMetapage(), ginRedoVacuumDataLeafPage(), ginScanToDelete(), ginStepRight(), ginTraverseLock(), ginUpdateStats(), ginvacuumcleanup(), ginVacuumEntryPage(), ginVacuumPostingTreeLeaf(), ginVacuumPostingTreeLeaves(), gistBufferingFindCorrectParent(), gistbufferinginserttuples(), gistbuild(), gistcheckpage(), gistdeletepage(), gistdoinsert(), gistFindCorrectParent(), gistFindPath(), gistfixsplit(), gistformdownlink(), gistGetMaxLevel(), GISTInitBuffer(), gistkillitems(), gistMemorizeAllDownlinks(), gistNewBuffer(), gistplacetopage(), gistProcessItup(), gistRedoClearFollowRight(), gistRedoDeleteRecord(), gistRedoPageDelete(), gistRedoPageSplitRecord(), gistRedoPageUpdateRecord(), gistScanPage(), gistvacuum_delete_empty_pages(), gistvacuumpage(), hash_bitmap_info(), hash_xlog_add_ovfl_page(), hash_xlog_delete(), hash_xlog_init_bitmap_page(), hash_xlog_init_meta_page(), hash_xlog_insert(), hash_xlog_move_page_contents(), hash_xlog_split_allocate_page(), hash_xlog_split_cleanup(), hash_xlog_split_complete(), hash_xlog_squeeze_page(), hash_xlog_update_meta_page(), hash_xlog_vacuum_one_page(), hashbucketcleanup(), hashbulkdelete(), heap_abort_speculative(), heap_delete(), heap_fetch(), heap_finish_speculative(), heap_force_common(), heap_freeze_prepared_tuples(), heap_get_latest_tid(), heap_hot_search_buffer(), heap_index_delete_tuples(), heap_inplace_update_and_unlock(), heap_insert(), heap_lock_tuple(), heap_lock_updated_tuple_rec(), heap_multi_insert(), heap_page_prune_and_freeze(), heap_page_prune_execute(), heap_page_prune_opt(), heap_page_would_be_all_visible(), heap_pre_freeze_checks(), heap_prepare_pagescan(), heap_update(), heap_xlog_confirm(), heap_xlog_delete(), heap_xlog_inplace(), heap_xlog_insert(), heap_xlog_lock(), heap_xlog_lock_updated(), heap_xlog_multi_insert(), heap_xlog_prune_freeze(), heap_xlog_update(), heap_xlog_visible(), heapam_index_build_range_scan(), heapam_index_validate_scan(), heapam_scan_analyze_next_tuple(), heapam_scan_bitmap_next_tuple(), heapam_scan_sample_next_tuple(), heapgettup_continue_page(), heapgettup_pagemode(), heapgettup_start_page(), index_compute_xid_horizon_for_tuples(), initBloomState(), lazy_scan_heap(), lazy_vacuum_heap_page(), lazy_vacuum_heap_rel(), log_heap_prune_and_freeze(), log_heap_update(), log_newpage_buffer(), log_newpage_range(), log_split_page(), MarkBufferDirtyHint(), modify_rel_block(), moveLeafs(), moveRightIfItNeeded(), nextval_internal(), palloc_btree_page(), pg_get_sequence_data(), pg_visibility(), pgstat_btree_page(), pgstat_gist_page(), pgstat_hash_page(), pgstat_heap(), pgstatginindex_internal(), pgstathashindex(), pgstatindex_impl(), prune_freeze_plan(), read_seq_tuple(), RelationAddBlocks(), RelationCopyStorageUsingBuffer(), RelationGetBufferForTuple(), RelationPutHeapTuple(), revmap_physical_extend(), scanGetCandidate(), scanPendingInsert(), scanPostingTree(), ScanSourceDatabasePgClass(), seq_redo(), SetSequence(), shiftList(), spgAddNodeAction(), spgbuild(), spgdoinsert(), spgGetCache(), SpGistGetBuffer(), SpGistInitBuffer(), SpGistNewBuffer(), SpGistSetLastUsedPage(), SpGistUpdateMetaPage(), spgprocesspending(), spgRedoAddLeaf(), spgRedoAddNode(), spgRedoMoveLeafs(), spgRedoPickSplit(), spgRedoSplitTuple(), spgRedoVacuumLeaf(), spgRedoVacuumRedirect(), spgRedoVacuumRoot(), spgSplitNodeAction(), spgvacuumpage(), spgWalk(), startScanEntry(), statapprox_heap(), terminate_brin_buildstate(), vacuumLeafPage(), vacuumLeafRoot(), vacuumRedirectAndPlaceholder(), verify_heapam(), verifyBackupPageConsistency(), visibilitymap_clear(), visibilitymap_count(), visibilitymap_get_status(), visibilitymap_prepare_truncate(), visibilitymap_set(), visibilitymap_set_vmbits(), vm_readbuf(), writeListPage(), XLogCheckBufferNeedsBackup(), XLogReadBufferExtended(), XLogReadBufferForRedoExtended(), XLogRecordPageWithFreeSpace(), XLogRegisterBuffer(), XLogSaveBufferForHint(), xlogVacuumPage(), and ZeroAndLockBuffer().

◆ BufferGetPageSize()

static Size BufferGetPageSize ( Buffer buffer )

inlinestatic

Definition at line 425 of file bufmgr.h.

{
    Assert(BufferIsValid(buffer));
    return (Size) BLCKSZ;
}

References Assert(), and BufferIsValid().

Referenced by _bt_allocbuf(), _bt_restore_meta(), _bt_split(), _hash_freeovflpage(), _hash_getinitbuf(), _hash_getnewbuf(), _hash_init_metabuffer(), _hash_initbitmapbuffer(), _hash_initbuf(), btree_xlog_mark_page_halfdead(), btree_xlog_newroot(), btree_xlog_split(), btree_xlog_unlink_page(), fill_seq_fork_with_data(), GinInitBuffer(), GinInitMetabuffer(), hash_xlog_squeeze_page(), heap_xlog_insert(), heap_xlog_multi_insert(), heap_xlog_update(), modify_rel_block(), RelationAddBlocks(), RelationGetBufferForTuple(), seq_redo(), and SpGistInitBuffer().

◆ BufferGetTag()

void BufferGetTag	(	Buffer	buffer,
		RelFileLocator *	rlocator,
		ForkNumber *	forknum,
		BlockNumber *	blknum
	)

Definition at line 4244 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    /* Do the same checks as BufferGetBlockNumber. */
    Assert(BufferIsPinned(buffer));
 
    if (BufferIsLocal(buffer))
        bufHdr = GetLocalBufferDescriptor(-buffer - 1);
    else
        bufHdr = GetBufferDescriptor(buffer - 1);
 
    /* pinned, so OK to read tag without spinlock */
    *rlocator = BufTagGetRelFileLocator(&bufHdr->tag);
    *forknum = BufTagGetForkNum(&bufHdr->tag);
    *blknum = bufHdr->tag.blockNum;
}

References Assert(), buftag::blockNum, PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsPinned, BufTagGetForkNum(), BufTagGetRelFileLocator(), GetBufferDescriptor(), GetLocalBufferDescriptor(), and BufferDesc::tag.

Referenced by fsm_search_avail(), ginRedoInsertEntry(), heap_inplace_update_and_unlock(), log_newpage_buffer(), ResolveCminCmaxDuringDecoding(), XLogRegisterBuffer(), and XLogSaveBufferForHint().

◆ BufferIsDirty()

bool BufferIsDirty ( Buffer buffer )

Definition at line 2911 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    Assert(BufferIsPinned(buffer));
 
    if (BufferIsLocal(buffer))
    {
        int         bufid = -buffer - 1;
 
        bufHdr = GetLocalBufferDescriptor(bufid);
        /* Content locks are not maintained for local buffers. */
    }
    else
    {
        bufHdr = GetBufferDescriptor(buffer - 1);
        Assert(BufferIsLockedByMeInMode(buffer, BUFFER_LOCK_EXCLUSIVE));
    }
 
    return pg_atomic_read_u32(&bufHdr->state) & BM_DIRTY;
}

References Assert(), BM_DIRTY, PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BufferIsLocal, BufferIsLockedByMeInMode(), BufferIsPinned, GetBufferDescriptor(), GetLocalBufferDescriptor(), pg_atomic_read_u32(), and BufferDesc::state.

Referenced by heap_multi_insert(), heap_xlog_multi_insert(), heap_xlog_prune_freeze(), log_heap_prune_and_freeze(), and XLogRegisterBuffer().

◆ BufferIsLockedByMe()

bool BufferIsLockedByMe ( Buffer buffer )

Definition at line 2843 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    Assert(BufferIsPinned(buffer));
 
    if (BufferIsLocal(buffer))
    {
        /* Content locks are not maintained for local buffers. */
        return true;
    }
    else
    {
        bufHdr = GetBufferDescriptor(buffer - 1);
        return LWLockHeldByMe(BufferDescriptorGetContentLock(bufHdr));
    }
}

References Assert(), PrivateRefCountEntry::buffer, BufferDescriptorGetContentLock(), BufferIsLocal, BufferIsPinned, GetBufferDescriptor(), and LWLockHeldByMe().

Referenced by FlushOneBuffer(), and MarkBufferDirtyHint().

◆ BufferIsLockedByMeInMode()

bool BufferIsLockedByMeInMode	(	Buffer	buffer,
		BufferLockMode	mode
	)

Definition at line 2869 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    Assert(BufferIsPinned(buffer));
 
    if (BufferIsLocal(buffer))
    {
        /* Content locks are not maintained for local buffers. */
        return true;
    }
    else
    {
        LWLockMode  lw_mode;
 
        switch (mode)
        {
            case BUFFER_LOCK_EXCLUSIVE:
                lw_mode = LW_EXCLUSIVE;
                break;
            case BUFFER_LOCK_SHARE:
                lw_mode = LW_SHARED;
                break;
            default:
                pg_unreachable();
        }
 
        bufHdr = GetBufferDescriptor(buffer - 1);
        return LWLockHeldByMeInMode(BufferDescriptorGetContentLock(bufHdr),
                                    lw_mode);
    }
}

References Assert(), PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BUFFER_LOCK_SHARE, BufferDescriptorGetContentLock(), BufferIsLocal, BufferIsPinned, GetBufferDescriptor(), LW_EXCLUSIVE, LW_SHARED, LWLockHeldByMeInMode(), mode, and pg_unreachable.

Referenced by BufferIsDirty(), IsBufferCleanupOK(), MarkBufferDirty(), visibilitymap_set(), visibilitymap_set_vmbits(), and XLogRegisterBuffer().

◆ BufferIsPermanent()

bool BufferIsPermanent ( Buffer buffer )

Definition at line 4469 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    /* Local buffers are used only for temp relations. */
    if (BufferIsLocal(buffer))
        return false;
 
    /* Make sure we've got a real buffer, and that we hold a pin on it. */
    Assert(BufferIsValid(buffer));
    Assert(BufferIsPinned(buffer));
 
    /*
     * BM_PERMANENT can't be changed while we hold a pin on the buffer, so we
     * need not bother with the buffer header spinlock.  Even if someone else
     * changes the buffer header state while we're doing this, the state is
     * changed atomically, so we'll read the old value or the new value, but
     * not random garbage.
     */
    bufHdr = GetBufferDescriptor(buffer - 1);
    return (pg_atomic_read_u32(&bufHdr->state) & BM_PERMANENT) != 0;
}

References Assert(), BM_PERMANENT, PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsPinned, BufferIsValid(), GetBufferDescriptor(), pg_atomic_read_u32(), and BufferDesc::state.

Referenced by SetHintBits().

◆ BufferIsValid()

static bool BufferIsValid ( Buffer bufnum )

inlinestatic

Definition at line 387 of file bufmgr.h.

{
    Assert(bufnum <= NBuffers);
    Assert(bufnum >= -NLocBuffer);
 
    return bufnum != InvalidBuffer;
}

References Assert(), InvalidBuffer, NBuffers, and NLocBuffer.

Referenced by _bt_clear_incomplete_split(), _bt_endpoint(), _bt_first(), _bt_get_endpoint(), _bt_insertonpg(), _bt_relandgetbuf(), _bt_search(), _bt_unlink_halfdead_page(), _hash_addovflpage(), _hash_dropscanbuf(), _hash_freeovflpage(), _hash_getbucketbuf_from_hashkey(), _hash_getcachedmetap(), _hash_readnext(), _hash_readpage(), _hash_readprev(), BitmapHeapScanNextBlock(), brin_doinsert(), brin_doupdate(), brin_getinsertbuffer(), brin_xlog_createidx(), brin_xlog_desummarize_page(), brin_xlog_insert_update(), brin_xlog_revmap_extend(), brin_xlog_samepage_update(), brin_xlog_update(), brinGetTupleForHeapBlock(), brininsert(), brinsummarize(), bt_recheck_sibling_links(), bt_rootdescend(), btree_xlog_dedup(), btree_xlog_delete(), btree_xlog_insert(), btree_xlog_mark_page_halfdead(), btree_xlog_split(), btree_xlog_unlink_page(), btree_xlog_vacuum(), btvacuumscan(), buffer_readv_complete(), BufferGetBlock(), BufferGetLSNAtomic(), BufferGetPageSize(), BufferIsPermanent(), ConditionalLockBufferForCleanup(), DebugPrintBufferRefcount(), doPickSplit(), entryGetItem(), entryLoadMoreItems(), EvictUnpinnedBuffer(), ExecStoreBufferHeapTuple(), ExecStorePinnedBufferHeapTuple(), FreeSpaceMapPrepareTruncateRel(), fsm_search(), fsm_vacuum_page(), generic_redo(), GetPrivateRefCount(), GetPrivateRefCountEntry(), GetRecordedFreeSpace(), GetVisibilityMapPins(), ginPlaceToPage(), ginRedoClearIncompleteSplit(), ginRedoDeletePage(), ginRedoInsert(), ginRedoUpdateMetapage(), ginRedoVacuumDataLeafPage(), ginScanToDelete(), gistinserttuples(), gistkillitems(), gistplacetopage(), gistRedoClearFollowRight(), gistRedoDeleteRecord(), gistRedoPageDelete(), gistRedoPageUpdateRecord(), gistvacuumscan(), gistXLogSplit(), gistXLogUpdate(), hash_xlog_add_ovfl_page(), hash_xlog_delete(), hash_xlog_init_bitmap_page(), hash_xlog_init_meta_page(), hash_xlog_insert(), hash_xlog_move_page_contents(), hash_xlog_split_allocate_page(), hash_xlog_split_cleanup(), hash_xlog_split_complete(), hash_xlog_squeeze_page(), hash_xlog_update_meta_page(), hash_xlog_vacuum_one_page(), heap_endscan(), heap_fetch_next_buffer(), heap_index_delete_tuples(), heap_inplace_lock(), heap_lock_tuple(), heap_rescan(), heap_update(), heap_vac_scan_next_block(), heap_xlog_confirm(), heap_xlog_delete(), heap_xlog_inplace(), heap_xlog_insert(), heap_xlog_lock(), heap_xlog_lock_updated(), heap_xlog_multi_insert(), heap_xlog_prune_freeze(), heap_xlog_update(), heap_xlog_visible(), heapam_index_fetch_reset(), heapam_scan_analyze_next_block(), heapam_scan_sample_next_block(), heapam_tuple_lock(), heapam_tuple_satisfies_snapshot(), heapgettup(), heapgettup_continue_page(), heapgettup_pagemode(), heapgettup_start_page(), invalidate_rel_block(), IsBufferCleanupOK(), lazy_scan_heap(), lazy_vacuum_heap_rel(), log_heap_visible(), MarkBufferDirty(), MarkBufferDirtyHint(), read_stream_next_buffer(), ReadRecentBuffer(), ReleaseAndReadBuffer(), ReleaseBuffer(), ResOwnerReleaseBufferPin(), spgRedoAddLeaf(), spgRedoAddNode(), spgRedoMoveLeafs(), spgRedoPickSplit(), spgRedoSplitTuple(), spgRedoVacuumLeaf(), spgRedoVacuumRedirect(), spgRedoVacuumRoot(), spgvacuumscan(), statapprox_heap(), tts_buffer_heap_clear(), tts_buffer_heap_copyslot(), tts_buffer_heap_materialize(), tts_buffer_heap_store_tuple(), verify_heapam(), verifyBackupPageConsistency(), visibilitymap_clear(), visibilitymap_count(), visibilitymap_get_status(), visibilitymap_pin(), visibilitymap_pin_ok(), visibilitymap_prepare_truncate(), visibilitymap_set(), visibilitymap_set_vmbits(), XLogPrefetcherNextBlock(), XLogReadBufferExtended(), and XLogReadBufferForRedoExtended().

◆ BufferManagerShmemInit()

void BufferManagerShmemInit ( void )

Definition at line 68 of file buf_init.c.

{
    bool        foundBufs,
                foundDescs,
                foundIOCV,
                foundBufCkpt;
 
    /* Align descriptors to a cacheline boundary. */
    BufferDescriptors = (BufferDescPadded *)
        ShmemInitStruct("Buffer Descriptors",
                        NBuffers * sizeof(BufferDescPadded),
                        &foundDescs);
 
    /* Align buffer pool on IO page size boundary. */
    BufferBlocks = (char *)
        TYPEALIGN(PG_IO_ALIGN_SIZE,
                  ShmemInitStruct("Buffer Blocks",
                                  NBuffers * (Size) BLCKSZ + PG_IO_ALIGN_SIZE,
                                  &foundBufs));
 
    /* Align condition variables to cacheline boundary. */
    BufferIOCVArray = (ConditionVariableMinimallyPadded *)
        ShmemInitStruct("Buffer IO Condition Variables",
                        NBuffers * sizeof(ConditionVariableMinimallyPadded),
                        &foundIOCV);
 
    /*
     * The array used to sort to-be-checkpointed buffer ids is located in
     * shared memory, to avoid having to allocate significant amounts of
     * memory at runtime. As that'd be in the middle of a checkpoint, or when
     * the checkpointer is restarted, memory allocation failures would be
     * painful.
     */
    CkptBufferIds = (CkptSortItem *)
        ShmemInitStruct("Checkpoint BufferIds",
                        NBuffers * sizeof(CkptSortItem), &foundBufCkpt);
 
    if (foundDescs || foundBufs || foundIOCV || foundBufCkpt)
    {
        /* should find all of these, or none of them */
        Assert(foundDescs && foundBufs && foundIOCV && foundBufCkpt);
        /* note: this path is only taken in EXEC_BACKEND case */
    }
    else
    {
        int         i;
 
        /*
         * Initialize all the buffer headers.
         */
        for (i = 0; i < NBuffers; i++)
        {
            BufferDesc *buf = GetBufferDescriptor(i);
 
            ClearBufferTag(&buf->tag);
 
            pg_atomic_init_u32(&buf->state, 0);
            buf->wait_backend_pgprocno = INVALID_PROC_NUMBER;
 
            buf->buf_id = i;
 
            pgaio_wref_clear(&buf->io_wref);
 
            LWLockInitialize(BufferDescriptorGetContentLock(buf),
                             LWTRANCHE_BUFFER_CONTENT);
 
            ConditionVariableInit(BufferDescriptorGetIOCV(buf));
        }
    }
 
    /* Init other shared buffer-management stuff */
    StrategyInitialize(!foundDescs);
 
    /* Initialize per-backend file flush context */
    WritebackContextInit(&BackendWritebackContext,
                         &backend_flush_after);
}

References Assert(), backend_flush_after, BackendWritebackContext, buf, BufferBlocks, BufferDescriptorGetContentLock(), BufferDescriptorGetIOCV(), BufferDescriptors, BufferIOCVArray, CkptBufferIds, ClearBufferTag(), ConditionVariableInit(), GetBufferDescriptor(), i, INVALID_PROC_NUMBER, LWLockInitialize(), NBuffers, pg_atomic_init_u32(), PG_IO_ALIGN_SIZE, pgaio_wref_clear(), ShmemInitStruct(), StrategyInitialize(), TYPEALIGN, and WritebackContextInit().

Referenced by CreateOrAttachShmemStructs().

◆ BufferManagerShmemSize()

Size BufferManagerShmemSize ( void )

Definition at line 153 of file buf_init.c.

{
    Size        size = 0;
 
    /* size of buffer descriptors */
    size = add_size(size, mul_size(NBuffers, sizeof(BufferDescPadded)));
    /* to allow aligning buffer descriptors */
    size = add_size(size, PG_CACHE_LINE_SIZE);
 
    /* size of data pages, plus alignment padding */
    size = add_size(size, PG_IO_ALIGN_SIZE);
    size = add_size(size, mul_size(NBuffers, BLCKSZ));
 
    /* size of stuff controlled by freelist.c */
    size = add_size(size, StrategyShmemSize());
 
    /* size of I/O condition variables */
    size = add_size(size, mul_size(NBuffers,
                                   sizeof(ConditionVariableMinimallyPadded)));
    /* to allow aligning the above */
    size = add_size(size, PG_CACHE_LINE_SIZE);
 
    /* size of checkpoint sort array in bufmgr.c */
    size = add_size(size, mul_size(NBuffers, sizeof(CkptSortItem)));
 
    return size;
}

References add_size(), mul_size(), NBuffers, PG_CACHE_LINE_SIZE, PG_IO_ALIGN_SIZE, and StrategyShmemSize().

Referenced by CalculateShmemSize().

◆ CheckBufferIsPinnedOnce()

void CheckBufferIsPinnedOnce ( Buffer buffer )

Definition at line 5651 of file bufmgr.c.

{
    if (BufferIsLocal(buffer))
    {
        if (LocalRefCount[-buffer - 1] != 1)
            elog(ERROR, "incorrect local pin count: %d",
                 LocalRefCount[-buffer - 1]);
    }
    else
    {
        if (GetPrivateRefCount(buffer) != 1)
            elog(ERROR, "incorrect local pin count: %d",
                 GetPrivateRefCount(buffer));
    }
}

References PrivateRefCountEntry::buffer, BufferIsLocal, elog, ERROR, GetPrivateRefCount(), and LocalRefCount.

Referenced by GetVictimBuffer(), lazy_scan_heap(), and LockBufferForCleanup().

◆ CheckPointBuffers()

void CheckPointBuffers ( int flags )

Definition at line 4209 of file bufmgr.c.

{
    BufferSync(flags);
}

References BufferSync().

Referenced by CheckPointGuts().

◆ ConditionalLockBuffer()

bool ConditionalLockBuffer ( Buffer buffer )

Definition at line 5630 of file bufmgr.c.

{
    BufferDesc *buf;
 
    Assert(BufferIsPinned(buffer));
    if (BufferIsLocal(buffer))
        return true;            /* act as though we got it */
 
    buf = GetBufferDescriptor(buffer - 1);
 
    return LWLockConditionalAcquire(BufferDescriptorGetContentLock(buf),
                                    LW_EXCLUSIVE);
}

References Assert(), buf, PrivateRefCountEntry::buffer, BufferDescriptorGetContentLock(), BufferIsLocal, BufferIsPinned, GetBufferDescriptor(), LW_EXCLUSIVE, and LWLockConditionalAcquire().

Referenced by _bt_conditionallockbuf(), BloomNewBuffer(), ConditionalLockBufferForCleanup(), GinNewBuffer(), gistNewBuffer(), RelationGetBufferForTuple(), spgdoinsert(), SpGistGetBuffer(), SpGistNewBuffer(), and SpGistUpdateMetaPage().

◆ ConditionalLockBufferForCleanup()

bool ConditionalLockBufferForCleanup ( Buffer buffer )

Definition at line 5857 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    uint32      buf_state,
                refcount;
 
    Assert(BufferIsValid(buffer));
 
    /* see AIO related comment in LockBufferForCleanup() */
 
    if (BufferIsLocal(buffer))
    {
        refcount = LocalRefCount[-buffer - 1];
        /* There should be exactly one pin */
        Assert(refcount > 0);
        if (refcount != 1)
            return false;
        /* Nobody else to wait for */
        return true;
    }
 
    /* There should be exactly one local pin */
    refcount = GetPrivateRefCount(buffer);
    Assert(refcount);
    if (refcount != 1)
        return false;
 
    /* Try to acquire lock */
    if (!ConditionalLockBuffer(buffer))
        return false;
 
    bufHdr = GetBufferDescriptor(buffer - 1);
    buf_state = LockBufHdr(bufHdr);
    refcount = BUF_STATE_GET_REFCOUNT(buf_state);
 
    Assert(refcount > 0);
    if (refcount == 1)
    {
        /* Successfully acquired exclusive lock with pincount 1 */
        UnlockBufHdr(bufHdr);
        return true;
    }
 
    /* Failed, so release the lock */
    UnlockBufHdr(bufHdr);
    LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
    return false;
}

References Assert(), BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BUFFER_LOCK_UNLOCK, BufferIsLocal, BufferIsValid(), ConditionalLockBuffer(), GetBufferDescriptor(), GetPrivateRefCount(), LocalRefCount, LockBuffer(), LockBufHdr(), PrivateRefCountEntry::refcount, and UnlockBufHdr().

Referenced by _hash_finish_split(), _hash_getbuf_with_condlock_cleanup(), heap_page_prune_opt(), and lazy_scan_heap().

◆ CreateAndCopyRelationData()

void CreateAndCopyRelationData	(	RelFileLocator	src_rlocator,
		RelFileLocator	dst_rlocator,
		bool	permanent
	)

Definition at line 5242 of file bufmgr.c.

{
    char        relpersistence;
    SMgrRelation src_rel;
    SMgrRelation dst_rel;
 
    /* Set the relpersistence. */
    relpersistence = permanent ?
        RELPERSISTENCE_PERMANENT : RELPERSISTENCE_UNLOGGED;
 
    src_rel = smgropen(src_rlocator, INVALID_PROC_NUMBER);
    dst_rel = smgropen(dst_rlocator, INVALID_PROC_NUMBER);
 
    /*
     * Create and copy all forks of the relation.  During create database we
     * have a separate cleanup mechanism which deletes complete database
     * directory.  Therefore, each individual relation doesn't need to be
     * registered for cleanup.
     */
    RelationCreateStorage(dst_rlocator, relpersistence, false);
 
    /* copy main fork. */
    RelationCopyStorageUsingBuffer(src_rlocator, dst_rlocator, MAIN_FORKNUM,
                                   permanent);
 
    /* copy those extra forks that exist */
    for (ForkNumber forkNum = MAIN_FORKNUM + 1;
         forkNum <= MAX_FORKNUM; forkNum++)
    {
        if (smgrexists(src_rel, forkNum))
        {
            smgrcreate(dst_rel, forkNum, false);
 
            /*
             * WAL log creation if the relation is persistent, or this is the
             * init fork of an unlogged relation.
             */
            if (permanent || forkNum == INIT_FORKNUM)
                log_smgrcreate(&dst_rlocator, forkNum);
 
            /* Copy a fork's data, block by block. */
            RelationCopyStorageUsingBuffer(src_rlocator, dst_rlocator, forkNum,
                                           permanent);
        }
    }
}

References INIT_FORKNUM, INVALID_PROC_NUMBER, log_smgrcreate(), MAIN_FORKNUM, MAX_FORKNUM, RelationCopyStorageUsingBuffer(), RelationCreateStorage(), smgrcreate(), smgrexists(), and smgropen().

Referenced by CreateDatabaseUsingWalLog().

◆ DebugPrintBufferRefcount()

char * DebugPrintBufferRefcount ( Buffer buffer )

Definition at line 4166 of file bufmgr.c.

{
    BufferDesc *buf;
    int32       loccount;
    char       *result;
    ProcNumber  backend;
    uint32      buf_state;
 
    Assert(BufferIsValid(buffer));
    if (BufferIsLocal(buffer))
    {
        buf = GetLocalBufferDescriptor(-buffer - 1);
        loccount = LocalRefCount[-buffer - 1];
        backend = MyProcNumber;
    }
    else
    {
        buf = GetBufferDescriptor(buffer - 1);
        loccount = GetPrivateRefCount(buffer);
        backend = INVALID_PROC_NUMBER;
    }
 
    /* theoretically we should lock the bufhdr here */
    buf_state = pg_atomic_read_u32(&buf->state);
 
    result = psprintf("[%03d] (rel=%s, blockNum=%u, flags=0x%x, refcount=%u %d)",
                      buffer,
                      relpathbackend(BufTagGetRelFileLocator(&buf->tag), backend,
                                     BufTagGetForkNum(&buf->tag)).str,
                      buf->tag.blockNum, buf_state & BUF_FLAG_MASK,
                      BUF_STATE_GET_REFCOUNT(buf_state), loccount);
    return result;
}

References Assert(), buf, BUF_FLAG_MASK, BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsValid(), BufTagGetForkNum(), BufTagGetRelFileLocator(), GetBufferDescriptor(), GetLocalBufferDescriptor(), GetPrivateRefCount(), INVALID_PROC_NUMBER, LocalRefCount, MyProcNumber, pg_atomic_read_u32(), psprintf(), and relpathbackend.

Referenced by buffer_call_start_io(), buffer_call_terminate_io(), CheckForBufferLeaks(), CheckForLocalBufferLeaks(), and ResOwnerPrintBufferPin().

◆ DropDatabaseBuffers()

void DropDatabaseBuffers ( Oid dbid )

Definition at line 4895 of file bufmgr.c.

{
    int         i;
 
    /*
     * We needn't consider local buffers, since by assumption the target
     * database isn't our own.
     */
 
    for (i = 0; i < NBuffers; i++)
    {
        BufferDesc *bufHdr = GetBufferDescriptor(i);
 
        /*
         * As in DropRelationBuffers, an unlocked precheck should be safe and
         * saves some cycles.
         */
        if (bufHdr->tag.dbOid != dbid)
            continue;
 
        LockBufHdr(bufHdr);
        if (bufHdr->tag.dbOid == dbid)
            InvalidateBuffer(bufHdr);   /* releases spinlock */
        else
            UnlockBufHdr(bufHdr);
    }
}

References buftag::dbOid, GetBufferDescriptor(), i, InvalidateBuffer(), LockBufHdr(), NBuffers, BufferDesc::tag, and UnlockBufHdr().

Referenced by createdb_failure_callback(), dbase_redo(), dropdb(), and movedb().

◆ DropRelationBuffers()

void DropRelationBuffers	(	SMgrRelation	smgr_reln,
		ForkNumber *	forkNum,
		int	nforks,
		BlockNumber *	firstDelBlock
	)

Definition at line 4545 of file bufmgr.c.

{
    int         i;
    int         j;
    RelFileLocatorBackend rlocator;
    BlockNumber nForkBlock[MAX_FORKNUM];
    uint64      nBlocksToInvalidate = 0;
 
    rlocator = smgr_reln->smgr_rlocator;
 
    /* If it's a local relation, it's localbuf.c's problem. */
    if (RelFileLocatorBackendIsTemp(rlocator))
    {
        if (rlocator.backend == MyProcNumber)
            DropRelationLocalBuffers(rlocator.locator, forkNum, nforks,
                                     firstDelBlock);
 
        return;
    }
 
    /*
     * To remove all the pages of the specified relation forks from the buffer
     * pool, we need to scan the entire buffer pool but we can optimize it by
     * finding the buffers from BufMapping table provided we know the exact
     * size of each fork of the relation. The exact size is required to ensure
     * that we don't leave any buffer for the relation being dropped as
     * otherwise the background writer or checkpointer can lead to a PANIC
     * error while flushing buffers corresponding to files that don't exist.
     *
     * To know the exact size, we rely on the size cached for each fork by us
     * during recovery which limits the optimization to recovery and on
     * standbys but we can easily extend it once we have shared cache for
     * relation size.
     *
     * In recovery, we cache the value returned by the first lseek(SEEK_END)
     * and the future writes keeps the cached value up-to-date. See
     * smgrextend. It is possible that the value of the first lseek is smaller
     * than the actual number of existing blocks in the file due to buggy
     * Linux kernels that might not have accounted for the recent write. But
     * that should be fine because there must not be any buffers after that
     * file size.
     */
    for (i = 0; i < nforks; i++)
    {
        /* Get the number of blocks for a relation's fork */
        nForkBlock[i] = smgrnblocks_cached(smgr_reln, forkNum[i]);
 
        if (nForkBlock[i] == InvalidBlockNumber)
        {
            nBlocksToInvalidate = InvalidBlockNumber;
            break;
        }
 
        /* calculate the number of blocks to be invalidated */
        nBlocksToInvalidate += (nForkBlock[i] - firstDelBlock[i]);
    }
 
    /*
     * We apply the optimization iff the total number of blocks to invalidate
     * is below the BUF_DROP_FULL_SCAN_THRESHOLD.
     */
    if (BlockNumberIsValid(nBlocksToInvalidate) &&
        nBlocksToInvalidate < BUF_DROP_FULL_SCAN_THRESHOLD)
    {
        for (j = 0; j < nforks; j++)
            FindAndDropRelationBuffers(rlocator.locator, forkNum[j],
                                       nForkBlock[j], firstDelBlock[j]);
        return;
    }
 
    for (i = 0; i < NBuffers; i++)
    {
        BufferDesc *bufHdr = GetBufferDescriptor(i);
 
        /*
         * We can make this a tad faster by prechecking the buffer tag before
         * we attempt to lock the buffer; this saves a lot of lock
         * acquisitions in typical cases.  It should be safe because the
         * caller must have AccessExclusiveLock on the relation, or some other
         * reason to be certain that no one is loading new pages of the rel
         * into the buffer pool.  (Otherwise we might well miss such pages
         * entirely.)  Therefore, while the tag might be changing while we
         * look at it, it can't be changing *to* a value we care about, only
         * *away* from such a value.  So false negatives are impossible, and
         * false positives are safe because we'll recheck after getting the
         * buffer lock.
         *
         * We could check forkNum and blockNum as well as the rlocator, but
         * the incremental win from doing so seems small.
         */
        if (!BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator.locator))
            continue;
 
        LockBufHdr(bufHdr);
 
        for (j = 0; j < nforks; j++)
        {
            if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rlocator.locator) &&
                BufTagGetForkNum(&bufHdr->tag) == forkNum[j] &&
                bufHdr->tag.blockNum >= firstDelBlock[j])
            {
                InvalidateBuffer(bufHdr);   /* releases spinlock */
                break;
            }
        }
        if (j >= nforks)
            UnlockBufHdr(bufHdr);
    }
}

References RelFileLocatorBackend::backend, buftag::blockNum, BlockNumberIsValid(), BUF_DROP_FULL_SCAN_THRESHOLD, BufTagGetForkNum(), BufTagMatchesRelFileLocator(), DropRelationLocalBuffers(), FindAndDropRelationBuffers(), GetBufferDescriptor(), i, InvalidateBuffer(), InvalidBlockNumber, j, RelFileLocatorBackend::locator, LockBufHdr(), MAX_FORKNUM, MyProcNumber, NBuffers, RelFileLocatorBackendIsTemp, SMgrRelationData::smgr_rlocator, smgrnblocks_cached(), BufferDesc::tag, and UnlockBufHdr().

Referenced by smgrtruncate().

◆ DropRelationsAllBuffers()

void DropRelationsAllBuffers	(	SMgrRelation *	smgr_reln,
		int	nlocators
	)

Definition at line 4665 of file bufmgr.c.

{
    int         i;
    int         n = 0;
    SMgrRelation *rels;
    BlockNumber (*block)[MAX_FORKNUM + 1];
    uint64      nBlocksToInvalidate = 0;
    RelFileLocator *locators;
    bool        cached = true;
    bool        use_bsearch;
 
    if (nlocators == 0)
        return;
 
    rels = palloc(sizeof(SMgrRelation) * nlocators);    /* non-local relations */
 
    /* If it's a local relation, it's localbuf.c's problem. */
    for (i = 0; i < nlocators; i++)
    {
        if (RelFileLocatorBackendIsTemp(smgr_reln[i]->smgr_rlocator))
        {
            if (smgr_reln[i]->smgr_rlocator.backend == MyProcNumber)
                DropRelationAllLocalBuffers(smgr_reln[i]->smgr_rlocator.locator);
        }
        else
            rels[n++] = smgr_reln[i];
    }
 
    /*
     * If there are no non-local relations, then we're done. Release the
     * memory and return.
     */
    if (n == 0)
    {
        pfree(rels);
        return;
    }
 
    /*
     * This is used to remember the number of blocks for all the relations
     * forks.
     */
    block = (BlockNumber (*)[MAX_FORKNUM + 1])
        palloc(sizeof(BlockNumber) * n * (MAX_FORKNUM + 1));
 
    /*
     * We can avoid scanning the entire buffer pool if we know the exact size
     * of each of the given relation forks. See DropRelationBuffers.
     */
    for (i = 0; i < n && cached; i++)
    {
        for (int j = 0; j <= MAX_FORKNUM; j++)
        {
            /* Get the number of blocks for a relation's fork. */
            block[i][j] = smgrnblocks_cached(rels[i], j);
 
            /* We need to only consider the relation forks that exists. */
            if (block[i][j] == InvalidBlockNumber)
            {
                if (!smgrexists(rels[i], j))
                    continue;
                cached = false;
                break;
            }
 
            /* calculate the total number of blocks to be invalidated */
            nBlocksToInvalidate += block[i][j];
        }
    }
 
    /*
     * We apply the optimization iff the total number of blocks to invalidate
     * is below the BUF_DROP_FULL_SCAN_THRESHOLD.
     */
    if (cached && nBlocksToInvalidate < BUF_DROP_FULL_SCAN_THRESHOLD)
    {
        for (i = 0; i < n; i++)
        {
            for (int j = 0; j <= MAX_FORKNUM; j++)
            {
                /* ignore relation forks that doesn't exist */
                if (!BlockNumberIsValid(block[i][j]))
                    continue;
 
                /* drop all the buffers for a particular relation fork */
                FindAndDropRelationBuffers(rels[i]->smgr_rlocator.locator,
                                           j, block[i][j], 0);
            }
        }
 
        pfree(block);
        pfree(rels);
        return;
    }
 
    pfree(block);
    locators = palloc(sizeof(RelFileLocator) * n);  /* non-local relations */
    for (i = 0; i < n; i++)
        locators[i] = rels[i]->smgr_rlocator.locator;
 
    /*
     * For low number of relations to drop just use a simple walk through, to
     * save the bsearch overhead. The threshold to use is rather a guess than
     * an exactly determined value, as it depends on many factors (CPU and RAM
     * speeds, amount of shared buffers etc.).
     */
    use_bsearch = n > RELS_BSEARCH_THRESHOLD;
 
    /* sort the list of rlocators if necessary */
    if (use_bsearch)
        qsort(locators, n, sizeof(RelFileLocator), rlocator_comparator);
 
    for (i = 0; i < NBuffers; i++)
    {
        RelFileLocator *rlocator = NULL;
        BufferDesc *bufHdr = GetBufferDescriptor(i);
 
        /*
         * As in DropRelationBuffers, an unlocked precheck should be safe and
         * saves some cycles.
         */
 
        if (!use_bsearch)
        {
            int         j;
 
            for (j = 0; j < n; j++)
            {
                if (BufTagMatchesRelFileLocator(&bufHdr->tag, &locators[j]))
                {
                    rlocator = &locators[j];
                    break;
                }
            }
        }
        else
        {
            RelFileLocator locator;
 
            locator = BufTagGetRelFileLocator(&bufHdr->tag);
            rlocator = bsearch(&locator,
                               locators, n, sizeof(RelFileLocator),
                               rlocator_comparator);
        }
 
        /* buffer doesn't belong to any of the given relfilelocators; skip it */
        if (rlocator == NULL)
            continue;
 
        LockBufHdr(bufHdr);
        if (BufTagMatchesRelFileLocator(&bufHdr->tag, rlocator))
            InvalidateBuffer(bufHdr);   /* releases spinlock */
        else
            UnlockBufHdr(bufHdr);
    }
 
    pfree(locators);
    pfree(rels);
}

References BlockNumberIsValid(), BUF_DROP_FULL_SCAN_THRESHOLD, BufTagGetRelFileLocator(), BufTagMatchesRelFileLocator(), DropRelationAllLocalBuffers(), FindAndDropRelationBuffers(), GetBufferDescriptor(), i, if(), InvalidateBuffer(), InvalidBlockNumber, j, LockBufHdr(), MAX_FORKNUM, MyProcNumber, NBuffers, palloc(), pfree(), qsort, RelFileLocatorBackendIsTemp, RELS_BSEARCH_THRESHOLD, rlocator_comparator(), smgrexists(), smgrnblocks_cached(), BufferDesc::tag, and UnlockBufHdr().

Referenced by smgrdounlinkall().

◆ EvictAllUnpinnedBuffers()

void EvictAllUnpinnedBuffers	(	int32 *	buffers_evicted,
		int32 *	buffers_flushed,
		int32 *	buffers_skipped
	)

Definition at line 6683 of file bufmgr.c.

{
    *buffers_evicted = 0;
    *buffers_skipped = 0;
    *buffers_flushed = 0;
 
    for (int buf = 1; buf <= NBuffers; buf++)
    {
        BufferDesc *desc = GetBufferDescriptor(buf - 1);
        uint32      buf_state;
        bool        buffer_flushed;
 
        CHECK_FOR_INTERRUPTS();
 
        buf_state = pg_atomic_read_u32(&desc->state);
        if (!(buf_state & BM_VALID))
            continue;
 
        ResourceOwnerEnlarge(CurrentResourceOwner);
        ReservePrivateRefCountEntry();
 
        LockBufHdr(desc);
 
        if (EvictUnpinnedBufferInternal(desc, &buffer_flushed))
            (*buffers_evicted)++;
        else
            (*buffers_skipped)++;
 
        if (buffer_flushed)
            (*buffers_flushed)++;
    }
}

References BM_VALID, buf, CHECK_FOR_INTERRUPTS, CurrentResourceOwner, EvictUnpinnedBufferInternal(), GetBufferDescriptor(), LockBufHdr(), NBuffers, pg_atomic_read_u32(), ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), and BufferDesc::state.

Referenced by pg_buffercache_evict_all().

◆ EvictRelUnpinnedBuffers()

void EvictRelUnpinnedBuffers	(	Relation	rel,
		int32 *	buffers_evicted,
		int32 *	buffers_flushed,
		int32 *	buffers_skipped
	)

Definition at line 6733 of file bufmgr.c.

{
    Assert(!RelationUsesLocalBuffers(rel));
 
    *buffers_skipped = 0;
    *buffers_evicted = 0;
    *buffers_flushed = 0;
 
    for (int buf = 1; buf <= NBuffers; buf++)
    {
        BufferDesc *desc = GetBufferDescriptor(buf - 1);
        uint32      buf_state = pg_atomic_read_u32(&(desc->state));
        bool        buffer_flushed;
 
        CHECK_FOR_INTERRUPTS();
 
        /* An unlocked precheck should be safe and saves some cycles. */
        if ((buf_state & BM_VALID) == 0 ||
            !BufTagMatchesRelFileLocator(&desc->tag, &rel->rd_locator))
            continue;
 
        /* Make sure we can pin the buffer. */
        ResourceOwnerEnlarge(CurrentResourceOwner);
        ReservePrivateRefCountEntry();
 
        buf_state = LockBufHdr(desc);
 
        /* recheck, could have changed without the lock */
        if ((buf_state & BM_VALID) == 0 ||
            !BufTagMatchesRelFileLocator(&desc->tag, &rel->rd_locator))
        {
            UnlockBufHdr(desc);
            continue;
        }
 
        if (EvictUnpinnedBufferInternal(desc, &buffer_flushed))
            (*buffers_evicted)++;
        else
            (*buffers_skipped)++;
 
        if (buffer_flushed)
            (*buffers_flushed)++;
    }
}

References Assert(), BM_VALID, buf, BufTagMatchesRelFileLocator(), CHECK_FOR_INTERRUPTS, CurrentResourceOwner, EvictUnpinnedBufferInternal(), GetBufferDescriptor(), LockBufHdr(), NBuffers, pg_atomic_read_u32(), RelationData::rd_locator, RelationUsesLocalBuffers, ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), BufferDesc::state, BufferDesc::tag, and UnlockBufHdr().

Referenced by pg_buffercache_evict_relation().

◆ EvictUnpinnedBuffer()

bool EvictUnpinnedBuffer	(	Buffer	buf,
		bool *	buffer_flushed
	)

Definition at line 6654 of file bufmgr.c.

{
    BufferDesc *desc;
 
    Assert(BufferIsValid(buf) && !BufferIsLocal(buf));
 
    /* Make sure we can pin the buffer. */
    ResourceOwnerEnlarge(CurrentResourceOwner);
    ReservePrivateRefCountEntry();
 
    desc = GetBufferDescriptor(buf - 1);
    LockBufHdr(desc);
 
    return EvictUnpinnedBufferInternal(desc, buffer_flushed);
}

References Assert(), buf, BufferIsLocal, BufferIsValid(), CurrentResourceOwner, EvictUnpinnedBufferInternal(), GetBufferDescriptor(), LockBufHdr(), ReservePrivateRefCountEntry(), and ResourceOwnerEnlarge().

Referenced by invalidate_rel_block(), modify_rel_block(), and pg_buffercache_evict().

◆ ExtendBufferedRel()

Buffer ExtendBufferedRel	(	BufferManagerRelation	bmr,
		ForkNumber	forkNum,
		BufferAccessStrategy	strategy,
		uint32	flags
	)

Definition at line 845 of file bufmgr.c.

{
    Buffer      buf;
    uint32      extend_by = 1;
 
    ExtendBufferedRelBy(bmr, forkNum, strategy, flags, extend_by,
                        &buf, &extend_by);
 
    return buf;
}

References buf, and ExtendBufferedRelBy().

Referenced by _bt_allocbuf(), _hash_getnewbuf(), BloomNewBuffer(), brinbuild(), brinbuildempty(), fill_seq_fork_with_data(), ginbuildempty(), GinNewBuffer(), gistbuildempty(), gistNewBuffer(), ReadBuffer_common(), revmap_physical_extend(), and SpGistNewBuffer().

◆ ExtendBufferedRelBy()

BlockNumber ExtendBufferedRelBy	(	BufferManagerRelation	bmr,
		ForkNumber	fork,
		BufferAccessStrategy	strategy,
		uint32	flags,
		uint32	extend_by,
		Buffer *	buffers,
		uint32 *	extended_by
	)

Definition at line 877 of file bufmgr.c.

{
    Assert((bmr.rel != NULL) != (bmr.smgr != NULL));
    Assert(bmr.smgr == NULL || bmr.relpersistence != '\0');
    Assert(extend_by > 0);
 
    if (bmr.relpersistence == '\0')
        bmr.relpersistence = bmr.rel->rd_rel->relpersistence;
 
    return ExtendBufferedRelCommon(bmr, fork, strategy, flags,
                                   extend_by, InvalidBlockNumber,
                                   buffers, extended_by);
}

References Assert(), ExtendBufferedRelCommon(), InvalidBlockNumber, RelationData::rd_rel, BufferManagerRelation::rel, BufferManagerRelation::relpersistence, and BufferManagerRelation::smgr.

Referenced by ExtendBufferedRel(), grow_rel(), and RelationAddBlocks().

◆ ExtendBufferedRelTo()

Buffer ExtendBufferedRelTo	(	BufferManagerRelation	bmr,
		ForkNumber	fork,
		BufferAccessStrategy	strategy,
		uint32	flags,
		BlockNumber	extend_to,
		ReadBufferMode	mode
	)

Definition at line 906 of file bufmgr.c.

{
    BlockNumber current_size;
    uint32      extended_by = 0;
    Buffer      buffer = InvalidBuffer;
    Buffer      buffers[64];
 
    Assert((bmr.rel != NULL) != (bmr.smgr != NULL));
    Assert(bmr.smgr == NULL || bmr.relpersistence != '\0');
    Assert(extend_to != InvalidBlockNumber && extend_to > 0);
 
    if (bmr.relpersistence == '\0')
        bmr.relpersistence = bmr.rel->rd_rel->relpersistence;
 
    /*
     * If desired, create the file if it doesn't exist.  If
     * smgr_cached_nblocks[fork] is positive then it must exist, no need for
     * an smgrexists call.
     */
    if ((flags & EB_CREATE_FORK_IF_NEEDED) &&
        (BMR_GET_SMGR(bmr)->smgr_cached_nblocks[fork] == 0 ||
         BMR_GET_SMGR(bmr)->smgr_cached_nblocks[fork] == InvalidBlockNumber) &&
        !smgrexists(BMR_GET_SMGR(bmr), fork))
    {
        LockRelationForExtension(bmr.rel, ExclusiveLock);
 
        /* recheck, fork might have been created concurrently */
        if (!smgrexists(BMR_GET_SMGR(bmr), fork))
            smgrcreate(BMR_GET_SMGR(bmr), fork, flags & EB_PERFORMING_RECOVERY);
 
        UnlockRelationForExtension(bmr.rel, ExclusiveLock);
    }
 
    /*
     * If requested, invalidate size cache, so that smgrnblocks asks the
     * kernel.
     */
    if (flags & EB_CLEAR_SIZE_CACHE)
        BMR_GET_SMGR(bmr)->smgr_cached_nblocks[fork] = InvalidBlockNumber;
 
    /*
     * Estimate how many pages we'll need to extend by. This avoids acquiring
     * unnecessarily many victim buffers.
     */
    current_size = smgrnblocks(BMR_GET_SMGR(bmr), fork);
 
    /*
     * Since no-one else can be looking at the page contents yet, there is no
     * difference between an exclusive lock and a cleanup-strength lock. Note
     * that we pass the original mode to ReadBuffer_common() below, when
     * falling back to reading the buffer to a concurrent relation extension.
     */
    if (mode == RBM_ZERO_AND_LOCK || mode == RBM_ZERO_AND_CLEANUP_LOCK)
        flags |= EB_LOCK_TARGET;
 
    while (current_size < extend_to)
    {
        uint32      num_pages = lengthof(buffers);
        BlockNumber first_block;
 
        if ((uint64) current_size + num_pages > extend_to)
            num_pages = extend_to - current_size;
 
        first_block = ExtendBufferedRelCommon(bmr, fork, strategy, flags,
                                              num_pages, extend_to,
                                              buffers, &extended_by);
 
        current_size = first_block + extended_by;
        Assert(num_pages != 0 || current_size >= extend_to);
 
        for (uint32 i = 0; i < extended_by; i++)
        {
            if (first_block + i != extend_to - 1)
                ReleaseBuffer(buffers[i]);
            else
                buffer = buffers[i];
        }
    }
 
    /*
     * It's possible that another backend concurrently extended the relation.
     * In that case read the buffer.
     *
     * XXX: Should we control this via a flag?
     */
    if (buffer == InvalidBuffer)
    {
        Assert(extended_by == 0);
        buffer = ReadBuffer_common(bmr.rel, BMR_GET_SMGR(bmr), bmr.relpersistence,
                                   fork, extend_to - 1, mode, strategy);
    }
 
    return buffer;
}

References Assert(), BMR_GET_SMGR, PrivateRefCountEntry::buffer, current_size, EB_CLEAR_SIZE_CACHE, EB_CREATE_FORK_IF_NEEDED, EB_LOCK_TARGET, EB_PERFORMING_RECOVERY, ExclusiveLock, ExtendBufferedRelCommon(), i, InvalidBlockNumber, InvalidBuffer, lengthof, LockRelationForExtension(), mode, RBM_ZERO_AND_CLEANUP_LOCK, RBM_ZERO_AND_LOCK, RelationData::rd_rel, ReadBuffer_common(), BufferManagerRelation::rel, ReleaseBuffer(), BufferManagerRelation::relpersistence, BufferManagerRelation::smgr, smgrcreate(), smgrexists(), smgrnblocks(), and UnlockRelationForExtension().

Referenced by fsm_extend(), vm_extend(), and XLogReadBufferExtended().

◆ FlushDatabaseBuffers()

void FlushDatabaseBuffers ( Oid dbid )

Definition at line 5306 of file bufmgr.c.

{
    int         i;
    BufferDesc *bufHdr;
 
    for (i = 0; i < NBuffers; i++)
    {
        uint32      buf_state;
 
        bufHdr = GetBufferDescriptor(i);
 
        /*
         * As in DropRelationBuffers, an unlocked precheck should be safe and
         * saves some cycles.
         */
        if (bufHdr->tag.dbOid != dbid)
            continue;
 
        /* Make sure we can handle the pin */
        ReservePrivateRefCountEntry();
        ResourceOwnerEnlarge(CurrentResourceOwner);
 
        buf_state = LockBufHdr(bufHdr);
        if (bufHdr->tag.dbOid == dbid &&
            (buf_state & (BM_VALID | BM_DIRTY)) == (BM_VALID | BM_DIRTY))
        {
            PinBuffer_Locked(bufHdr);
            FlushUnlockedBuffer(bufHdr, NULL, IOOBJECT_RELATION, IOCONTEXT_NORMAL);
            UnpinBuffer(bufHdr);
        }
        else
            UnlockBufHdr(bufHdr);
    }
}

References BM_DIRTY, BM_VALID, CurrentResourceOwner, buftag::dbOid, FlushUnlockedBuffer(), GetBufferDescriptor(), i, IOCONTEXT_NORMAL, IOOBJECT_RELATION, LockBufHdr(), NBuffers, PinBuffer_Locked(), ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), BufferDesc::tag, UnlockBufHdr(), and UnpinBuffer().

Referenced by dbase_redo().

◆ FlushOneBuffer()

void FlushOneBuffer ( Buffer buffer )

Definition at line 5346 of file bufmgr.c.

{
    BufferDesc *bufHdr;
 
    /* currently not needed, but no fundamental reason not to support */
    Assert(!BufferIsLocal(buffer));
 
    Assert(BufferIsPinned(buffer));
 
    bufHdr = GetBufferDescriptor(buffer - 1);
 
    Assert(BufferIsLockedByMe(buffer));
 
    FlushBuffer(bufHdr, NULL, IOOBJECT_RELATION, IOCONTEXT_NORMAL);
}

References Assert(), PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsLockedByMe(), BufferIsPinned, FlushBuffer(), GetBufferDescriptor(), IOCONTEXT_NORMAL, and IOOBJECT_RELATION.

Referenced by hash_xlog_init_bitmap_page(), hash_xlog_init_meta_page(), invalidate_rel_block(), and XLogReadBufferForRedoExtended().

◆ FlushRelationBuffers()

void FlushRelationBuffers ( Relation rel )

Definition at line 4942 of file bufmgr.c.

{
    int         i;
    BufferDesc *bufHdr;
    SMgrRelation srel = RelationGetSmgr(rel);
 
    if (RelationUsesLocalBuffers(rel))
    {
        for (i = 0; i < NLocBuffer; i++)
        {
            uint32      buf_state;
 
            bufHdr = GetLocalBufferDescriptor(i);
            if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rel->rd_locator) &&
                ((buf_state = pg_atomic_read_u32(&bufHdr->state)) &
                 (BM_VALID | BM_DIRTY)) == (BM_VALID | BM_DIRTY))
            {
                ErrorContextCallback errcallback;
 
                /* Setup error traceback support for ereport() */
                errcallback.callback = local_buffer_write_error_callback;
                errcallback.arg = bufHdr;
                errcallback.previous = error_context_stack;
                error_context_stack = &errcallback;
 
                /* Make sure we can handle the pin */
                ReservePrivateRefCountEntry();
                ResourceOwnerEnlarge(CurrentResourceOwner);
 
                /*
                 * Pin/unpin mostly to make valgrind work, but it also seems
                 * like the right thing to do.
                 */
                PinLocalBuffer(bufHdr, false);
 
 
                FlushLocalBuffer(bufHdr, srel);
 
                UnpinLocalBuffer(BufferDescriptorGetBuffer(bufHdr));
 
                /* Pop the error context stack */
                error_context_stack = errcallback.previous;
            }
        }
 
        return;
    }
 
    for (i = 0; i < NBuffers; i++)
    {
        uint32      buf_state;
 
        bufHdr = GetBufferDescriptor(i);
 
        /*
         * As in DropRelationBuffers, an unlocked precheck should be safe and
         * saves some cycles.
         */
        if (!BufTagMatchesRelFileLocator(&bufHdr->tag, &rel->rd_locator))
            continue;
 
        /* Make sure we can handle the pin */
        ReservePrivateRefCountEntry();
        ResourceOwnerEnlarge(CurrentResourceOwner);
 
        buf_state = LockBufHdr(bufHdr);
        if (BufTagMatchesRelFileLocator(&bufHdr->tag, &rel->rd_locator) &&
            (buf_state & (BM_VALID | BM_DIRTY)) == (BM_VALID | BM_DIRTY))
        {
            PinBuffer_Locked(bufHdr);
            FlushUnlockedBuffer(bufHdr, srel, IOOBJECT_RELATION, IOCONTEXT_NORMAL);
            UnpinBuffer(bufHdr);
        }
        else
            UnlockBufHdr(bufHdr);
    }
}

References ErrorContextCallback::arg, BM_DIRTY, BM_VALID, BufferDescriptorGetBuffer(), BufTagMatchesRelFileLocator(), ErrorContextCallback::callback, CurrentResourceOwner, error_context_stack, FlushLocalBuffer(), FlushUnlockedBuffer(), GetBufferDescriptor(), GetLocalBufferDescriptor(), i, IOCONTEXT_NORMAL, IOOBJECT_RELATION, local_buffer_write_error_callback(), LockBufHdr(), NBuffers, NLocBuffer, pg_atomic_read_u32(), PinBuffer_Locked(), PinLocalBuffer(), ErrorContextCallback::previous, RelationData::rd_locator, RelationGetSmgr(), RelationUsesLocalBuffers, ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), BufferDesc::state, BufferDesc::tag, UnlockBufHdr(), UnpinBuffer(), and UnpinLocalBuffer().

Referenced by fill_seq_with_data(), heapam_relation_copy_data(), and index_copy_data().

◆ FlushRelationsAllBuffers()

void FlushRelationsAllBuffers	(	SMgrRelation *	smgrs,
		int	nrels
	)

Definition at line 5030 of file bufmgr.c.

{
    int         i;
    SMgrSortArray *srels;
    bool        use_bsearch;
 
    if (nrels == 0)
        return;
 
    /* fill-in array for qsort */
    srels = palloc(sizeof(SMgrSortArray) * nrels);
 
    for (i = 0; i < nrels; i++)
    {
        Assert(!RelFileLocatorBackendIsTemp(smgrs[i]->smgr_rlocator));
 
        srels[i].rlocator = smgrs[i]->smgr_rlocator.locator;
        srels[i].srel = smgrs[i];
    }
 
    /*
     * Save the bsearch overhead for low number of relations to sync. See
     * DropRelationsAllBuffers for details.
     */
    use_bsearch = nrels > RELS_BSEARCH_THRESHOLD;
 
    /* sort the list of SMgrRelations if necessary */
    if (use_bsearch)
        qsort(srels, nrels, sizeof(SMgrSortArray), rlocator_comparator);
 
    for (i = 0; i < NBuffers; i++)
    {
        SMgrSortArray *srelent = NULL;
        BufferDesc *bufHdr = GetBufferDescriptor(i);
        uint32      buf_state;
 
        /*
         * As in DropRelationBuffers, an unlocked precheck should be safe and
         * saves some cycles.
         */
 
        if (!use_bsearch)
        {
            int         j;
 
            for (j = 0; j < nrels; j++)
            {
                if (BufTagMatchesRelFileLocator(&bufHdr->tag, &srels[j].rlocator))
                {
                    srelent = &srels[j];
                    break;
                }
            }
        }
        else
        {
            RelFileLocator rlocator;
 
            rlocator = BufTagGetRelFileLocator(&bufHdr->tag);
            srelent = bsearch(&rlocator,
                              srels, nrels, sizeof(SMgrSortArray),
                              rlocator_comparator);
        }
 
        /* buffer doesn't belong to any of the given relfilelocators; skip it */
        if (srelent == NULL)
            continue;
 
        /* Make sure we can handle the pin */
        ReservePrivateRefCountEntry();
        ResourceOwnerEnlarge(CurrentResourceOwner);
 
        buf_state = LockBufHdr(bufHdr);
        if (BufTagMatchesRelFileLocator(&bufHdr->tag, &srelent->rlocator) &&
            (buf_state & (BM_VALID | BM_DIRTY)) == (BM_VALID | BM_DIRTY))
        {
            PinBuffer_Locked(bufHdr);
            FlushUnlockedBuffer(bufHdr, srelent->srel, IOOBJECT_RELATION, IOCONTEXT_NORMAL);
            UnpinBuffer(bufHdr);
        }
        else
            UnlockBufHdr(bufHdr);
    }
 
    pfree(srels);
}

References Assert(), BM_DIRTY, BM_VALID, BufTagGetRelFileLocator(), BufTagMatchesRelFileLocator(), CurrentResourceOwner, FlushUnlockedBuffer(), GetBufferDescriptor(), i, IOCONTEXT_NORMAL, IOOBJECT_RELATION, j, RelFileLocatorBackend::locator, LockBufHdr(), NBuffers, palloc(), pfree(), PinBuffer_Locked(), qsort, RelFileLocatorBackendIsTemp, RELS_BSEARCH_THRESHOLD, ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), SMgrSortArray::rlocator, rlocator_comparator(), SMgrRelationData::smgr_rlocator, SMgrSortArray::srel, BufferDesc::tag, UnlockBufHdr(), and UnpinBuffer().

Referenced by smgrdosyncall().

◆ FreeAccessStrategy()

void FreeAccessStrategy ( BufferAccessStrategy strategy )

Definition at line 643 of file freelist.c.

{
    /* don't crash if called on a "default" strategy */
    if (strategy != NULL)
        pfree(strategy);
}

References pfree().

Referenced by blgetbitmap(), FreeBulkInsertState(), heap_endscan(), initscan(), parallel_vacuum_main(), and RelationCopyStorageUsingBuffer().

◆ GetAccessStrategy()

BufferAccessStrategy GetAccessStrategy ( BufferAccessStrategyType btype )

Definition at line 461 of file freelist.c.

{
    int         ring_size_kb;
 
    /*
     * Select ring size to use.  See buffer/README for rationales.
     *
     * Note: if you change the ring size for BAS_BULKREAD, see also
     * SYNC_SCAN_REPORT_INTERVAL in access/heap/syncscan.c.
     */
    switch (btype)
    {
        case BAS_NORMAL:
            /* if someone asks for NORMAL, just give 'em a "default" object */
            return NULL;
 
        case BAS_BULKREAD:
            {
                int         ring_max_kb;
 
                /*
                 * The ring always needs to be large enough to allow some
                 * separation in time between providing a buffer to the user
                 * of the strategy and that buffer being reused. Otherwise the
                 * user's pin will prevent reuse of the buffer, even without
                 * concurrent activity.
                 *
                 * We also need to ensure the ring always is large enough for
                 * SYNC_SCAN_REPORT_INTERVAL, as noted above.
                 *
                 * Thus we start out a minimal size and increase the size
                 * further if appropriate.
                 */
                ring_size_kb = 256;
 
                /*
                 * There's no point in a larger ring if we won't be allowed to
                 * pin sufficiently many buffers.  But we never limit to less
                 * than the minimal size above.
                 */
                ring_max_kb = GetPinLimit() * (BLCKSZ / 1024);
                ring_max_kb = Max(ring_size_kb, ring_max_kb);
 
                /*
                 * We would like the ring to additionally have space for the
                 * configured degree of IO concurrency. While being read in,
                 * buffers can obviously not yet be reused.
                 *
                 * Each IO can be up to io_combine_limit blocks large, and we
                 * want to start up to effective_io_concurrency IOs.
                 *
                 * Note that effective_io_concurrency may be 0, which disables
                 * AIO.
                 */
                ring_size_kb += (BLCKSZ / 1024) *
                    io_combine_limit * effective_io_concurrency;
 
                if (ring_size_kb > ring_max_kb)
                    ring_size_kb = ring_max_kb;
                break;
            }
        case BAS_BULKWRITE:
            ring_size_kb = 16 * 1024;
            break;
        case BAS_VACUUM:
            ring_size_kb = 2048;
            break;
 
        default:
            elog(ERROR, "unrecognized buffer access strategy: %d",
                 (int) btype);
            return NULL;        /* keep compiler quiet */
    }
 
    return GetAccessStrategyWithSize(btype, ring_size_kb);
}

References BAS_BULKREAD, BAS_BULKWRITE, BAS_NORMAL, BAS_VACUUM, effective_io_concurrency, elog, ERROR, GetAccessStrategyWithSize(), GetPinLimit(), io_combine_limit, and Max.

Referenced by blgetbitmap(), bt_check_every_level(), collect_corrupt_items(), collect_visibility_data(), GetBulkInsertState(), gin_check_parent_keys_consistency(), gin_check_posting_tree_parent_keys_consistency(), initscan(), pgstat_index(), pgstathashindex(), pgstatindex_impl(), RelationCopyStorageUsingBuffer(), ScanSourceDatabasePgClass(), statapprox_heap(), and verify_heapam().

◆ GetAccessStrategyBufferCount()

int GetAccessStrategyBufferCount ( BufferAccessStrategy strategy )

Definition at line 586 of file freelist.c.

{
    if (strategy == NULL)
        return 0;
 
    return strategy->nbuffers;
}

References BufferAccessStrategyData::nbuffers.

Referenced by parallel_vacuum_init().

◆ GetAccessStrategyPinLimit()

int GetAccessStrategyPinLimit ( BufferAccessStrategy strategy )

Definition at line 609 of file freelist.c.

{
    if (strategy == NULL)
        return NBuffers;
 
    switch (strategy->btype)
    {
        case BAS_BULKREAD:
 
            /*
             * Since BAS_BULKREAD uses StrategyRejectBuffer(), dirty buffers
             * shouldn't be a problem and the caller is free to pin up to the
             * entire ring at once.
             */
            return strategy->nbuffers;
 
        default:
 
            /*
             * Tell caller not to pin more than half the buffers in the ring.
             * This is a trade-off between look ahead distance and deferring
             * writeback and associated WAL traffic.
             */
            return strategy->nbuffers / 2;
    }
}

References BAS_BULKREAD, BufferAccessStrategyData::btype, BufferAccessStrategyData::nbuffers, and NBuffers.

Referenced by read_stream_begin_impl().

◆ GetAccessStrategyWithSize()

BufferAccessStrategy GetAccessStrategyWithSize	(	BufferAccessStrategyType	btype,
		int	ring_size_kb
	)

Definition at line 546 of file freelist.c.

{
    int         ring_buffers;
    BufferAccessStrategy strategy;
 
    Assert(ring_size_kb >= 0);
 
    /* Figure out how many buffers ring_size_kb is */
    ring_buffers = ring_size_kb / (BLCKSZ / 1024);
 
    /* 0 means unlimited, so no BufferAccessStrategy required */
    if (ring_buffers == 0)
        return NULL;
 
    /* Cap to 1/8th of shared_buffers */
    ring_buffers = Min(NBuffers / 8, ring_buffers);
 
    /* NBuffers should never be less than 16, so this shouldn't happen */
    Assert(ring_buffers > 0);
 
    /* Allocate the object and initialize all elements to zeroes */
    strategy = (BufferAccessStrategy)
        palloc0(offsetof(BufferAccessStrategyData, buffers) +
                ring_buffers * sizeof(Buffer));
 
    /* Set fields that don't start out zero */
    strategy->btype = btype;
    strategy->nbuffers = ring_buffers;
 
    return strategy;
}

References Assert(), BufferAccessStrategyData::btype, Min, BufferAccessStrategyData::nbuffers, NBuffers, and palloc0().

Referenced by do_autovacuum(), ExecVacuum(), GetAccessStrategy(), and parallel_vacuum_main().

◆ GetAdditionalLocalPinLimit()

uint32 GetAdditionalLocalPinLimit ( void )

Definition at line 315 of file localbuf.c.

{
    Assert(NLocalPinnedBuffers <= num_temp_buffers);
    return num_temp_buffers - NLocalPinnedBuffers;
}

References Assert(), NLocalPinnedBuffers, and num_temp_buffers.

Referenced by read_stream_start_pending_read().

◆ GetAdditionalPinLimit()

uint32 GetAdditionalPinLimit ( void )

Definition at line 2487 of file bufmgr.c.

{
    uint32      estimated_pins_held;
 
    /*
     * We get the number of "overflowed" pins for free, but don't know the
     * number of pins in PrivateRefCountArray.  The cost of calculating that
     * exactly doesn't seem worth it, so just assume the max.
     */
    estimated_pins_held = PrivateRefCountOverflowed + REFCOUNT_ARRAY_ENTRIES;
 
    /* Is this backend already holding more than its fair share? */
    if (estimated_pins_held > MaxProportionalPins)
        return 0;
 
    return MaxProportionalPins - estimated_pins_held;
}

References MaxProportionalPins, PrivateRefCountOverflowed, and REFCOUNT_ARRAY_ENTRIES.

Referenced by LimitAdditionalPins(), and read_stream_start_pending_read().

◆ GetLocalPinLimit()

uint32 GetLocalPinLimit ( void )

Definition at line 307 of file localbuf.c.

{
    /* Every backend has its own temporary buffers, and can pin them all. */
    return num_temp_buffers;
}

References num_temp_buffers.

Referenced by read_stream_begin_impl().

◆ GetPinLimit()

uint32 GetPinLimit ( void )

Definition at line 2475 of file bufmgr.c.

{
    return MaxProportionalPins;
}

References MaxProportionalPins.

Referenced by GetAccessStrategy(), and read_stream_begin_impl().

◆ HoldingBufferPinThatDelaysRecovery()

bool HoldingBufferPinThatDelaysRecovery ( void )

Definition at line 5831 of file bufmgr.c.

{
    int         bufid = GetStartupBufferPinWaitBufId();
 
    /*
     * If we get woken slowly then it's possible that the Startup process was
     * already woken by other backends before we got here. Also possible that
     * we get here by multiple interrupts or interrupts at inappropriate
     * times, so make sure we do nothing if the bufid is not set.
     */
    if (bufid < 0)
        return false;
 
    if (GetPrivateRefCount(bufid + 1) > 0)
        return true;
 
    return false;
}

References GetPrivateRefCount(), and GetStartupBufferPinWaitBufId().

Referenced by CheckRecoveryConflictDeadlock(), and ProcessRecoveryConflictInterrupt().

◆ IncrBufferRefCount()

void IncrBufferRefCount ( Buffer buffer )

Definition at line 5398 of file bufmgr.c.

{
    Assert(BufferIsPinned(buffer));
    ResourceOwnerEnlarge(CurrentResourceOwner);
    if (BufferIsLocal(buffer))
        LocalRefCount[-buffer - 1]++;
    else
    {
        PrivateRefCountEntry *ref;
 
        ref = GetPrivateRefCountEntry(buffer, true);
        Assert(ref != NULL);
        ref->refcount++;
    }
    ResourceOwnerRememberBuffer(CurrentResourceOwner, buffer);
}

References Assert(), PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsPinned, CurrentResourceOwner, GetPrivateRefCountEntry(), LocalRefCount, PrivateRefCountEntry::refcount, ResourceOwnerEnlarge(), and ResourceOwnerRememberBuffer().

Referenced by _bt_steppage(), btrestrpos(), entryLoadMoreItems(), ReadBufferBI(), RelationAddBlocks(), scanPostingTree(), startScanEntry(), and tts_buffer_heap_store_tuple().

◆ InitBufferManagerAccess()

void InitBufferManagerAccess ( void )

Definition at line 4007 of file bufmgr.c.

{
    HASHCTL     hash_ctl;
 
    /*
     * An advisory limit on the number of pins each backend should hold, based
     * on shared_buffers and the maximum number of connections possible.
     * That's very pessimistic, but outside toy-sized shared_buffers it should
     * allow plenty of pins.  LimitAdditionalPins() and
     * GetAdditionalPinLimit() can be used to check the remaining balance.
     */
    MaxProportionalPins = NBuffers / (MaxBackends + NUM_AUXILIARY_PROCS);
 
    memset(&PrivateRefCountArray, 0, sizeof(PrivateRefCountArray));
 
    hash_ctl.keysize = sizeof(Buffer);
    hash_ctl.entrysize = sizeof(PrivateRefCountEntry);
 
    PrivateRefCountHash = hash_create("PrivateRefCount", 100, &hash_ctl,
                                      HASH_ELEM | HASH_BLOBS);
 
    /*
     * AtProcExit_Buffers needs LWLock access, and thereby has to be called at
     * the corresponding phase of backend shutdown.
     */
    Assert(MyProc != NULL);
    on_shmem_exit(AtProcExit_Buffers, 0);
}

References Assert(), AtProcExit_Buffers(), HASHCTL::entrysize, HASH_BLOBS, hash_create(), HASH_ELEM, HASHCTL::keysize, MaxBackends, MaxProportionalPins, MyProc, NBuffers, NUM_AUXILIARY_PROCS, on_shmem_exit(), PrivateRefCountArray, and PrivateRefCountHash.

Referenced by BaseInit().

◆ IsBufferCleanupOK()

bool IsBufferCleanupOK ( Buffer buffer )

Definition at line 5915 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    uint32      buf_state;
 
    Assert(BufferIsValid(buffer));
 
    /* see AIO related comment in LockBufferForCleanup() */
 
    if (BufferIsLocal(buffer))
    {
        /* There should be exactly one pin */
        if (LocalRefCount[-buffer - 1] != 1)
            return false;
        /* Nobody else to wait for */
        return true;
    }
 
    /* There should be exactly one local pin */
    if (GetPrivateRefCount(buffer) != 1)
        return false;
 
    bufHdr = GetBufferDescriptor(buffer - 1);
 
    /* caller must hold exclusive lock on buffer */
    Assert(BufferIsLockedByMeInMode(buffer, BUFFER_LOCK_EXCLUSIVE));
 
    buf_state = LockBufHdr(bufHdr);
 
    Assert(BUF_STATE_GET_REFCOUNT(buf_state) > 0);
    if (BUF_STATE_GET_REFCOUNT(buf_state) == 1)
    {
        /* pincount is OK. */
        UnlockBufHdr(bufHdr);
        return true;
    }
 
    UnlockBufHdr(bufHdr);
    return false;
}

References Assert(), BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BufferIsLocal, BufferIsLockedByMeInMode(), BufferIsValid(), GetBufferDescriptor(), GetPrivateRefCount(), LocalRefCount, LockBufHdr(), and UnlockBufHdr().

Referenced by _hash_doinsert(), _hash_expandtable(), _hash_splitbucket(), and hashbucketcleanup().

◆ LimitAdditionalLocalPins()

void LimitAdditionalLocalPins ( uint32 * additional_pins )

Definition at line 323 of file localbuf.c.

{
    uint32      max_pins;
 
    if (*additional_pins <= 1)
        return;
 
    /*
     * In contrast to LimitAdditionalPins() other backends don't play a role
     * here. We can allow up to NLocBuffer pins in total, but it might not be
     * initialized yet so read num_temp_buffers.
     */
    max_pins = (num_temp_buffers - NLocalPinnedBuffers);
 
    if (*additional_pins >= max_pins)
        *additional_pins = max_pins;
}

References NLocalPinnedBuffers, and num_temp_buffers.

Referenced by ExtendBufferedRelLocal().

◆ LimitAdditionalPins()

void LimitAdditionalPins ( uint32 * additional_pins )

Definition at line 2513 of file bufmgr.c.

{
    uint32      limit;
 
    if (*additional_pins <= 1)
        return;
 
    limit = GetAdditionalPinLimit();
    limit = Max(limit, 1);
    if (limit < *additional_pins)
        *additional_pins = limit;
}

References GetAdditionalPinLimit(), and Max.

Referenced by ExtendBufferedRelShared().

◆ LockBuffer()

void LockBuffer	(	Buffer	buffer,
		BufferLockMode	mode
	)

Definition at line 5604 of file bufmgr.c.

{
    BufferDesc *buf;
 
    Assert(BufferIsPinned(buffer));
    if (BufferIsLocal(buffer))
        return;                 /* local buffers need no lock */
 
    buf = GetBufferDescriptor(buffer - 1);
 
    if (mode == BUFFER_LOCK_UNLOCK)
        LWLockRelease(BufferDescriptorGetContentLock(buf));
    else if (mode == BUFFER_LOCK_SHARE)
        LWLockAcquire(BufferDescriptorGetContentLock(buf), LW_SHARED);
    else if (mode == BUFFER_LOCK_EXCLUSIVE)
        LWLockAcquire(BufferDescriptorGetContentLock(buf), LW_EXCLUSIVE);
    else
        elog(ERROR, "unrecognized buffer lock mode: %d", mode);
}

References Assert(), buf, PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BUFFER_LOCK_SHARE, BUFFER_LOCK_UNLOCK, BufferDescriptorGetContentLock(), BufferIsLocal, BufferIsPinned, elog, ERROR, GetBufferDescriptor(), LW_EXCLUSIVE, LW_SHARED, LWLockAcquire(), LWLockRelease(), and mode.

Referenced by _bt_lockbuf(), _bt_unlockbuf(), _bt_upgradelockbufcleanup(), _hash_addovflpage(), _hash_doinsert(), _hash_expandtable(), _hash_finish_split(), _hash_first(), _hash_freeovflpage(), _hash_getbuf(), _hash_getbuf_with_strategy(), _hash_getcachedmetap(), _hash_init(), _hash_kill_items(), _hash_readnext(), _hash_readpage(), _hash_readprev(), _hash_splitbucket(), _hash_squeezebucket(), _hash_vacuum_one_page(), BitmapHeapScanNextBlock(), blbulkdelete(), blgetbitmap(), blinsert(), BloomInitMetapage(), BloomNewBuffer(), blvacuumcleanup(), brin_doinsert(), brin_doupdate(), brin_evacuate_page(), brin_getinsertbuffer(), brin_page_cleanup(), bringetbitmap(), brinGetStats(), brinGetTupleForHeapBlock(), brininsert(), brinLockRevmapPageForUpdate(), brinRevmapDesummarizeRange(), brinRevmapInitialize(), brinsummarize(), bt_metap(), bt_multi_page_stats(), bt_page_items_internal(), bt_page_stats_internal(), bt_recheck_sibling_links(), collect_corrupt_items(), collect_visibility_data(), collectMatchBitmap(), ConditionalLockBufferForCleanup(), count_nondeletable_pages(), create_toy_buffer(), entryLoadMoreItems(), ExtendBufferedRelShared(), FreeSpaceMapPrepareTruncateRel(), fsm_readbuf(), fsm_search(), fsm_search_avail(), fsm_set_and_search(), fsm_vacuum_page(), get_raw_page_internal(), GetVisibilityMapPins(), gin_check_parent_keys_consistency(), gin_check_posting_tree_parent_keys_consistency(), gin_refind_parent(), ginbulkdelete(), ginEntryInsert(), ginFindLeafPage(), ginFindParents(), ginFinishOldSplit(), ginFinishSplit(), ginGetStats(), ginHeapTupleFastInsert(), ginInsertCleanup(), ginInsertValue(), GinNewBuffer(), ginScanToDelete(), ginStepRight(), ginTraverseLock(), ginUpdateStats(), ginvacuumcleanup(), ginVacuumPostingTreeLeaves(), gistBufferingFindCorrectParent(), gistbufferinginserttuples(), gistdoinsert(), gistFindCorrectParent(), gistFindPath(), gistfinishsplit(), gistfixsplit(), gistformdownlink(), gistGetMaxLevel(), gistinserttuples(), gistkillitems(), gistNewBuffer(), gistProcessItup(), gistScanPage(), gistvacuum_delete_empty_pages(), gistvacuumpage(), hashbucketcleanup(), hashbulkdelete(), heap_abort_speculative(), heap_delete(), heap_fetch(), heap_finish_speculative(), heap_get_latest_tid(), heap_index_delete_tuples(), heap_inplace_lock(), heap_inplace_unlock(), heap_inplace_update_and_unlock(), heap_lock_tuple(), heap_lock_updated_tuple_rec(), heap_multi_insert(), heap_page_prune_opt(), heap_prepare_pagescan(), heap_update(), heap_xlog_visible(), heapam_index_build_range_scan(), heapam_index_fetch_tuple(), heapam_index_validate_scan(), heapam_relation_copy_for_cluster(), heapam_scan_analyze_next_block(), heapam_scan_sample_next_tuple(), heapam_tuple_satisfies_snapshot(), heapgettup(), initBloomState(), invalidate_rel_block(), lazy_scan_heap(), lazy_scan_new_or_empty(), lazy_vacuum_heap_page(), lazy_vacuum_heap_rel(), LockBufferForCleanup(), log_newpage_range(), modify_rel_block(), palloc_btree_page(), pg_visibility(), pgrowlocks(), pgstat_btree_page(), pgstat_gist_page(), pgstat_hash_page(), pgstat_heap(), pgstatginindex_internal(), pgstathashindex(), pgstatindex_impl(), read_seq_tuple(), RelationAddBlocks(), RelationCopyStorageUsingBuffer(), RelationGetBufferForTuple(), revmap_physical_extend(), scanGetCandidate(), scanPendingInsert(), ScanSourceDatabasePgClass(), shiftList(), spgdoinsert(), spgGetCache(), SpGistNewBuffer(), spgprocesspending(), spgvacuumpage(), spgWalk(), startScanEntry(), statapprox_heap(), summarize_range(), UnlockReleaseBuffer(), update_most_recent_deletion_info(), verify_heapam(), verifyBackupPageConsistency(), visibilitymap_clear(), visibilitymap_prepare_truncate(), visibilitymap_set(), vm_readbuf(), XLogReadBufferForRedoExtended(), XLogRecordPageWithFreeSpace(), and ZeroAndLockBuffer().

◆ LockBufferForCleanup()

void LockBufferForCleanup ( Buffer buffer )

Definition at line 5684 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    TimestampTz waitStart = 0;
    bool        waiting = false;
    bool        logged_recovery_conflict = false;
 
    Assert(BufferIsPinned(buffer));
    Assert(PinCountWaitBuf == NULL);
 
    CheckBufferIsPinnedOnce(buffer);
 
    /*
     * We do not yet need to be worried about in-progress AIOs holding a pin,
     * as we, so far, only support doing reads via AIO and this function can
     * only be called once the buffer is valid (i.e. no read can be in
     * flight).
     */
 
    /* Nobody else to wait for */
    if (BufferIsLocal(buffer))
        return;
 
    bufHdr = GetBufferDescriptor(buffer - 1);
 
    for (;;)
    {
        uint32      buf_state;
        uint32      unset_bits = 0;
 
        /* Try to acquire lock */
        LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
        buf_state = LockBufHdr(bufHdr);
 
        Assert(BUF_STATE_GET_REFCOUNT(buf_state) > 0);
        if (BUF_STATE_GET_REFCOUNT(buf_state) == 1)
        {
            /* Successfully acquired exclusive lock with pincount 1 */
            UnlockBufHdr(bufHdr);
 
            /*
             * Emit the log message if recovery conflict on buffer pin was
             * resolved but the startup process waited longer than
             * deadlock_timeout for it.
             */
            if (logged_recovery_conflict)
                LogRecoveryConflict(PROCSIG_RECOVERY_CONFLICT_BUFFERPIN,
                                    waitStart, GetCurrentTimestamp(),
                                    NULL, false);
 
            if (waiting)
            {
                /* reset ps display to remove the suffix if we added one */
                set_ps_display_remove_suffix();
                waiting = false;
            }
            return;
        }
        /* Failed, so mark myself as waiting for pincount 1 */
        if (buf_state & BM_PIN_COUNT_WAITER)
        {
            UnlockBufHdr(bufHdr);
            LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
            elog(ERROR, "multiple backends attempting to wait for pincount 1");
        }
        bufHdr->wait_backend_pgprocno = MyProcNumber;
        PinCountWaitBuf = bufHdr;
        UnlockBufHdrExt(bufHdr, buf_state,
                        BM_PIN_COUNT_WAITER, 0,
                        0);
        LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
 
        /* Wait to be signaled by UnpinBuffer() */
        if (InHotStandby)
        {
            if (!waiting)
            {
                /* adjust the process title to indicate that it's waiting */
                set_ps_display_suffix("waiting");
                waiting = true;
            }
 
            /*
             * Emit the log message if the startup process is waiting longer
             * than deadlock_timeout for recovery conflict on buffer pin.
             *
             * Skip this if first time through because the startup process has
             * not started waiting yet in this case. So, the wait start
             * timestamp is set after this logic.
             */
            if (waitStart != 0 && !logged_recovery_conflict)
            {
                TimestampTz now = GetCurrentTimestamp();
 
                if (TimestampDifferenceExceeds(waitStart, now,
                                               DeadlockTimeout))
                {
                    LogRecoveryConflict(PROCSIG_RECOVERY_CONFLICT_BUFFERPIN,
                                        waitStart, now, NULL, true);
                    logged_recovery_conflict = true;
                }
            }
 
            /*
             * Set the wait start timestamp if logging is enabled and first
             * time through.
             */
            if (log_recovery_conflict_waits && waitStart == 0)
                waitStart = GetCurrentTimestamp();
 
            /* Publish the bufid that Startup process waits on */
            SetStartupBufferPinWaitBufId(buffer - 1);
            /* Set alarm and then wait to be signaled by UnpinBuffer() */
            ResolveRecoveryConflictWithBufferPin();
            /* Reset the published bufid */
            SetStartupBufferPinWaitBufId(-1);
        }
        else
            ProcWaitForSignal(WAIT_EVENT_BUFFER_CLEANUP);
 
        /*
         * Remove flag marking us as waiter. Normally this will not be set
         * anymore, but ProcWaitForSignal() can return for other signals as
         * well.  We take care to only reset the flag if we're the waiter, as
         * theoretically another backend could have started waiting. That's
         * impossible with the current usages due to table level locking, but
         * better be safe.
         */
        buf_state = LockBufHdr(bufHdr);
        if ((buf_state & BM_PIN_COUNT_WAITER) != 0 &&
            bufHdr->wait_backend_pgprocno == MyProcNumber)
            unset_bits |= BM_PIN_COUNT_WAITER;
 
        UnlockBufHdrExt(bufHdr, buf_state,
                        0, unset_bits,
                        0);
 
        PinCountWaitBuf = NULL;
        /* Loop back and try again */
    }
}

References Assert(), BM_PIN_COUNT_WAITER, BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BUFFER_LOCK_UNLOCK, BufferIsLocal, BufferIsPinned, CheckBufferIsPinnedOnce(), DeadlockTimeout, elog, ERROR, GetBufferDescriptor(), GetCurrentTimestamp(), InHotStandby, LockBuffer(), LockBufHdr(), log_recovery_conflict_waits, LogRecoveryConflict(), MyProcNumber, now(), PinCountWaitBuf, PROCSIG_RECOVERY_CONFLICT_BUFFERPIN, ProcWaitForSignal(), ResolveRecoveryConflictWithBufferPin(), set_ps_display_remove_suffix(), set_ps_display_suffix(), SetStartupBufferPinWaitBufId(), TimestampDifferenceExceeds(), UnlockBufHdr(), UnlockBufHdrExt(), BufferDesc::wait_backend_pgprocno, and waiting.

Referenced by _bt_upgradelockbufcleanup(), ginVacuumPostingTree(), hashbulkdelete(), heap_force_common(), lazy_scan_heap(), XLogReadBufferForRedoExtended(), and ZeroAndLockBuffer().

◆ MarkBufferDirty()

void MarkBufferDirty ( Buffer buffer )

Definition at line 2943 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    uint32      buf_state;
    uint32      old_buf_state;
 
    if (!BufferIsValid(buffer))
        elog(ERROR, "bad buffer ID: %d", buffer);
 
    if (BufferIsLocal(buffer))
    {
        MarkLocalBufferDirty(buffer);
        return;
    }
 
    bufHdr = GetBufferDescriptor(buffer - 1);
 
    Assert(BufferIsPinned(buffer));
    Assert(BufferIsLockedByMeInMode(buffer, BUFFER_LOCK_EXCLUSIVE));
 
    /*
     * NB: We have to wait for the buffer header spinlock to be not held, as
     * TerminateBufferIO() relies on the spinlock.
     */
    old_buf_state = pg_atomic_read_u32(&bufHdr->state);
    for (;;)
    {
        if (old_buf_state & BM_LOCKED)
            old_buf_state = WaitBufHdrUnlocked(bufHdr);
 
        buf_state = old_buf_state;
 
        Assert(BUF_STATE_GET_REFCOUNT(buf_state) > 0);
        buf_state |= BM_DIRTY | BM_JUST_DIRTIED;
 
        if (pg_atomic_compare_exchange_u32(&bufHdr->state, &old_buf_state,
                                           buf_state))
            break;
    }
 
    /*
     * If the buffer was not dirty already, do vacuum accounting.
     */
    if (!(old_buf_state & BM_DIRTY))
    {
        pgBufferUsage.shared_blks_dirtied++;
        if (VacuumCostActive)
            VacuumCostBalance += VacuumCostPageDirty;
    }
}

References Assert(), BM_DIRTY, BM_JUST_DIRTIED, BM_LOCKED, BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BUFFER_LOCK_EXCLUSIVE, BufferIsLocal, BufferIsLockedByMeInMode(), BufferIsPinned, BufferIsValid(), elog, ERROR, GetBufferDescriptor(), MarkLocalBufferDirty(), pg_atomic_compare_exchange_u32(), pg_atomic_read_u32(), pgBufferUsage, BufferUsage::shared_blks_dirtied, BufferDesc::state, VacuumCostActive, VacuumCostBalance, VacuumCostPageDirty, and WaitBufHdrUnlocked().

Referenced by _bt_clear_incomplete_split(), _bt_dedup_pass(), _bt_delitems_delete(), _bt_delitems_vacuum(), _bt_getroot(), _bt_insertonpg(), _bt_mark_page_halfdead(), _bt_newlevel(), _bt_restore_meta(), _bt_set_cleanup_info(), _bt_split(), _bt_unlink_halfdead_page(), _hash_addovflpage(), _hash_doinsert(), _hash_expandtable(), _hash_freeovflpage(), _hash_init(), _hash_splitbucket(), _hash_squeezebucket(), _hash_vacuum_one_page(), addLeafTuple(), brin_doinsert(), brin_doupdate(), brin_initialize_empty_new_buffer(), brin_xlog_createidx(), brin_xlog_desummarize_page(), brin_xlog_insert_update(), brin_xlog_revmap_extend(), brin_xlog_samepage_update(), brin_xlog_update(), brinbuild(), brinbuildempty(), brinRevmapDesummarizeRange(), btree_xlog_dedup(), btree_xlog_delete(), btree_xlog_insert(), btree_xlog_mark_page_halfdead(), btree_xlog_newroot(), btree_xlog_split(), btree_xlog_unlink_page(), btree_xlog_vacuum(), createPostingTree(), dataExecPlaceToPageInternal(), dataExecPlaceToPageLeaf(), doPickSplit(), entryExecPlaceToPage(), fill_seq_fork_with_data(), FreeSpaceMapPrepareTruncateRel(), generic_redo(), GenericXLogFinish(), ginbuild(), ginbuildempty(), ginbulkdelete(), ginDeletePage(), ginHeapTupleFastInsert(), ginPlaceToPage(), ginRedoClearIncompleteSplit(), ginRedoCreatePTree(), ginRedoDeleteListPages(), ginRedoDeletePage(), ginRedoInsert(), ginRedoInsertListPage(), ginRedoUpdateMetapage(), ginRedoVacuumDataLeafPage(), ginUpdateStats(), ginVacuumPostingTreeLeaf(), gistbuild(), gistbuildempty(), gistdeletepage(), gistplacetopage(), gistprunepage(), gistRedoClearFollowRight(), gistRedoDeleteRecord(), gistRedoPageDelete(), gistRedoPageSplitRecord(), gistRedoPageUpdateRecord(), gistvacuumpage(), hash_xlog_add_ovfl_page(), hash_xlog_delete(), hash_xlog_init_bitmap_page(), hash_xlog_init_meta_page(), hash_xlog_insert(), hash_xlog_move_page_contents(), hash_xlog_split_allocate_page(), hash_xlog_split_cleanup(), hash_xlog_split_complete(), hash_xlog_squeeze_page(), hash_xlog_update_meta_page(), hash_xlog_vacuum_one_page(), hashbucketcleanup(), hashbulkdelete(), heap_abort_speculative(), heap_delete(), heap_finish_speculative(), heap_force_common(), heap_inplace_update_and_unlock(), heap_insert(), heap_lock_tuple(), heap_lock_updated_tuple_rec(), heap_multi_insert(), heap_page_prune_and_freeze(), heap_update(), heap_xlog_confirm(), heap_xlog_delete(), heap_xlog_inplace(), heap_xlog_insert(), heap_xlog_lock(), heap_xlog_lock_updated(), heap_xlog_multi_insert(), heap_xlog_prune_freeze(), heap_xlog_update(), heap_xlog_visible(), lazy_scan_new_or_empty(), lazy_scan_prune(), lazy_vacuum_heap_page(), log_newpage_range(), MarkDirtyUnpinnedBufferInternal(), moveLeafs(), nextval_internal(), RelationAddBlocks(), RelationCopyStorageUsingBuffer(), RelationGetBufferForTuple(), revmap_physical_extend(), saveNodeLink(), seq_redo(), SetSequence(), shiftList(), spgAddNodeAction(), spgbuild(), SpGistUpdateMetaPage(), spgRedoAddLeaf(), spgRedoAddNode(), spgRedoMoveLeafs(), spgRedoPickSplit(), spgRedoSplitTuple(), spgRedoVacuumLeaf(), spgRedoVacuumRedirect(), spgRedoVacuumRoot(), spgSplitNodeAction(), vacuumLeafPage(), vacuumLeafRoot(), vacuumRedirectAndPlaceholder(), visibilitymap_clear(), visibilitymap_prepare_truncate(), visibilitymap_set(), visibilitymap_set_vmbits(), writeListPage(), and XLogReadBufferForRedoExtended().

◆ MarkBufferDirtyHint()

void MarkBufferDirtyHint	(	Buffer	buffer,
		bool	buffer_std
	)

Definition at line 5430 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    Page        page = BufferGetPage(buffer);
 
    if (!BufferIsValid(buffer))
        elog(ERROR, "bad buffer ID: %d", buffer);
 
    if (BufferIsLocal(buffer))
    {
        MarkLocalBufferDirty(buffer);
        return;
    }
 
    bufHdr = GetBufferDescriptor(buffer - 1);
 
    Assert(GetPrivateRefCount(buffer) > 0);
    /* here, either share or exclusive lock is OK */
    Assert(BufferIsLockedByMe(buffer));
 
    /*
     * This routine might get called many times on the same page, if we are
     * making the first scan after commit of an xact that added/deleted many
     * tuples. So, be as quick as we can if the buffer is already dirty.  We
     * do this by not acquiring spinlock if it looks like the status bits are
     * already set.  Since we make this test unlocked, there's a chance we
     * might fail to notice that the flags have just been cleared, and failed
     * to reset them, due to memory-ordering issues.  But since this function
     * is only intended to be used in cases where failing to write out the
     * data would be harmless anyway, it doesn't really matter.
     */
    if ((pg_atomic_read_u32(&bufHdr->state) & (BM_DIRTY | BM_JUST_DIRTIED)) !=
        (BM_DIRTY | BM_JUST_DIRTIED))
    {
        XLogRecPtr  lsn = InvalidXLogRecPtr;
        bool        dirtied = false;
        bool        delayChkptFlags = false;
        uint32      buf_state;
 
        /*
         * If we need to protect hint bit updates from torn writes, WAL-log a
         * full page image of the page. This full page image is only necessary
         * if the hint bit update is the first change to the page since the
         * last checkpoint.
         *
         * We don't check full_page_writes here because that logic is included
         * when we call XLogInsert() since the value changes dynamically.
         */
        if (XLogHintBitIsNeeded() &&
            (pg_atomic_read_u32(&bufHdr->state) & BM_PERMANENT))
        {
            /*
             * If we must not write WAL, due to a relfilelocator-specific
             * condition or being in recovery, don't dirty the page.  We can
             * set the hint, just not dirty the page as a result so the hint
             * is lost when we evict the page or shutdown.
             *
             * See src/backend/storage/page/README for longer discussion.
             */
            if (RecoveryInProgress() ||
                RelFileLocatorSkippingWAL(BufTagGetRelFileLocator(&bufHdr->tag)))
                return;
 
            /*
             * If the block is already dirty because we either made a change
             * or set a hint already, then we don't need to write a full page
             * image.  Note that aggressive cleaning of blocks dirtied by hint
             * bit setting would increase the call rate. Bulk setting of hint
             * bits would reduce the call rate...
             *
             * We must issue the WAL record before we mark the buffer dirty.
             * Otherwise we might write the page before we write the WAL. That
             * causes a race condition, since a checkpoint might occur between
             * writing the WAL record and marking the buffer dirty. We solve
             * that with a kluge, but one that is already in use during
             * transaction commit to prevent race conditions. Basically, we
             * simply prevent the checkpoint WAL record from being written
             * until we have marked the buffer dirty. We don't start the
             * checkpoint flush until we have marked dirty, so our checkpoint
             * must flush the change to disk successfully or the checkpoint
             * never gets written, so crash recovery will fix.
             *
             * It's possible we may enter here without an xid, so it is
             * essential that CreateCheckPoint waits for virtual transactions
             * rather than full transactionids.
             */
            Assert((MyProc->delayChkptFlags & DELAY_CHKPT_START) == 0);
            MyProc->delayChkptFlags |= DELAY_CHKPT_START;
            delayChkptFlags = true;
            lsn = XLogSaveBufferForHint(buffer, buffer_std);
        }
 
        buf_state = LockBufHdr(bufHdr);
 
        Assert(BUF_STATE_GET_REFCOUNT(buf_state) > 0);
 
        if (!(buf_state & BM_DIRTY))
        {
            dirtied = true;     /* Means "will be dirtied by this action" */
 
            /*
             * Set the page LSN if we wrote a backup block. We aren't supposed
             * to set this when only holding a share lock but as long as we
             * serialise it somehow we're OK. We choose to set LSN while
             * holding the buffer header lock, which causes any reader of an
             * LSN who holds only a share lock to also obtain a buffer header
             * lock before using PageGetLSN(), which is enforced in
             * BufferGetLSNAtomic().
             *
             * If checksums are enabled, you might think we should reset the
             * checksum here. That will happen when the page is written
             * sometime later in this checkpoint cycle.
             */
            if (XLogRecPtrIsValid(lsn))
                PageSetLSN(page, lsn);
        }
 
        UnlockBufHdrExt(bufHdr, buf_state,
                        BM_DIRTY | BM_JUST_DIRTIED,
                        0, 0);
 
        if (delayChkptFlags)
            MyProc->delayChkptFlags &= ~DELAY_CHKPT_START;
 
        if (dirtied)
        {
            pgBufferUsage.shared_blks_dirtied++;
            if (VacuumCostActive)
                VacuumCostBalance += VacuumCostPageDirty;
        }
    }
}

References Assert(), BM_DIRTY, BM_JUST_DIRTIED, BM_PERMANENT, BUF_STATE_GET_REFCOUNT, PrivateRefCountEntry::buffer, BufferGetPage(), BufferIsLocal, BufferIsLockedByMe(), BufferIsValid(), BufTagGetRelFileLocator(), DELAY_CHKPT_START, PGPROC::delayChkptFlags, elog, ERROR, GetBufferDescriptor(), GetPrivateRefCount(), InvalidXLogRecPtr, LockBufHdr(), MarkLocalBufferDirty(), MyProc, PageSetLSN(), pg_atomic_read_u32(), pgBufferUsage, RecoveryInProgress(), RelFileLocatorSkippingWAL(), BufferUsage::shared_blks_dirtied, BufferDesc::state, BufferDesc::tag, UnlockBufHdrExt(), VacuumCostActive, VacuumCostBalance, VacuumCostPageDirty, XLogHintBitIsNeeded, XLogRecPtrIsValid, and XLogSaveBufferForHint().

Referenced by _bt_check_unique(), _bt_killitems(), _hash_kill_items(), brin_start_evacuating_page(), btvacuumpage(), fsm_search(), fsm_search_avail(), fsm_set_and_search(), fsm_vacuum_page(), gistkillitems(), heap_page_prune_and_freeze(), read_seq_tuple(), SetHintBits(), and XLogRecordPageWithFreeSpace().

◆ MarkDirtyAllUnpinnedBuffers()

void MarkDirtyAllUnpinnedBuffers	(	int32 *	buffers_dirtied,
		int32 *	buffers_already_dirty,
		int32 *	buffers_skipped
	)

Definition at line 6933 of file bufmgr.c.

{
    *buffers_dirtied = 0;
    *buffers_already_dirty = 0;
    *buffers_skipped = 0;
 
    for (int buf = 1; buf <= NBuffers; buf++)
    {
        BufferDesc *desc = GetBufferDescriptor(buf - 1);
        uint32      buf_state;
        bool        buffer_already_dirty;
 
        CHECK_FOR_INTERRUPTS();
 
        buf_state = pg_atomic_read_u32(&desc->state);
        if (!(buf_state & BM_VALID))
            continue;
 
        ResourceOwnerEnlarge(CurrentResourceOwner);
        ReservePrivateRefCountEntry();
 
        LockBufHdr(desc);
 
        if (MarkDirtyUnpinnedBufferInternal(buf, desc, &buffer_already_dirty))
            (*buffers_dirtied)++;
        else if (buffer_already_dirty)
            (*buffers_already_dirty)++;
        else
            (*buffers_skipped)++;
    }
}

References BM_VALID, buf, CHECK_FOR_INTERRUPTS, CurrentResourceOwner, GetBufferDescriptor(), LockBufHdr(), MarkDirtyUnpinnedBufferInternal(), NBuffers, pg_atomic_read_u32(), ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), and BufferDesc::state.

Referenced by pg_buffercache_mark_dirty_all().

◆ MarkDirtyRelUnpinnedBuffers()

void MarkDirtyRelUnpinnedBuffers	(	Relation	rel,
		int32 *	buffers_dirtied,
		int32 *	buffers_already_dirty,
		int32 *	buffers_skipped
	)

Definition at line 6876 of file bufmgr.c.

{
    Assert(!RelationUsesLocalBuffers(rel));
 
    *buffers_dirtied = 0;
    *buffers_already_dirty = 0;
    *buffers_skipped = 0;
 
    for (int buf = 1; buf <= NBuffers; buf++)
    {
        BufferDesc *desc = GetBufferDescriptor(buf - 1);
        uint32      buf_state = pg_atomic_read_u32(&(desc->state));
        bool        buffer_already_dirty;
 
        CHECK_FOR_INTERRUPTS();
 
        /* An unlocked precheck should be safe and saves some cycles. */
        if ((buf_state & BM_VALID) == 0 ||
            !BufTagMatchesRelFileLocator(&desc->tag, &rel->rd_locator))
            continue;
 
        /* Make sure we can pin the buffer. */
        ResourceOwnerEnlarge(CurrentResourceOwner);
        ReservePrivateRefCountEntry();
 
        buf_state = LockBufHdr(desc);
 
        /* recheck, could have changed without the lock */
        if ((buf_state & BM_VALID) == 0 ||
            !BufTagMatchesRelFileLocator(&desc->tag, &rel->rd_locator))
        {
            UnlockBufHdr(desc);
            continue;
        }
 
        if (MarkDirtyUnpinnedBufferInternal(buf, desc, &buffer_already_dirty))
            (*buffers_dirtied)++;
        else if (buffer_already_dirty)
            (*buffers_already_dirty)++;
        else
            (*buffers_skipped)++;
    }
}

References Assert(), BM_VALID, buf, BufTagMatchesRelFileLocator(), CHECK_FOR_INTERRUPTS, CurrentResourceOwner, GetBufferDescriptor(), LockBufHdr(), MarkDirtyUnpinnedBufferInternal(), NBuffers, pg_atomic_read_u32(), RelationData::rd_locator, RelationUsesLocalBuffers, ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), BufferDesc::state, BufferDesc::tag, and UnlockBufHdr().

Referenced by pg_buffercache_mark_dirty_relation().

◆ MarkDirtyUnpinnedBuffer()

bool MarkDirtyUnpinnedBuffer	(	Buffer	buf,
		bool *	buffer_already_dirty
	)

Definition at line 6840 of file bufmgr.c.

{
    BufferDesc *desc;
    bool        buffer_dirtied = false;
 
    Assert(!BufferIsLocal(buf));
 
    /* Make sure we can pin the buffer. */
    ResourceOwnerEnlarge(CurrentResourceOwner);
    ReservePrivateRefCountEntry();
 
    desc = GetBufferDescriptor(buf - 1);
    LockBufHdr(desc);
 
    buffer_dirtied = MarkDirtyUnpinnedBufferInternal(buf, desc, buffer_already_dirty);
    /* Both can not be true at the same time */
    Assert(!(buffer_dirtied && *buffer_already_dirty));
 
    return buffer_dirtied;
}

References Assert(), buf, BufferIsLocal, CurrentResourceOwner, GetBufferDescriptor(), LockBufHdr(), MarkDirtyUnpinnedBufferInternal(), ReservePrivateRefCountEntry(), and ResourceOwnerEnlarge().

Referenced by pg_buffercache_mark_dirty().

◆ PrefetchBuffer()

PrefetchBufferResult PrefetchBuffer	(	Relation	reln,
		ForkNumber	forkNum,
		BlockNumber	blockNum
	)

Definition at line 653 of file bufmgr.c.

{
    Assert(RelationIsValid(reln));
    Assert(BlockNumberIsValid(blockNum));
 
    if (RelationUsesLocalBuffers(reln))
    {
        /* see comments in ReadBufferExtended */
        if (RELATION_IS_OTHER_TEMP(reln))
            ereport(ERROR,
                    (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                     errmsg("cannot access temporary tables of other sessions")));
 
        /* pass it off to localbuf.c */
        return PrefetchLocalBuffer(RelationGetSmgr(reln), forkNum, blockNum);
    }
    else
    {
        /* pass it to the shared buffer version */
        return PrefetchSharedBuffer(RelationGetSmgr(reln), forkNum, blockNum);
    }
}

References Assert(), BlockNumberIsValid(), ereport, errcode(), errmsg(), ERROR, PrefetchLocalBuffer(), PrefetchSharedBuffer(), RELATION_IS_OTHER_TEMP, RelationGetSmgr(), RelationIsValid, and RelationUsesLocalBuffers.

Referenced by count_nondeletable_pages(), invalidate_rel_block(), and pg_prewarm().

◆ PrefetchSharedBuffer()

PrefetchBufferResult PrefetchSharedBuffer	(	SMgrRelation	smgr_reln,
		ForkNumber	forkNum,
		BlockNumber	blockNum
	)

Definition at line 563 of file bufmgr.c.

{
    PrefetchBufferResult result = {InvalidBuffer, false};
    BufferTag   newTag;         /* identity of requested block */
    uint32      newHash;        /* hash value for newTag */
    LWLock     *newPartitionLock;   /* buffer partition lock for it */
    int         buf_id;
 
    Assert(BlockNumberIsValid(blockNum));
 
    /* create a tag so we can lookup the buffer */
    InitBufferTag(&newTag, &smgr_reln->smgr_rlocator.locator,
                  forkNum, blockNum);
 
    /* determine its hash code and partition lock ID */
    newHash = BufTableHashCode(&newTag);
    newPartitionLock = BufMappingPartitionLock(newHash);
 
    /* see if the block is in the buffer pool already */
    LWLockAcquire(newPartitionLock, LW_SHARED);
    buf_id = BufTableLookup(&newTag, newHash);
    LWLockRelease(newPartitionLock);
 
    /* If not in buffers, initiate prefetch */
    if (buf_id < 0)
    {
#ifdef USE_PREFETCH
        /*
         * Try to initiate an asynchronous read.  This returns false in
         * recovery if the relation file doesn't exist.
         */
        if ((io_direct_flags & IO_DIRECT_DATA) == 0 &&
            smgrprefetch(smgr_reln, forkNum, blockNum, 1))
        {
            result.initiated_io = true;
        }
#endif                          /* USE_PREFETCH */
    }
    else
    {
        /*
         * Report the buffer it was in at that time.  The caller may be able
         * to avoid a buffer table lookup, but it's not pinned and it must be
         * rechecked!
         */
        result.recent_buffer = buf_id + 1;
    }
 
    /*
     * If the block *is* in buffers, we do nothing.  This is not really ideal:
     * the block might be just about to be evicted, which would be stupid
     * since we know we are going to need it soon.  But the only easy answer
     * is to bump the usage_count, which does not seem like a great solution:
     * when the caller does ultimately touch the block, usage_count would get
     * bumped again, resulting in too much favoritism for blocks that are
     * involved in a prefetch sequence. A real fix would involve some
     * additional per-buffer state, and it's not clear that there's enough of
     * a problem to justify that.
     */
 
    return result;
}

References Assert(), BlockNumberIsValid(), BufMappingPartitionLock(), BufTableHashCode(), BufTableLookup(), InitBufferTag(), PrefetchBufferResult::initiated_io, InvalidBuffer, IO_DIRECT_DATA, io_direct_flags, RelFileLocatorBackend::locator, LW_SHARED, LWLockAcquire(), LWLockRelease(), PrefetchBufferResult::recent_buffer, SMgrRelationData::smgr_rlocator, and smgrprefetch().

Referenced by PrefetchBuffer(), and XLogPrefetcherNextBlock().

◆ ReadBuffer()

Buffer ReadBuffer	(	Relation	reln,
		BlockNumber	blockNum
	)

Definition at line 745 of file bufmgr.c.

{
    return ReadBufferExtended(reln, MAIN_FORKNUM, blockNum, RBM_NORMAL, NULL);
}

References MAIN_FORKNUM, RBM_NORMAL, and ReadBufferExtended().

Referenced by _bt_allocbuf(), _bt_getbuf(), _bt_search_insert(), _hash_getbuf(), _hash_getbuf_with_condlock_cleanup(), blbulkdelete(), blinsert(), BloomNewBuffer(), brin_getinsertbuffer(), brinGetStats(), brinGetTupleForHeapBlock(), brinRevmapDesummarizeRange(), brinRevmapInitialize(), bt_metap(), bt_multi_page_stats(), bt_page_items_internal(), bt_page_stats_internal(), ginFindLeafPage(), ginFindParents(), ginGetStats(), ginHeapTupleFastInsert(), ginInsertCleanup(), GinNewBuffer(), ginStepRight(), ginUpdateStats(), gistBufferingFindCorrectParent(), gistbufferinginserttuples(), gistdoinsert(), gistFindCorrectParent(), gistFindPath(), gistfixsplit(), gistGetMaxLevel(), gistkillitems(), gistNewBuffer(), gistProcessItup(), gistScanPage(), heap_abort_speculative(), heap_delete(), heap_fetch(), heap_finish_speculative(), heap_force_common(), heap_get_latest_tid(), heap_index_delete_tuples(), heap_lock_tuple(), heap_update(), initBloomState(), pg_visibility(), pgstatginindex_internal(), read_seq_tuple(), RelationGetBufferForTuple(), ReleaseAndReadBuffer(), revmap_get_buffer(), revmap_physical_extend(), scanGetCandidate(), scanPendingInsert(), shiftList(), spgdoinsert(), spgGetCache(), SpGistGetBuffer(), SpGistNewBuffer(), SpGistUpdateMetaPage(), and spgWalk().

◆ ReadBufferExtended()

Buffer ReadBufferExtended	(	Relation	reln,
		ForkNumber	forkNum,
		BlockNumber	blockNum,
		ReadBufferMode	mode,
		BufferAccessStrategy	strategy
	)

inline

Definition at line 792 of file bufmgr.c.

{
    Buffer      buf;
 
    /*
     * Reject attempts to read non-local temporary relations; we would be
     * likely to get wrong data since we have no visibility into the owning
     * session's local buffers.
     */
    if (RELATION_IS_OTHER_TEMP(reln))
        ereport(ERROR,
                (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                 errmsg("cannot access temporary tables of other sessions")));
 
    /*
     * Read the buffer, and update pgstat counters to reflect a cache hit or
     * miss.
     */
    buf = ReadBuffer_common(reln, RelationGetSmgr(reln), 0,
                            forkNum, blockNum, mode, strategy);
 
    return buf;
}

References buf, ereport, errcode(), errmsg(), ERROR, mode, ReadBuffer_common(), RELATION_IS_OTHER_TEMP, and RelationGetSmgr().

Referenced by _hash_getbuf_with_strategy(), _hash_getinitbuf(), _hash_getnewbuf(), blbulkdelete(), blgetbitmap(), BloomInitMetapage(), blvacuumcleanup(), bt_recheck_sibling_links(), btvacuumpage(), count_nondeletable_pages(), create_toy_buffer(), fsm_readbuf(), get_raw_page_internal(), gin_check_parent_keys_consistency(), gin_check_posting_tree_parent_keys_consistency(), gin_refind_parent(), ginbulkdelete(), ginDeletePage(), ginScanToDelete(), ginvacuumcleanup(), ginVacuumPostingTree(), ginVacuumPostingTreeLeaves(), gistvacuum_delete_empty_pages(), gistvacuumpage(), hashbulkdelete(), heapam_scan_sample_next_block(), log_newpage_range(), modify_rel_block(), palloc_btree_page(), pgstat_btree_page(), pgstat_gist_page(), pgstat_hash_page(), pgstat_heap(), pgstathashindex(), pgstatindex_impl(), ReadBuffer(), ReadBufferBI(), spgprocesspending(), statapprox_heap(), and vm_readbuf().

◆ ReadBufferWithoutRelcache()

Buffer ReadBufferWithoutRelcache	(	RelFileLocator	rlocator,
		ForkNumber	forkNum,
		BlockNumber	blockNum,
		ReadBufferMode	mode,
		BufferAccessStrategy	strategy,
		bool	permanent
	)

Definition at line 829 of file bufmgr.c.

{
    SMgrRelation smgr = smgropen(rlocator, INVALID_PROC_NUMBER);
 
    return ReadBuffer_common(NULL, smgr,
                             permanent ? RELPERSISTENCE_PERMANENT : RELPERSISTENCE_UNLOGGED,
                             forkNum, blockNum,
                             mode, strategy);
}

References INVALID_PROC_NUMBER, mode, ReadBuffer_common(), and smgropen().

Referenced by RelationCopyStorageUsingBuffer(), ScanSourceDatabasePgClass(), and XLogReadBufferExtended().

◆ ReadRecentBuffer()

bool ReadRecentBuffer	(	RelFileLocator	rlocator,
		ForkNumber	forkNum,
		BlockNumber	blockNum,
		Buffer	recent_buffer
	)

Definition at line 684 of file bufmgr.c.

{
    BufferDesc *bufHdr;
    BufferTag   tag;
    uint32      buf_state;
 
    Assert(BufferIsValid(recent_buffer));
 
    ResourceOwnerEnlarge(CurrentResourceOwner);
    ReservePrivateRefCountEntry();
    InitBufferTag(&tag, &rlocator, forkNum, blockNum);
 
    if (BufferIsLocal(recent_buffer))
    {
        int         b = -recent_buffer - 1;
 
        bufHdr = GetLocalBufferDescriptor(b);
        buf_state = pg_atomic_read_u32(&bufHdr->state);
 
        /* Is it still valid and holding the right tag? */
        if ((buf_state & BM_VALID) && BufferTagsEqual(&tag, &bufHdr->tag))
        {
            PinLocalBuffer(bufHdr, true);
 
            pgBufferUsage.local_blks_hit++;
 
            return true;
        }
    }
    else
    {
        bufHdr = GetBufferDescriptor(recent_buffer - 1);
 
        /*
         * Is it still valid and holding the right tag?  We do an unlocked tag
         * comparison first, to make it unlikely that we'll increment the
         * usage counter of the wrong buffer, if someone calls us with a very
         * out of date recent_buffer.  Then we'll check it again if we get the
         * pin.
         */
        if (BufferTagsEqual(&tag, &bufHdr->tag) &&
            PinBuffer(bufHdr, NULL, true))
        {
            if (BufferTagsEqual(&tag, &bufHdr->tag))
            {
                pgBufferUsage.shared_blks_hit++;
                return true;
            }
            UnpinBuffer(bufHdr);
        }
    }
 
    return false;
}

References Assert(), b, BM_VALID, BufferIsLocal, BufferIsValid(), BufferTagsEqual(), CurrentResourceOwner, GetBufferDescriptor(), GetLocalBufferDescriptor(), InitBufferTag(), BufferUsage::local_blks_hit, pg_atomic_read_u32(), pgBufferUsage, PinBuffer(), PinLocalBuffer(), ReservePrivateRefCountEntry(), ResourceOwnerEnlarge(), BufferUsage::shared_blks_hit, BufferDesc::state, BufferDesc::tag, and UnpinBuffer().

Referenced by invalidate_rel_block(), and XLogReadBufferExtended().

◆ RelationGetNumberOfBlocksInFork()

BlockNumber RelationGetNumberOfBlocksInFork	(	Relation	relation,
		ForkNumber	forkNum
	)

Definition at line 4437 of file bufmgr.c.

{
    if (RELKIND_HAS_TABLE_AM(relation->rd_rel->relkind))
    {
        /*
         * Not every table AM uses BLCKSZ wide fixed size blocks. Therefore
         * tableam returns the size in bytes - but for the purpose of this
         * routine, we want the number of blocks. Therefore divide, rounding
         * up.
         */
        uint64      szbytes;
 
        szbytes = table_relation_size(relation, forkNum);
 
        return (szbytes + (BLCKSZ - 1)) / BLCKSZ;
    }
    else if (RELKIND_HAS_STORAGE(relation->rd_rel->relkind))
    {
        return smgrnblocks(RelationGetSmgr(relation), forkNum);
    }
    else
        Assert(false);
 
    return 0;                   /* keep compiler quiet */
}

References Assert(), RelationData::rd_rel, RelationGetSmgr(), smgrnblocks(), and table_relation_size().

Referenced by _hash_getnewbuf(), _hash_init(), autoprewarm_database_main(), get_raw_page_internal(), and pg_prewarm().

◆ ReleaseAndReadBuffer()

Buffer ReleaseAndReadBuffer	(	Buffer	buffer,
		Relation	relation,
		BlockNumber	blockNum
	)

Definition at line 3008 of file bufmgr.c.

{
    ForkNumber  forkNum = MAIN_FORKNUM;
    BufferDesc *bufHdr;
 
    if (BufferIsValid(buffer))
    {
        Assert(BufferIsPinned(buffer));
        if (BufferIsLocal(buffer))
        {
            bufHdr = GetLocalBufferDescriptor(-buffer - 1);
            if (bufHdr->tag.blockNum == blockNum &&
                BufTagMatchesRelFileLocator(&bufHdr->tag, &relation->rd_locator) &&
                BufTagGetForkNum(&bufHdr->tag) == forkNum)
                return buffer;
            UnpinLocalBuffer(buffer);
        }
        else
        {
            bufHdr = GetBufferDescriptor(buffer - 1);
            /* we have pin, so it's ok to examine tag without spinlock */
            if (bufHdr->tag.blockNum == blockNum &&
                BufTagMatchesRelFileLocator(&bufHdr->tag, &relation->rd_locator) &&
                BufTagGetForkNum(&bufHdr->tag) == forkNum)
                return buffer;
            UnpinBuffer(bufHdr);
        }
    }
 
    return ReadBuffer(relation, blockNum);
}

References Assert(), buftag::blockNum, PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsPinned, BufferIsValid(), BufTagGetForkNum(), BufTagMatchesRelFileLocator(), GetBufferDescriptor(), GetLocalBufferDescriptor(), MAIN_FORKNUM, RelationData::rd_locator, ReadBuffer(), BufferDesc::tag, UnpinBuffer(), and UnpinLocalBuffer().

Referenced by _bt_relandgetbuf(), ginFindLeafPage(), and heapam_index_fetch_tuple().

◆ ReleaseBuffer()

void ReleaseBuffer ( Buffer buffer )

Definition at line 5366 of file bufmgr.c.

{
    if (!BufferIsValid(buffer))
        elog(ERROR, "bad buffer ID: %d", buffer);
 
    if (BufferIsLocal(buffer))
        UnpinLocalBuffer(buffer);
    else
        UnpinBuffer(GetBufferDescriptor(buffer - 1));
}

References PrivateRefCountEntry::buffer, BufferIsLocal, BufferIsValid(), elog, ERROR, GetBufferDescriptor(), UnpinBuffer(), and UnpinLocalBuffer().

Referenced by _bt_allocbuf(), _bt_pagedel(), _bt_relbuf(), _bt_search_insert(), _bt_unlink_halfdead_page(), _hash_dropbuf(), _hash_getbuf_with_condlock_cleanup(), autoprewarm_database_main(), BitmapHeapScanNextBlock(), blinsert(), BloomNewBuffer(), brin_getinsertbuffer(), brin_vacuum_scan(), bringetbitmap(), brinGetTupleForHeapBlock(), brininsert(), brinRevmapTerminate(), brinsummarize(), buffer_create_toy(), collect_corrupt_items(), collect_visibility_data(), entryLoadMoreItems(), ExecEndIndexOnlyScan(), ExtendBufferedRelTo(), FreeBulkInsertState(), freeGinBtreeStack(), fsm_search(), fsm_vacuum_page(), get_actual_variable_endpoint(), get_raw_page_internal(), GetRecordedFreeSpace(), gin_check_parent_keys_consistency(), gin_check_posting_tree_parent_keys_consistency(), ginDeletePage(), ginFindParents(), ginFinishSplit(), ginFreeScanKeys(), ginInsertCleanup(), GinNewBuffer(), ginScanToDelete(), gistdoinsert(), gistFindCorrectParent(), gistNewBuffer(), gistvacuum_delete_empty_pages(), grow_rel(), heap_abort_speculative(), heap_delete(), heap_endscan(), heap_fetch(), heap_fetch_next_buffer(), heap_force_common(), heap_insert(), heap_lock_tuple(), heap_lock_updated_tuple_rec(), heap_multi_insert(), heap_rescan(), heap_update(), heap_vac_scan_next_block(), heap_xlog_delete(), heap_xlog_insert(), heap_xlog_lock(), heap_xlog_lock_updated(), heap_xlog_multi_insert(), heap_xlog_update(), heap_xlog_visible(), heapam_index_fetch_reset(), heapam_scan_sample_next_block(), heapam_tuple_lock(), heapgettup(), heapgettup_pagemode(), invalidate_rel_block(), lazy_scan_heap(), lazy_vacuum_heap_rel(), modify_rel_block(), pg_prewarm(), pg_visibility(), pg_visibility_map(), pg_visibility_map_summary(), pgstatindex_impl(), read_rel_block_ll(), read_stream_reset(), ReadBufferBI(), RelationAddBlocks(), RelationGetBufferForTuple(), ReleaseBulkInsertStatePin(), revmap_get_buffer(), spgdoinsert(), SpGistGetBuffer(), SpGistNewBuffer(), SpGistUpdateMetaPage(), statapprox_heap(), summarize_range(), terminate_brin_buildstate(), tts_buffer_heap_clear(), tts_buffer_heap_materialize(), tts_buffer_heap_store_tuple(), UnlockReleaseBuffer(), verify_heapam(), visibilitymap_count(), visibilitymap_get_status(), visibilitymap_pin(), and XLogReadBufferExtended().

◆ StartReadBuffer()

bool StartReadBuffer	(	ReadBuffersOperation *	operation,
		Buffer *	buffer,
		BlockNumber	blocknum,
		int	flags
	)

Definition at line 1489 of file bufmgr.c.

{
    int         nblocks = 1;
    bool        result;
 
    result = StartReadBuffersImpl(operation, buffer, blocknum, &nblocks, flags,
                                  false /* single block, no forwarding */ );
    Assert(nblocks == 1);       /* single block can't be short */
 
    return result;
}

References Assert(), PrivateRefCountEntry::buffer, and StartReadBuffersImpl().

Referenced by read_stream_next_buffer(), and ReadBuffer_common().

◆ StartReadBuffers()

bool StartReadBuffers	(	ReadBuffersOperation *	operation,
		Buffer *	buffers,
		BlockNumber	blockNum,
		int *	nblocks,
		int	flags
	)

Definition at line 1470 of file bufmgr.c.

{
    return StartReadBuffersImpl(operation, buffers, blockNum, nblocks, flags,
                                true /* expect forwarded buffers */ );
}

References StartReadBuffersImpl().

Referenced by read_stream_start_pending_read().

◆ UnlockBuffers()

void UnlockBuffers ( void )

Definition at line 5573 of file bufmgr.c.

{
    BufferDesc *buf = PinCountWaitBuf;
 
    if (buf)
    {
        uint32      buf_state;
        uint32      unset_bits = 0;
 
        buf_state = LockBufHdr(buf);
 
        /*
         * Don't complain if flag bit not set; it could have been reset but we
         * got a cancel/die interrupt before getting the signal.
         */
        if ((buf_state & BM_PIN_COUNT_WAITER) != 0 &&
            buf->wait_backend_pgprocno == MyProcNumber)
            unset_bits = BM_PIN_COUNT_WAITER;
 
        UnlockBufHdrExt(buf, buf_state,
                        0, unset_bits,
                        0);
 
        PinCountWaitBuf = NULL;
    }
}

References BM_PIN_COUNT_WAITER, buf, LockBufHdr(), MyProcNumber, PinCountWaitBuf, and UnlockBufHdrExt().

Referenced by AbortSubTransaction(), AbortTransaction(), AtProcExit_Buffers(), AutoVacLauncherMain(), BackgroundWriterMain(), CheckpointerMain(), and WalWriterMain().

◆ UnlockReleaseBuffer()

void UnlockReleaseBuffer ( Buffer buffer )

Definition at line 5383 of file bufmgr.c.

{
    LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
    ReleaseBuffer(buffer);
}

References PrivateRefCountEntry::buffer, BUFFER_LOCK_UNLOCK, LockBuffer(), and ReleaseBuffer().

Referenced by _bt_clear_incomplete_split(), _bt_restore_meta(), _hash_relbuf(), allocNewBuffer(), AlterSequence(), blbulkdelete(), blgetbitmap(), blinsert(), BloomInitMetapage(), blvacuumcleanup(), brin_doinsert(), brin_doupdate(), brin_evacuate_page(), brin_getinsertbuffer(), brin_xlog_createidx(), brin_xlog_desummarize_page(), brin_xlog_insert_update(), brin_xlog_revmap_extend(), brin_xlog_samepage_update(), brin_xlog_update(), brinbuild(), brinbuildempty(), brinGetStats(), brinRevmapDesummarizeRange(), bt_metap(), bt_multi_page_stats(), bt_page_items_internal(), bt_page_stats_internal(), bt_recheck_sibling_links(), btree_xlog_dedup(), btree_xlog_delete(), btree_xlog_insert(), btree_xlog_mark_page_halfdead(), btree_xlog_newroot(), btree_xlog_split(), btree_xlog_unlink_page(), btree_xlog_vacuum(), collect_corrupt_items(), collect_visibility_data(), count_nondeletable_pages(), createPostingTree(), doPickSplit(), entryLoadMoreItems(), fill_seq_fork_with_data(), flushCachedPage(), FreeSpaceMapPrepareTruncateRel(), fsm_search(), fsm_set_and_search(), generic_redo(), gin_refind_parent(), ginbuild(), ginbuildempty(), ginbulkdelete(), ginGetStats(), ginHeapTupleFastInsert(), ginInsertCleanup(), ginPlaceToPage(), ginRedoClearIncompleteSplit(), ginRedoCreatePTree(), ginRedoDeleteListPages(), ginRedoDeletePage(), ginRedoInsert(), ginRedoInsertListPage(), ginRedoSplit(), ginRedoUpdateMetapage(), ginRedoVacuumDataLeafPage(), ginRedoVacuumPage(), ginScanToDelete(), ginStepRight(), ginUpdateStats(), ginvacuumcleanup(), ginVacuumPostingTree(), ginVacuumPostingTreeLeaves(), gistbufferinginserttuples(), gistbuild(), gistbuildempty(), gistdoinsert(), gistFindCorrectParent(), gistFindPath(), gistGetMaxLevel(), gistinserttuples(), gistkillitems(), gistplacetopage(), gistProcessItup(), gistRedoClearFollowRight(), gistRedoDeleteRecord(), gistRedoPageDelete(), gistRedoPageSplitRecord(), gistRedoPageUpdateRecord(), gistScanPage(), gistvacuum_delete_empty_pages(), gistvacuumpage(), hash_xlog_add_ovfl_page(), hash_xlog_delete(), hash_xlog_init_bitmap_page(), hash_xlog_init_meta_page(), hash_xlog_insert(), hash_xlog_move_page_contents(), hash_xlog_split_allocate_page(), hash_xlog_split_cleanup(), hash_xlog_split_complete(), hash_xlog_split_page(), hash_xlog_squeeze_page(), hash_xlog_update_meta_page(), hash_xlog_vacuum_one_page(), heap_delete(), heap_finish_speculative(), heap_force_common(), heap_get_latest_tid(), heap_index_delete_tuples(), heap_insert(), heap_lock_updated_tuple_rec(), heap_multi_insert(), heap_update(), heap_xlog_confirm(), heap_xlog_delete(), heap_xlog_inplace(), heap_xlog_insert(), heap_xlog_lock(), heap_xlog_lock_updated(), heap_xlog_multi_insert(), heap_xlog_prune_freeze(), heap_xlog_update(), heap_xlog_visible(), heapam_scan_analyze_next_tuple(), initBloomState(), lazy_scan_heap(), lazy_scan_new_or_empty(), lazy_vacuum_heap_rel(), log_newpage_range(), moveLeafs(), nextval_internal(), palloc_btree_page(), pg_get_sequence_data(), pg_sequence_last_value(), pg_visibility(), pgstat_gist_page(), pgstat_heap(), pgstatginindex_internal(), pgstathashindex(), RelationCopyStorageUsingBuffer(), RelationGetBufferForTuple(), ResetSequence(), revmap_physical_extend(), scanGetCandidate(), scanPendingInsert(), scanPostingTree(), ScanSourceDatabasePgClass(), seq_redo(), SequenceChangePersistence(), SetSequence(), shiftList(), spgAddNodeAction(), spgbuild(), spgdoinsert(), spgGetCache(), SpGistGetBuffer(), SpGistUpdateMetaPage(), spgMatchNodeAction(), spgprocesspending(), spgRedoAddLeaf(), spgRedoAddNode(), spgRedoMoveLeafs(), spgRedoPickSplit(), spgRedoSplitTuple(), spgRedoVacuumLeaf(), spgRedoVacuumRedirect(), spgRedoVacuumRoot(), spgSplitNodeAction(), spgvacuumpage(), spgWalk(), statapprox_heap(), verify_heapam(), verifyBackupPageConsistency(), visibilitymap_prepare_truncate(), writeListPage(), xlog_redo(), and XLogRecordPageWithFreeSpace().

◆ WaitReadBuffers()

void WaitReadBuffers ( ReadBuffersOperation * operation )

Definition at line 1613 of file bufmgr.c.

{
    PgAioReturn *aio_ret = &operation->io_return;
    IOContext   io_context;
    IOObject    io_object;
 
    if (operation->persistence == RELPERSISTENCE_TEMP)
    {
        io_context = IOCONTEXT_NORMAL;
        io_object = IOOBJECT_TEMP_RELATION;
    }
    else
    {
        io_context = IOContextForStrategy(operation->strategy);
        io_object = IOOBJECT_RELATION;
    }
 
    /*
     * If we get here without an IO operation having been issued, the
     * io_method == IOMETHOD_SYNC path must have been used. Otherwise the
     * caller should not have called WaitReadBuffers().
     *
     * In the case of IOMETHOD_SYNC, we start - as we used to before the
     * introducing of AIO - the IO in WaitReadBuffers(). This is done as part
     * of the retry logic below, no extra code is required.
     *
     * This path is expected to eventually go away.
     */
    if (!pgaio_wref_valid(&operation->io_wref) && io_method != IOMETHOD_SYNC)
        elog(ERROR, "waiting for read operation that didn't read");
 
    /*
     * To handle partial reads, and IOMETHOD_SYNC, we re-issue IO until we're
     * done. We may need multiple retries, not just because we could get
     * multiple partial reads, but also because some of the remaining
     * to-be-read buffers may have been read in by other backends, limiting
     * the IO size.
     */
    while (true)
    {
        int         ignored_nblocks_progress;
 
        CheckReadBuffersOperation(operation, false);
 
        /*
         * If there is an IO associated with the operation, we may need to
         * wait for it.
         */
        if (pgaio_wref_valid(&operation->io_wref))
        {
            /*
             * Track the time spent waiting for the IO to complete. As
             * tracking a wait even if we don't actually need to wait
             *
             * a) is not cheap, due to the timestamping overhead
             *
             * b) reports some time as waiting, even if we never waited
             *
             * we first check if we already know the IO is complete.
             */
            if (aio_ret->result.status == PGAIO_RS_UNKNOWN &&
                !pgaio_wref_check_done(&operation->io_wref))
            {
                instr_time  io_start = pgstat_prepare_io_time(track_io_timing);
 
                pgaio_wref_wait(&operation->io_wref);
 
                /*
                 * The IO operation itself was already counted earlier, in
                 * AsyncReadBuffers(), this just accounts for the wait time.
                 */
                pgstat_count_io_op_time(io_object, io_context, IOOP_READ,
                                        io_start, 0, 0);
            }
            else
            {
                Assert(pgaio_wref_check_done(&operation->io_wref));
            }
 
            /*
             * We now are sure the IO completed. Check the results. This
             * includes reporting on errors if there were any.
             */
            ProcessReadBuffersResult(operation);
        }
 
        /*
         * Most of the time, the one IO we already started, will read in
         * everything.  But we need to deal with partial reads and buffers not
         * needing IO anymore.
         */
        if (operation->nblocks_done == operation->nblocks)
            break;
 
        CHECK_FOR_INTERRUPTS();
 
        /*
         * This may only complete the IO partially, either because some
         * buffers were already valid, or because of a partial read.
         *
         * NB: In contrast to after the AsyncReadBuffers() call in
         * StartReadBuffers(), we do *not* reduce
         * ReadBuffersOperation->nblocks here, callers expect the full
         * operation to be completed at this point (as more operations may
         * have been queued).
         */
        AsyncReadBuffers(operation, &ignored_nblocks_progress);
    }
 
    CheckReadBuffersOperation(operation, true);
 
    /* NB: READ_DONE tracepoint was already executed in completion callback */
}

References Assert(), AsyncReadBuffers(), CHECK_FOR_INTERRUPTS, CheckReadBuffersOperation(), elog, ERROR, io_method, ReadBuffersOperation::io_return, ReadBuffersOperation::io_wref, IOCONTEXT_NORMAL, IOContextForStrategy(), IOMETHOD_SYNC, IOOBJECT_RELATION, IOOBJECT_TEMP_RELATION, IOOP_READ, ReadBuffersOperation::nblocks, ReadBuffersOperation::nblocks_done, ReadBuffersOperation::persistence, PGAIO_RS_UNKNOWN, pgaio_wref_check_done(), pgaio_wref_valid(), pgaio_wref_wait(), pgstat_count_io_op_time(), pgstat_prepare_io_time(), ProcessReadBuffersResult(), PgAioReturn::result, PgAioResult::status, ReadBuffersOperation::strategy, and track_io_timing.

Referenced by read_stream_next_buffer(), and ReadBuffer_common().

Variable Documentation

◆ aio_local_buffer_readv_cb

PGDLLIMPORT const PgAioHandleCallbacks aio_local_buffer_readv_cb

extern

Definition at line 7645 of file bufmgr.c.

◆ aio_shared_buffer_readv_cb

PGDLLIMPORT const PgAioHandleCallbacks aio_shared_buffer_readv_cb

extern

Definition at line 7636 of file bufmgr.c.

◆ backend_flush_after

PGDLLIMPORT int backend_flush_after

extern

Definition at line 180 of file bufmgr.c.

Referenced by BufferManagerShmemInit().

◆ bgwriter_flush_after

PGDLLIMPORT int bgwriter_flush_after

extern

Definition at line 179 of file bufmgr.c.

Referenced by BackgroundWriterMain().

◆ bgwriter_lru_maxpages

PGDLLIMPORT int bgwriter_lru_maxpages

extern

Definition at line 145 of file bufmgr.c.

Referenced by BgBufferSync().

◆ bgwriter_lru_multiplier

PGDLLIMPORT double bgwriter_lru_multiplier

extern

Definition at line 146 of file bufmgr.c.

Referenced by BgBufferSync().

◆ BufferBlocks

PGDLLIMPORT char* BufferBlocks

extern

Definition at line 22 of file buf_init.c.

Referenced by BufferGetBlock(), and BufferManagerShmemInit().

◆ checkpoint_flush_after

PGDLLIMPORT int checkpoint_flush_after

extern

Definition at line 178 of file bufmgr.c.

Referenced by BufferSync().

◆ effective_io_concurrency

PGDLLIMPORT int effective_io_concurrency

extern

Definition at line 155 of file bufmgr.c.

Referenced by get_tablespace_io_concurrency(), GetAccessStrategy(), read_stream_begin_impl(), and tablespace_reloptions().

◆ io_combine_limit

PGDLLIMPORT int io_combine_limit

extern

Definition at line 170 of file bufmgr.c.

Referenced by assign_io_combine_limit(), assign_io_max_combine_limit(), GetAccessStrategy(), read_stream_begin_impl(), read_stream_next_buffer(), and read_stream_reset().

◆ io_combine_limit_guc

PGDLLIMPORT int io_combine_limit_guc

extern

Definition at line 171 of file bufmgr.c.

Referenced by assign_io_max_combine_limit().

◆ io_max_combine_limit

PGDLLIMPORT int io_max_combine_limit

extern

Definition at line 172 of file bufmgr.c.

Referenced by AioHandleDataShmemSize(), AioHandleIOVShmemSize(), AioShmemInit(), assign_io_combine_limit(), pgaio_io_set_handle_data_32(), and pgaio_io_set_handle_data_64().

◆ LocalBufferBlockPointers

PGDLLIMPORT Block* LocalBufferBlockPointers

extern

Definition at line 48 of file localbuf.c.

Referenced by BufferGetBlock(), and InitLocalBuffers().

◆ LocalRefCount

PGDLLIMPORT int32* LocalRefCount

extern

Definition at line 49 of file localbuf.c.

Referenced by CheckBufferIsPinnedOnce(), CheckForLocalBufferLeaks(), ConditionalLockBufferForCleanup(), DebugPrintBufferRefcount(), FlushLocalBuffer(), GetLocalVictimBuffer(), IncrBufferRefCount(), InitLocalBuffers(), InvalidateLocalBuffer(), IsBufferCleanupOK(), MarkLocalBufferDirty(), PinLocalBuffer(), and UnpinLocalBufferNoOwner().

◆ maintenance_io_concurrency

PGDLLIMPORT int maintenance_io_concurrency

extern

Definition at line 162 of file bufmgr.c.

Referenced by assign_maintenance_io_concurrency(), get_tablespace_maintenance_io_concurrency(), heap_index_delete_tuples(), tablespace_reloptions(), and XLogPrefetcherReadRecord().

◆ NBuffers

PGDLLIMPORT int NBuffers

extern

Definition at line 142 of file globals.c.

Referenced by AioChooseMaxConcurrency(), apw_dump_now(), apw_load_buffers(), BgBufferSync(), BufferIsValid(), BufferManagerShmemInit(), BufferManagerShmemSize(), BufferSync(), CheckpointerShmemInit(), CheckpointerShmemSize(), ClockSweepTick(), CreateCheckPoint(), DropDatabaseBuffers(), DropRelationBuffers(), DropRelationsAllBuffers(), EvictAllUnpinnedBuffers(), EvictRelUnpinnedBuffers(), FlushDatabaseBuffers(), FlushRelationBuffers(), FlushRelationsAllBuffers(), GetAccessStrategyPinLimit(), GetAccessStrategyWithSize(), hashbuild(), InitBufferManagerAccess(), initscan(), LogCheckpointEnd(), MarkDirtyAllUnpinnedBuffers(), MarkDirtyRelUnpinnedBuffers(), pg_buffercache_evict(), pg_buffercache_mark_dirty(), pg_buffercache_os_pages_internal(), pg_buffercache_pages(), pg_buffercache_summary(), SimpleLruAutotuneBuffers(), StrategyGetBuffer(), StrategyInitialize(), StrategyShmemSize(), StrategySyncStart(), table_block_parallelscan_initialize(), and XLOGChooseNumBuffers().

◆ NLocBuffer

PGDLLIMPORT int NLocBuffer

extern

Definition at line 45 of file localbuf.c.

Referenced by BufferIsValid(), check_temp_buffers(), CheckForLocalBufferLeaks(), DropRelationAllLocalBuffers(), DropRelationLocalBuffers(), FlushRelationBuffers(), GetLocalBufferStorage(), GetLocalVictimBuffer(), hashbuild(), and InitLocalBuffers().

◆ track_io_timing

PGDLLIMPORT bool track_io_timing

extern

Definition at line 147 of file bufmgr.c.

Referenced by AsyncReadBuffers(), BufFileDumpBuffer(), BufFileLoadBuffer(), do_analyze_rel(), ExtendBufferedRelLocal(), ExtendBufferedRelShared(), FlushBuffer(), FlushLocalBuffer(), heap_vacuum_rel(), IssuePendingWritebacks(), mdsyncfiletag(), pgstat_count_backend_io_op_time(), register_dirty_segment(), show_buffer_usage(), and WaitReadBuffers().

◆ zero_damaged_pages

PGDLLIMPORT bool zero_damaged_pages

extern

Definition at line 144 of file bufmgr.c.

Referenced by AsyncReadBuffers(), mdreadv(), and read_rel_block_ll().

Variables
PGDLLIMPORT int	NBuffers

PGDLLIMPORT bool	zero_damaged_pages

PGDLLIMPORT int	bgwriter_lru_maxpages

PGDLLIMPORT double	bgwriter_lru_multiplier

PGDLLIMPORT bool	track_io_timing

PGDLLIMPORT int	effective_io_concurrency

PGDLLIMPORT int	maintenance_io_concurrency

PGDLLIMPORT int	io_combine_limit

PGDLLIMPORT int	io_combine_limit_guc

PGDLLIMPORT int	io_max_combine_limit

PGDLLIMPORT int	checkpoint_flush_after

PGDLLIMPORT int	backend_flush_after

PGDLLIMPORT int	bgwriter_flush_after

PGDLLIMPORT const PgAioHandleCallbacks	aio_shared_buffer_readv_cb

PGDLLIMPORT const PgAioHandleCallbacks	aio_local_buffer_readv_cb

PGDLLIMPORT char *	BufferBlocks

PGDLLIMPORT int	NLocBuffer

PGDLLIMPORT Block *	LocalBufferBlockPointers

PGDLLIMPORT int32 *	LocalRefCount

Data Structures

Macros

Typedefs

Enumerations

Functions

Variables

Macro Definition Documentation

◆ BMR_GET_SMGR

◆ BMR_REL

◆ BMR_SMGR

◆ DEFAULT_EFFECTIVE_IO_CONCURRENCY

◆ DEFAULT_IO_COMBINE_LIMIT

◆ DEFAULT_MAINTENANCE_IO_CONCURRENCY

◆ MAX_IO_COMBINE_LIMIT

◆ MAX_IO_CONCURRENCY

◆ P_NEW

◆ READ_BUFFERS_IGNORE_CHECKSUM_FAILURES

◆ READ_BUFFERS_ISSUE_ADVICE

◆ READ_BUFFERS_SYNCHRONOUSLY

◆ READ_BUFFERS_ZERO_ON_ERROR

◆ RelationGetNumberOfBlocks

Typedef Documentation

◆ Block

◆ BufferAccessStrategyType

◆ BufferLockMode

◆ BufferManagerRelation

◆ ExtendBufferedFlags

◆ PrefetchBufferResult

◆ ReadBuffersOperation

◆ SMgrRelation

◆ WritebackContext

Enumeration Type Documentation

◆ BufferAccessStrategyType

◆ BufferLockMode

◆ ExtendBufferedFlags

◆ ReadBufferMode

Function Documentation

◆ AtEOXact_Buffers()

◆ AtProcExit_LocalBuffers()

◆ BgBufferSync()

◆ BufferGetBlock()

◆ BufferGetBlockNumber()

◆ BufferGetLSNAtomic()

◆ BufferGetPage()

◆ BufferGetPageSize()

◆ BufferGetTag()

◆ BufferIsDirty()

◆ BufferIsLockedByMe()

◆ BufferIsLockedByMeInMode()

◆ BufferIsPermanent()

◆ BufferIsValid()

◆ BufferManagerShmemInit()

◆ BufferManagerShmemSize()

◆ CheckBufferIsPinnedOnce()

◆ CheckPointBuffers()

◆ ConditionalLockBuffer()

◆ ConditionalLockBufferForCleanup()

◆ CreateAndCopyRelationData()

◆ DebugPrintBufferRefcount()

◆ DropDatabaseBuffers()

◆ DropRelationBuffers()

◆ DropRelationsAllBuffers()

◆ EvictAllUnpinnedBuffers()

◆ EvictRelUnpinnedBuffers()

◆ EvictUnpinnedBuffer()

◆ ExtendBufferedRel()

◆ ExtendBufferedRelBy()

◆ ExtendBufferedRelTo()

◆ FlushDatabaseBuffers()

◆ FlushOneBuffer()

◆ FlushRelationBuffers()

◆ FlushRelationsAllBuffers()

◆ FreeAccessStrategy()

◆ GetAccessStrategy()

◆ GetAccessStrategyBufferCount()

◆ GetAccessStrategyPinLimit()

◆ GetAccessStrategyWithSize()

◆ GetAdditionalLocalPinLimit()

◆ GetAdditionalPinLimit()

◆ GetLocalPinLimit()