opensourcemirror
/
postgresql
mirror of git://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Releases Wiki Activity
postgresql/src/backend/storage/lmgr/lock.c

2026 lines
53 KiB
C
Raw Blame History

/*-------------------------------------------------------------------------
*
* lock.c--
* simple lock acquisition
*
* Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/storage/lmgr/lock.c,v 1.38.2.1 1999/03/07 02:00:49 tgl Exp $
*
* NOTES
* Outside modules can create a lock table and acquire/release
* locks. A lock table is a shared memory hash table. When
* a process tries to acquire a lock of a type that conflictRs
* with existing locks, it is put to sleep using the routines
* in storage/lmgr/proc.c.
*
* Interface:
*
* LockAcquire(), LockRelease(), LockMethodTableInit(),
* LockMethodTableRename(), LockReleaseAll, LockOwners()
* LockResolveConflicts(), GrantLock()
*
* NOTE: This module is used to define new lock tables. The
* multi-level lock table (multi.c) used by the heap
* access methods calls these routines. See multi.c for
* examples showing how to use this interface.
*
*-------------------------------------------------------------------------
*/
#include <stdio.h> /* for sprintf() */
#include <string.h>
#include <sys/types.h>
#include <unistd.h>
#include <signal.h>
#include "postgres.h"
#include "miscadmin.h"
#include "storage/shmem.h"
#include "storage/sinvaladt.h"
#include "storage/spin.h"
#include "storage/proc.h"
#include "storage/lock.h"
#include "utils/hsearch.h"
#include "utils/memutils.h"
#include "utils/palloc.h"
#include "access/xact.h"
#include "access/transam.h"
#include "utils/trace.h"
#include "utils/ps_status.h"
static int WaitOnLock(LOCKMETHOD lockmethod, LOCK *lock, LOCKMODE lockmode,
TransactionId xid);
/*
* lockDebugRelation can be used to trace unconditionally a single relation,
* for example pg_listener, if you suspect there are locking problems.
*
* lockDebugOidMin is is used to avoid tracing postgres relations, which
* would produce a lot of output. Unfortunately most system relations are
* created after bootstrap and have oid greater than BootstrapObjectIdData.
* If you are using tprintf you should specify a value greater than the max
* oid of system relations, which can be found with the following query:
*
* select max(int4in(int4out(oid))) from pg_class where relname ~ '^pg_';
*
* To get a useful lock trace you can use the following pg_options:
*
* -T "verbose,query,locks,userlocks,lock_debug_oidmin=17500"
*/
#define LOCKDEBUG(lockmethod) (pg_options[TRACE_SHORTLOCKS+lockmethod])
#define lockDebugRelation (pg_options[TRACE_LOCKRELATION])
#define lockDebugOidMin (pg_options[TRACE_LOCKOIDMIN])
#define lockReadPriority (pg_options[OPT_LOCKREADPRIORITY])
#ifdef LOCK_MGR_DEBUG
#define LOCK_PRINT(where,lock,type) \
if (((LOCKDEBUG(LOCK_LOCKMETHOD(*(lock))) >= 1) \
&& (lock->tag.relId >= lockDebugOidMin)) \
|| (lock->tag.relId == lockDebugRelation)) \
LOCK_PRINT_AUX(where,lock,type)
#define LOCK_PRINT_AUX(where,lock,type) \
TPRINTF(TRACE_ALL, \
"%s: lock(%x) tbl(%d) rel(%d) db(%d) tid(%d,%d) mask(%x) " \
"hold(%d,%d,%d,%d,%d)=%d " \
"act(%d,%d,%d,%d,%d)=%d wait(%d) type(%s)", \
where, \
MAKE_OFFSET(lock), \
lock->tag.lockmethod, \
lock->tag.relId, \
lock->tag.dbId, \
((lock->tag.tupleId.ip_blkid.bi_hi<<16)+ \
lock->tag.tupleId.ip_blkid.bi_lo), \
lock->tag.tupleId.ip_posid, \
lock->mask, \
lock->holders[1], \
lock->holders[2], \
lock->holders[3], \
lock->holders[4], \
lock->holders[5], \
lock->nHolding, \
lock->activeHolders[1], \
lock->activeHolders[2], \
lock->activeHolders[3], \
lock->activeHolders[4], \
lock->activeHolders[5], \
lock->nActive, \
lock->waitProcs.size, \
lock_types[type])
#define XID_PRINT(where,xidentP) \
if (((LOCKDEBUG(XIDENT_LOCKMETHOD(*(xidentP))) >= 1) \
&& (((LOCK *)MAKE_PTR(xidentP->tag.lock))->tag.relId \
>= lockDebugOidMin)) \
|| (((LOCK *)MAKE_PTR(xidentP->tag.lock))->tag.relId \
== lockDebugRelation)) \
XID_PRINT_AUX(where,xidentP)
#define XID_PRINT_AUX(where,xidentP) \
TPRINTF(TRACE_ALL, \
"%s: xid(%x) lock(%x) tbl(%d) pid(%d) xid(%d) " \
"hold(%d,%d,%d,%d,%d)=%d", \
where, \
MAKE_OFFSET(xidentP), \
xidentP->tag.lock, \
XIDENT_LOCKMETHOD(*(xidentP)), \
xidentP->tag.pid, \
xidentP->tag.xid, \
xidentP->holders[1], \
xidentP->holders[2], \
xidentP->holders[3], \
xidentP->holders[4], \
xidentP->holders[5], \
xidentP->nHolding)
#else /* !LOCK_MGR_DEBUG */
#define LOCK_PRINT(where,lock,type)
#define LOCK_PRINT_AUX(where,lock,type)
#define XID_PRINT(where,xidentP)
#define XID_PRINT_AUX(where,xidentP)
#endif /* !LOCK_MGR_DEBUG */
static char *lock_types[] = {
"",
"WRITE",
"READ",
"WRITE INTENT",
"READ INTENT",
"EXTEND"
};
SPINLOCK LockMgrLock; /* in Shmem or created in
* CreateSpinlocks() */
/* This is to simplify/speed up some bit arithmetic */
static MASK BITS_OFF[MAX_LOCKMODES];
static MASK BITS_ON[MAX_LOCKMODES];
/* -----------------
* XXX Want to move this to this file
* -----------------
*/
static bool LockingIsDisabled;
/* -------------------
* map from lockmethod to the lock table structure
* -------------------
*/
static LOCKMETHODTABLE *LockMethodTable[MAX_LOCK_METHODS];
static int NumLockMethods;
/* -------------------
* InitLocks -- Init the lock module. Create a private data
* structure for constructing conflict masks.
* -------------------
*/
void
InitLocks()
{
int i;
int bit;
bit = 1;
/* -------------------
* remember 0th lockmode is invalid
* -------------------
*/
for (i = 0; i < MAX_LOCKMODES; i++, bit <<= 1)
{
BITS_ON[i] = bit;
BITS_OFF[i] = ~bit;
}
#ifdef LOCK_MGR_DEBUG
/*
* If lockDebugOidMin value has not been specified in pg_options set a
* default value.
*/
if (!lockDebugOidMin)
lockDebugOidMin = BootstrapObjectIdData;
#endif
}
/* -------------------
* LockDisable -- sets LockingIsDisabled flag to TRUE or FALSE.
* ------------------
*/
void
LockDisable(int status)
{
LockingIsDisabled = status;
}
/*
* LockMethodInit -- initialize the lock table's lock type
* structures
*
* Notes: just copying. Should only be called once.
*/
static void
LockMethodInit(LOCKMETHODTABLE * lockMethodTable,
MASK *conflictsP,
int *prioP,
int numModes)
{
int i;
lockMethodTable->ctl->numLockModes = numModes;
numModes++;
for (i = 0; i < numModes; i++, prioP++, conflictsP++)
{
lockMethodTable->ctl->conflictTab[i] = *conflictsP;
lockMethodTable->ctl->prio[i] = *prioP;
}
}
/*
* LockMethodTableInit -- initialize a lock table structure
*
* Notes:
* (a) a lock table has four separate entries in the shmem index
* table. This is because every shared hash table and spinlock
* has its name stored in the shmem index at its creation. It
* is wasteful, in this case, but not much space is involved.
*
*/
LOCKMETHOD
LockMethodTableInit(char *tabName,
MASK *conflictsP,
int *prioP,
int numModes)
{
LOCKMETHODTABLE *lockMethodTable;
char *shmemName;
HASHCTL info;
int hash_flags;
bool found;
int status = TRUE;
if (numModes > MAX_LOCKMODES)
{
elog(NOTICE, "LockMethodTableInit: too many lock types %d greater than %d",
numModes, MAX_LOCKMODES);
return INVALID_LOCKMETHOD;
}
/* allocate a string for the shmem index table lookup */
shmemName = (char *) palloc((unsigned) (strlen(tabName) + 32));
if (!shmemName)
{
elog(NOTICE, "LockMethodTableInit: couldn't malloc string %s \n", tabName);
return INVALID_LOCKMETHOD;
}
/* each lock table has a non-shared header */
lockMethodTable = (LOCKMETHODTABLE *) palloc((unsigned) sizeof(LOCKMETHODTABLE));
if (!lockMethodTable)
{
elog(NOTICE, "LockMethodTableInit: couldn't malloc lock table %s\n", tabName);
pfree(shmemName);
return INVALID_LOCKMETHOD;
}
/* ------------------------
* find/acquire the spinlock for the table
* ------------------------
*/
SpinAcquire(LockMgrLock);
/* -----------------------
* allocate a control structure from shared memory or attach to it
* if it already exists.
* -----------------------
*/
sprintf(shmemName, "%s (ctl)", tabName);
lockMethodTable->ctl = (LOCKMETHODCTL *)
ShmemInitStruct(shmemName, (unsigned) sizeof(LOCKMETHODCTL), &found);
if (!lockMethodTable->ctl)
{
elog(FATAL, "LockMethodTableInit: couldn't initialize %s", tabName);
status = FALSE;
}
/* -------------------
* no zero-th table
* -------------------
*/
NumLockMethods = 1;
/* ----------------
* we're first - initialize
* ----------------
*/
if (!found)
{
MemSet(lockMethodTable->ctl, 0, sizeof(LOCKMETHODCTL));
lockMethodTable->ctl->masterLock = LockMgrLock;
lockMethodTable->ctl->lockmethod = NumLockMethods;
}
/* --------------------
* other modules refer to the lock table by a lockmethod
* --------------------
*/
LockMethodTable[NumLockMethods] = lockMethodTable;
NumLockMethods++;
Assert(NumLockMethods <= MAX_LOCK_METHODS);
/* ----------------------
* allocate a hash table for the lock tags. This is used
* to find the different locks.
* ----------------------
*/
info.keysize = SHMEM_LOCKTAB_KEYSIZE;
info.datasize = SHMEM_LOCKTAB_DATASIZE;
info.hash = tag_hash;
hash_flags = (HASH_ELEM | HASH_FUNCTION);
sprintf(shmemName, "%s (lock hash)", tabName);
lockMethodTable->lockHash = (HTAB *) ShmemInitHash(shmemName,
INIT_TABLE_SIZE, MAX_TABLE_SIZE,
&info, hash_flags);
Assert(lockMethodTable->lockHash->hash == tag_hash);
if (!lockMethodTable->lockHash)
{
elog(FATAL, "LockMethodTableInit: couldn't initialize %s", tabName);
status = FALSE;
}
/* -------------------------
* allocate an xid table. When different transactions hold
* the same lock, additional information must be saved (locks per tx).
* -------------------------
*/
info.keysize = SHMEM_XIDTAB_KEYSIZE;
info.datasize = SHMEM_XIDTAB_DATASIZE;
info.hash = tag_hash;
hash_flags = (HASH_ELEM | HASH_FUNCTION);
sprintf(shmemName, "%s (xid hash)", tabName);
lockMethodTable->xidHash = (HTAB *) ShmemInitHash(shmemName,
INIT_TABLE_SIZE, MAX_TABLE_SIZE,
&info, hash_flags);
if (!lockMethodTable->xidHash)
{
elog(FATAL, "LockMethodTableInit: couldn't initialize %s", tabName);
status = FALSE;
}
/* init ctl data structures */
LockMethodInit(lockMethodTable, conflictsP, prioP, numModes);
SpinRelease(LockMgrLock);
pfree(shmemName);
if (status)
return lockMethodTable->ctl->lockmethod;
else
return INVALID_LOCKMETHOD;
}
/*
* LockMethodTableRename -- allocate another lockmethod to the same
* lock table.
*
* NOTES: Both the lock module and the lock chain (lchain.c)
* module use table id's to distinguish between different
* kinds of locks. Short term and long term locks look
* the same to the lock table, but are handled differently
* by the lock chain manager. This function allows the
* client to use different lockmethods when acquiring/releasing
* short term and long term locks.
*/
LOCKMETHOD
LockMethodTableRename(LOCKMETHOD lockmethod)
{
LOCKMETHOD newLockMethod;
if (NumLockMethods >= MAX_LOCK_METHODS)
return INVALID_LOCKMETHOD;
if (LockMethodTable[lockmethod] == INVALID_LOCKMETHOD)
return INVALID_LOCKMETHOD;
/* other modules refer to the lock table by a lockmethod */
newLockMethod = NumLockMethods;
NumLockMethods++;
LockMethodTable[newLockMethod] = LockMethodTable[lockmethod];
return newLockMethod;
}
/*
* LockAcquire -- Check for lock conflicts, sleep if conflict found,
* set lock if/when no conflicts.
*
* Returns: TRUE if parameters are correct, FALSE otherwise.
*
* Side Effects: The lock is always acquired. No way to abort
* a lock acquisition other than aborting the transaction.
* Lock is recorded in the lkchain.
*
#ifdef USER_LOCKS
*
* Note on User Locks:
*
* User locks are handled totally on the application side as
* long term cooperative locks which extend beyond the normal
* transaction boundaries. Their purpose is to indicate to an
* application that someone is `working' on an item. So it is
* possible to put an user lock on a tuple's oid, retrieve the
* tuple, work on it for an hour and then update it and remove
* the lock. While the lock is active other clients can still
* read and write the tuple but they can be aware that it has
* been locked at the application level by someone.
* User locks use lock tags made of an uint16 and an uint32, for
* example 0 and a tuple oid, or any other arbitrary pair of
* numbers following a convention established by the application.
* In this sense tags don't refer to tuples or database entities.
* User locks and normal locks are completely orthogonal and
* they don't interfere with each other, so it is possible
* to acquire a normal lock on an user-locked tuple or user-lock
* a tuple for which a normal write lock already exists.
* User locks are always non blocking, therefore they are never
* acquired if already held by another process. They must be
* released explicitly by the application but they are released
* automatically when a backend terminates.
* They are indicated by a lockmethod 2 which is an alias for the
* normal lock table, and are distinguished from normal locks
* for the following differences:
*
* normal lock user lock
*
* lockmethod 1 2
* tag.relId rel oid 0
* tag.ItemPointerData.ip_blkid block id lock id2
* tag.ItemPointerData.ip_posid tuple offset lock id1
* xid.pid 0 backend pid
* xid.xid xid or 0 0
* persistence transaction user or backend
*
* The lockmode parameter can have the same values for normal locks
* although probably only WRITE_LOCK can have some practical use.
*
* DZ - 22 Nov 1997
#endif
*/
bool
LockAcquire(LOCKMETHOD lockmethod, LOCKTAG *locktag, LOCKMODE lockmode)
{
XIDLookupEnt *result,
item;
HTAB *xidTable;
bool found;
LOCK *lock = NULL;
SPINLOCK masterLock;
LOCKMETHODTABLE *lockMethodTable;
int status;
TransactionId xid;
#ifdef USER_LOCKS
int is_user_lock;
is_user_lock = (lockmethod == USER_LOCKMETHOD);
if (is_user_lock)
{
#ifdef USER_LOCKS_DEBUG
TPRINTF(TRACE_USERLOCKS, "LockAcquire: user lock [%u,%u] %s",
locktag->tupleId.ip_posid,
((locktag->tupleId.ip_blkid.bi_hi << 16) +
locktag->tupleId.ip_blkid.bi_lo),
lock_types[lockmode]);
#endif
}
#endif
/* ???????? This must be changed when short term locks will be used */
locktag->lockmethod = lockmethod;
Assert(lockmethod < NumLockMethods);
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
{
elog(NOTICE, "LockAcquire: bad lock table %d", lockmethod);
return FALSE;
}
if (LockingIsDisabled)
return TRUE;
masterLock = lockMethodTable->ctl->masterLock;
SpinAcquire(masterLock);
/*
* Find or create a lock with this tag
*/
Assert(lockMethodTable->lockHash->hash == tag_hash);
lock = (LOCK *) hash_search(lockMethodTable->lockHash, (Pointer) locktag,
HASH_ENTER, &found);
if (!lock)
{
SpinRelease(masterLock);
elog(FATAL, "LockAcquire: lock table %d is corrupted", lockmethod);
return FALSE;
}
/* --------------------
* if there was nothing else there, complete initialization
* --------------------
*/
if (!found)
{
lock->mask = 0;
lock->nHolding = 0;
lock->nActive = 0;
MemSet((char *) lock->holders, 0, sizeof(int) * MAX_LOCKMODES);
MemSet((char *) lock->activeHolders, 0, sizeof(int) * MAX_LOCKMODES);
ProcQueueInit(&(lock->waitProcs));
Assert(BlockIdEquals(&(lock->tag.tupleId.ip_blkid),
&(locktag->tupleId.ip_blkid)));
LOCK_PRINT("LockAcquire: new", lock, lockmode);
}
else
{
LOCK_PRINT("LockAcquire: found", lock, lockmode);
Assert((lock->nHolding > 0) && (lock->holders[lockmode] >= 0));
Assert((lock->nActive > 0) && (lock->activeHolders[lockmode] >= 0));
Assert(lock->nActive <= lock->nHolding);
}
/* ------------------
* add an element to the lock queue so that we can clear the
* locks at end of transaction.
* ------------------
*/
xidTable = lockMethodTable->xidHash;
/* ------------------
* Zero out all of the tag bytes (this clears the padding bytes for long
* word alignment and ensures hashing consistency).
* ------------------
*/
MemSet(&item, 0, XID_TAGSIZE); /* must clear padding, needed */
item.tag.lock = MAKE_OFFSET(lock);
#ifdef USE_XIDTAG_LOCKMETHOD
item.tag.lockmethod = lockmethod;
#endif
#ifdef USER_LOCKS
if (is_user_lock)
{
item.tag.pid = MyProcPid;
item.tag.xid = xid = 0;
}
else
{
xid = GetCurrentTransactionId();
TransactionIdStore(xid, &item.tag.xid);
}
#else
xid = GetCurrentTransactionId();
TransactionIdStore(xid, &item.tag.xid);
#endif
/*
* Find or create an xid entry with this tag
*/
result = (XIDLookupEnt *) hash_search(xidTable, (Pointer) &item,
HASH_ENTER, &found);
if (!result)
{
elog(NOTICE, "LockAcquire: xid table corrupted");
return STATUS_ERROR;
}
/*
* If not found initialize the new entry
*/
if (!found)
{
result->nHolding = 0;
MemSet((char *) result->holders, 0, sizeof(int) * MAX_LOCKMODES);
ProcAddLock(&result->queue);
XID_PRINT("LockAcquire: new", result);
}
else
{
XID_PRINT("LockAcquire: found", result);
Assert((result->nHolding > 0) && (result->holders[lockmode] >= 0));
Assert(result->nHolding <= lock->nActive);
}
/* ----------------
* lock->nholding tells us how many processes have _tried_ to
* acquire this lock, Regardless of whether they succeeded or
* failed in doing so.
* ----------------
*/
lock->nHolding++;
lock->holders[lockmode]++;
Assert((lock->nHolding > 0) && (lock->holders[lockmode] > 0));
/* --------------------
* If I'm the only one holding a lock, then there
* cannot be a conflict. Need to subtract one from the
* lock's count since we just bumped the count up by 1
* above.
* --------------------
*/
if (result->nHolding == lock->nActive)
{
result->holders[lockmode]++;
result->nHolding++;
XID_PRINT("LockAcquire: owning", result);
Assert((result->nHolding > 0) && (result->holders[lockmode] > 0));
GrantLock(lock, lockmode);
SpinRelease(masterLock);
return TRUE;
}
status = LockResolveConflicts(lockmethod, lock, lockmode, xid, result);
if (status == STATUS_OK)
GrantLock(lock, lockmode);
else if (status == STATUS_FOUND)
{
#ifdef USER_LOCKS
/*
* User locks are non blocking. If we can't acquire a lock we must
* remove the xid entry and return FALSE without waiting.
*/
if (is_user_lock)
{
if (!result->nHolding)
{
SHMQueueDelete(&result->queue);
result = (XIDLookupEnt *) hash_search(xidTable,
(Pointer) result,
HASH_REMOVE, &found);
if (!result || !found)
elog(NOTICE, "LockAcquire: remove xid, table corrupted");
}
else
XID_PRINT_AUX("LockAcquire: NHOLDING", result);
lock->nHolding--;
lock->holders[lockmode]--;
LOCK_PRINT("LockAcquire: user lock failed", lock, lockmode);
Assert((lock->nHolding > 0) && (lock->holders[lockmode] >= 0));
Assert(lock->nActive <= lock->nHolding);
SpinRelease(masterLock);
return FALSE;
}
#endif
status = WaitOnLock(lockmethod, lock, lockmode, xid);
/*
* Check the xid entry status, in case something in the ipc
* communication doesn't work correctly.
*/
if (!((result->nHolding > 0) && (result->holders[lockmode] > 0)))
{
XID_PRINT_AUX("LockAcquire: INCONSISTENT ", result);
LOCK_PRINT_AUX("LockAcquire: INCONSISTENT ", lock, lockmode);
/* Should we retry ? */
return FALSE;
}
XID_PRINT("LockAcquire: granted", result);
LOCK_PRINT("LockAcquire: granted", lock, lockmode);
}
SpinRelease(masterLock);
return status == STATUS_OK;
}
/* ----------------------------
* LockResolveConflicts -- test for lock conflicts
*
* NOTES:
* Here's what makes this complicated: one transaction's
* locks don't conflict with one another. When many processes
* hold locks, each has to subtract off the other's locks when
* determining whether or not any new lock acquired conflicts with
* the old ones.
*
* For example, if I am already holding a WRITE_INTENT lock,
* there will not be a conflict with my own READ_LOCK. If I
* don't consider the intent lock when checking for conflicts,
* I find no conflict.
* ----------------------------
*/
int
LockResolveConflicts(LOCKMETHOD lockmethod,
LOCK *lock,
LOCKMODE lockmode,
TransactionId xid,
XIDLookupEnt *xidentP) /* xident ptr or NULL */
{
XIDLookupEnt *result,
item;
int *myHolders;
int numLockModes;
HTAB *xidTable;
bool found;
int bitmask;
int i,
tmpMask;
#ifdef USER_LOCKS
int is_user_lock;
#endif
numLockModes = LockMethodTable[lockmethod]->ctl->numLockModes;
xidTable = LockMethodTable[lockmethod]->xidHash;
if (xidentP)
{
/*
* A pointer to the xid entry was supplied from the caller.
* Actually only LockAcquire can do it.
*/
result = xidentP;
}
else
{
/* ---------------------
* read my own statistics from the xid table. If there
* isn't an entry, then we'll just add one.
*
* Zero out the tag, this clears the padding bytes for long
* word alignment and ensures hashing consistency.
* ------------------
*/
MemSet(&item, 0, XID_TAGSIZE);
item.tag.lock = MAKE_OFFSET(lock);
#ifdef USE_XIDTAG_LOCKMETHOD
item.tag.lockmethod = lockmethod;
#endif
#ifdef USER_LOCKS
is_user_lock = (lockmethod == 2);
if (is_user_lock)
{
item.tag.pid = MyProcPid;
item.tag.xid = 0;
}
else
TransactionIdStore(xid, &item.tag.xid);
#else
TransactionIdStore(xid, &item.tag.xid);
#endif
/*
* Find or create an xid entry with this tag
*/
result = (XIDLookupEnt *) hash_search(xidTable, (Pointer) &item,
HASH_ENTER, &found);
if (!result)
{
elog(NOTICE, "LockResolveConflicts: xid table corrupted");
return STATUS_ERROR;
}
/*
* If not found initialize the new entry. THIS SHOULD NEVER
* HAPPEN, if we are trying to resolve a conflict we must already
* have allocated an xid entry for this lock. dz 21-11-1997
*/
if (!found)
{
/* ---------------
* we're not holding any type of lock yet. Clear
* the lock stats.
* ---------------
*/
MemSet(result->holders, 0, numLockModes * sizeof(*(lock->holders)));
result->nHolding = 0;
XID_PRINT_AUX("LockResolveConflicts: NOT FOUND", result);
}
else
XID_PRINT("LockResolveConflicts: found", result);
}
Assert((result->nHolding >= 0) && (result->holders[lockmode] >= 0));
/*
* We can control runtime this option. Default is lockReadPriority=0
*/
if (!lockReadPriority)
{
/* ------------------------
* If someone with a greater priority is waiting for the lock,
* do not continue and share the lock, even if we can. bjm
* ------------------------
*/
int myprio = LockMethodTable[lockmethod]->ctl->prio[lockmode];
PROC_QUEUE *waitQueue = &(lock->waitProcs);
PROC *topproc = (PROC *) MAKE_PTR(waitQueue->links.prev);
if (waitQueue->size && topproc->prio > myprio)
{
XID_PRINT("LockResolveConflicts: higher priority proc waiting",
result);
return STATUS_FOUND;
}
}
/* ----------------------------
* first check for global conflicts: If no locks conflict
* with mine, then I get the lock.
*
* Checking for conflict: lock->mask represents the types of
* currently held locks. conflictTable[lockmode] has a bit
* set for each type of lock that conflicts with mine. Bitwise
* compare tells if there is a conflict.
* ----------------------------
*/
if (!(LockMethodTable[lockmethod]->ctl->conflictTab[lockmode] & lock->mask))
{
result->holders[lockmode]++;
result->nHolding++;
XID_PRINT("LockResolveConflicts: no conflict", result);
Assert((result->nHolding > 0) && (result->holders[lockmode] > 0));
return STATUS_OK;
}
/* ------------------------
* Rats. Something conflicts. But it could still be my own
* lock. We have to construct a conflict mask
* that does not reflect our own locks.
* ------------------------
*/
myHolders = result->holders;
bitmask = 0;
tmpMask = 2;
for (i = 1; i <= numLockModes; i++, tmpMask <<= 1)
{
if (lock->activeHolders[i] != myHolders[i])
bitmask |= tmpMask;
}
/* ------------------------
* now check again for conflicts. 'bitmask' describes the types
* of locks held by other processes. If one of these
* conflicts with the kind of lock that I want, there is a
* conflict and I have to sleep.
* ------------------------
*/
if (!(LockMethodTable[lockmethod]->ctl->conflictTab[lockmode] & bitmask))
{
/* no conflict. Get the lock and go on */
result->holders[lockmode]++;
result->nHolding++;
XID_PRINT("LockResolveConflicts: resolved", result);
Assert((result->nHolding > 0) && (result->holders[lockmode] > 0));
return STATUS_OK;
}
XID_PRINT("LockResolveConflicts: conflicting", result);
return STATUS_FOUND;
}
/*
* GrantLock -- update the lock data structure to show
* the new lock holder.
*/
void
GrantLock(LOCK *lock, LOCKMODE lockmode)
{
lock->nActive++;
lock->activeHolders[lockmode]++;
lock->mask |= BITS_ON[lockmode];
LOCK_PRINT("GrantLock", lock, lockmode);
Assert((lock->nActive > 0) && (lock->activeHolders[lockmode] > 0));
Assert(lock->nActive <= lock->nHolding);
}
static int
WaitOnLock(LOCKMETHOD lockmethod, LOCK *lock, LOCKMODE lockmode,
TransactionId xid)
{
PROC_QUEUE *waitQueue = &(lock->waitProcs);
LOCKMETHODTABLE *lockMethodTable = LockMethodTable[lockmethod];
int prio = lockMethodTable->ctl->prio[lockmode];
char old_status[64],
new_status[64];
Assert(lockmethod < NumLockMethods);
/*
* the waitqueue is ordered by priority. I insert myself according to
* the priority of the lock I am acquiring.
*
* SYNC NOTE: I am assuming that the lock table spinlock is sufficient
* synchronization for this queue. That will not be true if/when
* people can be deleted from the queue by a SIGINT or something.
*/
LOCK_PRINT_AUX("WaitOnLock: sleeping on lock", lock, lockmode);
strcpy(old_status, PS_STATUS);
strcpy(new_status, PS_STATUS);
strcat(new_status, " waiting");
PS_SET_STATUS(new_status);
if (ProcSleep(waitQueue,
lockMethodTable->ctl->masterLock,
lockmode,
prio,
lock,
xid) != NO_ERROR)
{
/* -------------------
* This could have happend as a result of a deadlock,
* see HandleDeadLock().
* Decrement the lock nHolding and holders fields as
* we are no longer waiting on this lock.
* -------------------
*/
lock->nHolding--;
lock->holders[lockmode]--;
LOCK_PRINT_AUX("WaitOnLock: aborting on lock", lock, lockmode);
Assert((lock->nHolding >= 0) && (lock->holders[lockmode] >= 0));
Assert(lock->nActive <= lock->nHolding);
SpinRelease(lockMethodTable->ctl->masterLock);
elog(ERROR, "WaitOnLock: error on wakeup - Aborting this transaction");
/* not reached */
}
PS_SET_STATUS(old_status);
LOCK_PRINT_AUX("WaitOnLock: wakeup on lock", lock, lockmode);
return STATUS_OK;
}
/*
* LockRelease -- look up 'locktag' in lock table 'lockmethod' and
* release it.
*
* Side Effects: if the lock no longer conflicts with the highest
* priority waiting process, that process is granted the lock
* and awoken. (We have to grant the lock here to avoid a
* race between the waking process and any new process to
* come along and request the lock).
*/
bool
LockRelease(LOCKMETHOD lockmethod, LOCKTAG *locktag, LOCKMODE lockmode)
{
LOCK *lock = NULL;
SPINLOCK masterLock;
bool found;
LOCKMETHODTABLE *lockMethodTable;
XIDLookupEnt *result,
item;
HTAB *xidTable;
TransactionId xid;
bool wakeupNeeded = true;
int trace_flag;
#ifdef USER_LOCKS
int is_user_lock;
is_user_lock = (lockmethod == USER_LOCKMETHOD);
if (is_user_lock)
{
TPRINTF(TRACE_USERLOCKS, "LockRelease: user lock tag [%u,%u] %d",
locktag->tupleId.ip_posid,
((locktag->tupleId.ip_blkid.bi_hi << 16) +
locktag->tupleId.ip_blkid.bi_lo),
lockmode);
}
#endif
/* ???????? This must be changed when short term locks will be used */
locktag->lockmethod = lockmethod;
#ifdef USER_LOCKS
trace_flag = \
(lockmethod == USER_LOCKMETHOD) ? TRACE_USERLOCKS : TRACE_LOCKS;
#else
trace_flag = TRACE_LOCKS;
#endif
Assert(lockmethod < NumLockMethods);
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
{
elog(NOTICE, "lockMethodTable is null in LockRelease");
return FALSE;
}
if (LockingIsDisabled)
return TRUE;
masterLock = lockMethodTable->ctl->masterLock;
SpinAcquire(masterLock);
/*
* Find a lock with this tag
*/
Assert(lockMethodTable->lockHash->hash == tag_hash);
lock = (LOCK *) hash_search(lockMethodTable->lockHash, (Pointer) locktag,
HASH_FIND, &found);
/*
* let the caller print its own error message, too. Do not
* elog(ERROR).
*/
if (!lock)
{
SpinRelease(masterLock);
elog(NOTICE, "LockRelease: locktable corrupted");
return FALSE;
}
if (!found)
{
SpinRelease(masterLock);
#ifdef USER_LOCKS
if (is_user_lock)
{
TPRINTF(TRACE_USERLOCKS, "LockRelease: no lock with this tag");
return FALSE;
}
#endif
elog(NOTICE, "LockRelease: locktable lookup failed, no lock");
return FALSE;
}
LOCK_PRINT("LockRelease: found", lock, lockmode);
Assert((lock->nHolding > 0) && (lock->holders[lockmode] >= 0));
Assert((lock->nActive > 0) && (lock->activeHolders[lockmode] >= 0));
Assert(lock->nActive <= lock->nHolding);
/* ------------------
* Zero out all of the tag bytes (this clears the padding bytes for long
* word alignment and ensures hashing consistency).
* ------------------
*/
MemSet(&item, 0, XID_TAGSIZE);
item.tag.lock = MAKE_OFFSET(lock);
#ifdef USE_XIDTAG_LOCKMETHOD
item.tag.lockmethod = lockmethod;
#endif
#ifdef USER_LOCKS
if (is_user_lock)
{
item.tag.pid = MyProcPid;
item.tag.xid = xid = 0;
}
else
{
xid = GetCurrentTransactionId();
TransactionIdStore(xid, &item.tag.xid);
}
#else
xid = GetCurrentTransactionId();
TransactionIdStore(xid, &item.tag.xid);
#endif
/*
* Find an xid entry with this tag
*/
xidTable = lockMethodTable->xidHash;
result = (XIDLookupEnt *) hash_search(xidTable, (Pointer) &item,
HASH_FIND_SAVE, &found);
if (!result || !found)
{
SpinRelease(masterLock);
#ifdef USER_LOCKS
if (!found && is_user_lock)
TPRINTF(TRACE_USERLOCKS, "LockRelease: no lock with this tag");
else
#endif
elog(NOTICE, "LockReplace: xid table corrupted");
return FALSE;
}
XID_PRINT("LockRelease: found", result);
Assert(result->tag.lock == MAKE_OFFSET(lock));
/*
* Check that we are actually holding a lock of the type we want to
* release.
*/
if (!(result->holders[lockmode] > 0))
{
SpinRelease(masterLock);
XID_PRINT_AUX("LockAcquire: WRONGTYPE", result);
elog(NOTICE, "LockRelease: you don't own a lock of type %s",
lock_types[lockmode]);
Assert(result->holders[lockmode] >= 0);
return FALSE;
}
Assert(result->nHolding > 0);
/*
* fix the general lock stats
*/
lock->nHolding--;
lock->holders[lockmode]--;
lock->nActive--;
lock->activeHolders[lockmode]--;
/* --------------------------
* If there are still active locks of the type I just released, no one
* should be woken up. Whoever is asleep will still conflict
* with the remaining locks.
* --------------------------
*/
if (lock->activeHolders[lockmode])
wakeupNeeded = false;
else
{
/* change the conflict mask. No more of this lock type. */
lock->mask &= BITS_OFF[lockmode];
}
LOCK_PRINT("LockRelease: updated", lock, lockmode);
Assert((lock->nHolding >= 0) && (lock->holders[lockmode] >= 0));
Assert((lock->nActive >= 0) && (lock->activeHolders[lockmode] >= 0));
Assert(lock->nActive <= lock->nHolding);
if (!lock->nHolding)
{
/* ------------------
* if there's no one waiting in the queue,
* we just released the last lock.
* Delete it from the lock table.
* ------------------
*/
Assert(lockMethodTable->lockHash->hash == tag_hash);
lock = (LOCK *) hash_search(lockMethodTable->lockHash,
(Pointer) &(lock->tag),
HASH_REMOVE,
&found);
Assert(lock && found);
wakeupNeeded = false;
}
/*
* now check to see if I have any private locks. If I do, decrement
* the counts associated with them.
*/
result->holders[lockmode]--;
result->nHolding--;
XID_PRINT("LockRelease: updated", result);
Assert((result->nHolding >= 0) && (result->holders[lockmode] >= 0));
/*
* If this was my last hold on this lock, delete my entry in the XID
* table.
*/
if (!result->nHolding)
{
if (result->queue.prev == INVALID_OFFSET)
elog(NOTICE, "LockRelease: xid.prev == INVALID_OFFSET");
if (result->queue.next == INVALID_OFFSET)
elog(NOTICE, "LockRelease: xid.next == INVALID_OFFSET");
if (result->queue.next != INVALID_OFFSET)
SHMQueueDelete(&result->queue);
XID_PRINT("LockRelease: deleting", result);
result = (XIDLookupEnt *) hash_search(xidTable, (Pointer) &result,
HASH_REMOVE_SAVED, &found);
if (!result || !found)
{
SpinRelease(masterLock);
elog(NOTICE, "LockRelease: remove xid, table corrupted");
return FALSE;
}
}
if (wakeupNeeded)
{
/* --------------------------
* Wake the first waiting process and grant him the lock if it
* doesn't conflict. The woken process must record the lock
* himself.
* --------------------------
*/
ProcLockWakeup(&(lock->waitProcs), lockmethod, lock);
}
else
{
if (((LOCKDEBUG(LOCK_LOCKMETHOD(*(lock))) >= 1) \
&& (lock->tag.relId >= lockDebugOidMin)) \
|| (lock->tag.relId == lockDebugRelation))
TPRINTF(TRACE_ALL, "LockRelease: no wakeup needed");
}
SpinRelease(masterLock);
return TRUE;
}
/*
* LockReleaseAll -- Release all locks in a process lock queue.
*/
bool
LockReleaseAll(LOCKMETHOD lockmethod, SHM_QUEUE *lockQueue)
{
PROC_QUEUE *waitQueue;
int done;
XIDLookupEnt *xidLook = NULL;
XIDLookupEnt *tmp = NULL;
XIDLookupEnt *result;
SHMEM_OFFSET end = MAKE_OFFSET(lockQueue);
SPINLOCK masterLock;
LOCKMETHODTABLE *lockMethodTable;
int i,
numLockModes;
LOCK *lock;
bool found;
int trace_flag;
int xidtag_lockmethod;
#ifdef USER_LOCKS
int is_user_lock_table,
count,
nleft;
count = nleft = 0;
is_user_lock_table = (lockmethod == USER_LOCKMETHOD);
trace_flag = (lockmethod == 2) ? TRACE_USERLOCKS : TRACE_LOCKS;
#else
trace_flag = TRACE_LOCKS;
#endif
TPRINTF(trace_flag, "LockReleaseAll: lockmethod=%d, pid=%d",
lockmethod, MyProcPid);
Assert(lockmethod < NumLockMethods);
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
{
elog(NOTICE, "LockAcquire: bad lockmethod %d", lockmethod);
return FALSE;
}
if (SHMQueueEmpty(lockQueue))
return TRUE;
numLockModes = lockMethodTable->ctl->numLockModes;
masterLock = lockMethodTable->ctl->masterLock;
SpinAcquire(masterLock);
SHMQueueFirst(lockQueue, (Pointer *) &xidLook, &xidLook->queue);
for (;;)
{
/*
* Sometimes the queue appears to be messed up.
*/
if (count++ > 1000)
{
elog(NOTICE, "LockReleaseAll: xid loop detected, giving up");
nleft = 0;
break;
}
/* ---------------------------
* XXX Here we assume the shared memory queue is circular and
* that we know its internal structure. Should have some sort of
* macros to allow one to walk it. mer 20 July 1991
* ---------------------------
*/
done = (xidLook->queue.next == end);
lock = (LOCK *) MAKE_PTR(xidLook->tag.lock);
xidtag_lockmethod = XIDENT_LOCKMETHOD(*xidLook);
if ((xidtag_lockmethod == lockmethod) || (trace_flag >= 2))
{
XID_PRINT("LockReleaseAll", xidLook);
LOCK_PRINT("LockReleaseAll", lock, 0);
}
#ifdef USE_XIDTAG_LOCKMETHOD
if (xidtag_lockmethod != LOCK_LOCKMETHOD(*lock))
elog(NOTICE, "LockReleaseAll: xid/lock method mismatch: %d != %d",
xidtag_lockmethod, lock->tag.lockmethod);
#endif
if ((xidtag_lockmethod != lockmethod) && (trace_flag >= 2))
{
TPRINTF(trace_flag, "LockReleaseAll: skipping other table");
nleft++;
goto next_item;
}
Assert(lock->nHolding > 0);
Assert(lock->nActive > 0);
Assert(lock->nActive <= lock->nHolding);
Assert(xidLook->nHolding >= 0);
Assert(xidLook->nHolding <= lock->nHolding);
#ifdef USER_LOCKS
if (is_user_lock_table)
{
if ((xidLook->tag.pid == 0) || (xidLook->tag.xid != 0))
{
TPRINTF(TRACE_USERLOCKS,
"LockReleaseAll: skiping normal lock [%d,%d,%d]",
xidLook->tag.lock, xidLook->tag.pid, xidLook->tag.xid);
nleft++;
goto next_item;
}
if (xidLook->tag.pid != MyProcPid)
{
/* Should never happen */
elog(NOTICE,
"LockReleaseAll: INVALID PID: [%u,%u] [%d,%d,%d]",
lock->tag.tupleId.ip_posid,
((lock->tag.tupleId.ip_blkid.bi_hi << 16) +
lock->tag.tupleId.ip_blkid.bi_lo),
xidLook->tag.lock, xidLook->tag.pid, xidLook->tag.xid);
nleft++;
goto next_item;
}
TPRINTF(TRACE_USERLOCKS,
"LockReleaseAll: releasing user lock [%u,%u] [%d,%d,%d]",
lock->tag.tupleId.ip_posid,
((lock->tag.tupleId.ip_blkid.bi_hi << 16) +
lock->tag.tupleId.ip_blkid.bi_lo),
xidLook->tag.lock, xidLook->tag.pid, xidLook->tag.xid);
}
else
{
/*
* Can't check xidLook->tag.xid, can be 0 also for normal
* locks
*/
if (xidLook->tag.pid != 0)
{
TPRINTF(TRACE_LOCKS,
"LockReleaseAll: skiping user lock [%u,%u] [%d,%d,%d]",
lock->tag.tupleId.ip_posid,
((lock->tag.tupleId.ip_blkid.bi_hi << 16) +
lock->tag.tupleId.ip_blkid.bi_lo),
xidLook->tag.lock, xidLook->tag.pid, xidLook->tag.xid);
nleft++;
goto next_item;
}
}
#endif
/* ------------------
* fix the general lock stats
* ------------------
*/
if (lock->nHolding != xidLook->nHolding)
{
for (i = 1; i <= numLockModes; i++)
{
Assert(xidLook->holders[i] >= 0);
lock->holders[i] -= xidLook->holders[i];
lock->activeHolders[i] -= xidLook->holders[i];
Assert((lock->holders[i] >= 0) \
&&(lock->activeHolders[i] >= 0));
if (!lock->activeHolders[i])
lock->mask &= BITS_OFF[i];
}
lock->nHolding -= xidLook->nHolding;
lock->nActive -= xidLook->nHolding;
Assert((lock->nHolding >= 0) && (lock->nActive >= 0));
Assert(lock->nActive <= lock->nHolding);
}
else
{
/* --------------
* set nHolding to zero so that we can garbage collect the lock
* down below...
* --------------
*/
lock->nHolding = 0;
/* Fix the lock status, just for next LOCK_PRINT message. */
for (i = 1; i <= numLockModes; i++)
{
Assert(lock->holders[i] == lock->activeHolders[i]);
lock->holders[i] = lock->activeHolders[i] = 0;
}
}
LOCK_PRINT("LockReleaseAll: updated", lock, 0);
/*
* Remove the xid from the process lock queue
*/
SHMQueueDelete(&xidLook->queue);
/* ----------------
* always remove the xidLookup entry, we're done with it now
* ----------------
*/
XID_PRINT("LockReleaseAll: deleting", xidLook);
result = (XIDLookupEnt *) hash_search(lockMethodTable->xidHash,
(Pointer) xidLook,
HASH_REMOVE,
&found);
if (!result || !found)
{
SpinRelease(masterLock);
elog(NOTICE, "LockReleaseAll: xid table corrupted");
return FALSE;
}
if (!lock->nHolding)
{
/* --------------------
* if there's no one waiting in the queue, we've just released
* the last lock.
* --------------------
*/
LOCK_PRINT("LockReleaseAll: deleting", lock, 0);
Assert(lockMethodTable->lockHash->hash == tag_hash);
lock = (LOCK *) hash_search(lockMethodTable->lockHash,
(Pointer) &(lock->tag),
HASH_REMOVE, &found);
if ((!lock) || (!found))
{
SpinRelease(masterLock);
elog(NOTICE, "LockReleaseAll: cannot remove lock from HTAB");
return FALSE;
}
}
else
{
/* --------------------
* Wake the first waiting process and grant him the lock if it
* doesn't conflict. The woken process must record the lock
* him/herself.
* --------------------
*/
waitQueue = &(lock->waitProcs);
ProcLockWakeup(waitQueue, lockmethod, lock);
}
#ifdef USER_LOCKS
next_item:
#endif
if (done)
break;
SHMQueueFirst(&xidLook->queue, (Pointer *) &tmp, &tmp->queue);
xidLook = tmp;
}
/*
* Reinitialize the queue only if nothing has been left in.
*/
if (nleft == 0)
{
TPRINTF(trace_flag, "LockReleaseAll: reinitializing lockQueue");
SHMQueueInit(lockQueue);
}
SpinRelease(masterLock);
TPRINTF(trace_flag, "LockReleaseAll: done");
return TRUE;
}
int
LockShmemSize()
{
int size = 0;
size += MAXALIGN(sizeof(PROC_HDR)); /* ProcGlobal */
size += MAXALIGN(MaxBackendId * sizeof(PROC)); /* each MyProc */
size += MAXALIGN(MaxBackendId * sizeof(LOCKMETHODCTL)); /* each
* lockMethodTable->ctl */
/* lockHash table */
size += hash_estimate_size(NLOCKENTS,
SHMEM_LOCKTAB_KEYSIZE,
SHMEM_LOCKTAB_DATASIZE);
/* xidHash table */
size += hash_estimate_size(MaxBackendId,
SHMEM_XIDTAB_KEYSIZE,
SHMEM_XIDTAB_DATASIZE);
/* Since the lockHash entry count above is only an estimate,
* add 10% safety margin.
*/
size += size / 10;
return size;
}
/* -----------------
* Boolean function to determine current locking status
* -----------------
*/
bool
LockingDisabled()
{
return LockingIsDisabled;
}
/*
* DeadlockCheck -- Checks for deadlocks for a given process
*
* We can't block on user locks, so no sense testing for deadlock
* because there is no blocking, and no timer for the block.
*
* This code takes a list of locks a process holds, and the lock that
* the process is sleeping on, and tries to find if any of the processes
* waiting on its locks hold the lock it is waiting for. If no deadlock
* is found, it goes on to look at all the processes waiting on their locks.
*
* We have already locked the master lock before being called.
*/
bool
DeadLockCheck(SHM_QUEUE *lockQueue, LOCK *findlock, bool skip_check)
{
int done;
XIDLookupEnt *xidLook = NULL;
XIDLookupEnt *tmp = NULL;
SHMEM_OFFSET end = MAKE_OFFSET(lockQueue);
LOCK *lock;
LOCKMETHODTABLE *lockMethodTable;
XIDLookupEnt *result,
item;
HTAB *xidTable;
bool found;
static PROC *checked_procs[MaxBackendId];
static int nprocs;
static bool MyNHolding;
/* initialize at start of recursion */
if (skip_check)
{
checked_procs[0] = MyProc;
nprocs = 1;
lockMethodTable = LockMethodTable[DEFAULT_LOCKMETHOD];
xidTable = lockMethodTable->xidHash;
MemSet(&item, 0, XID_TAGSIZE);
TransactionIdStore(MyProc->xid, &item.tag.xid);
item.tag.lock = MAKE_OFFSET(findlock);
#if 0
item.tag.pid = pid;
#endif
if (!(result = (XIDLookupEnt *)
hash_search(xidTable, (Pointer) &item, HASH_FIND, &found)) || !found)
{
elog(NOTICE, "LockAcquire: xid table corrupted");
return true;
}
MyNHolding = result->nHolding;
}
if (SHMQueueEmpty(lockQueue))
return false;
SHMQueueFirst(lockQueue, (Pointer *) &xidLook, &xidLook->queue);
XID_PRINT("DeadLockCheck", xidLook);
for (;;)
{
/* ---------------------------
* XXX Here we assume the shared memory queue is circular and
* that we know its internal structure. Should have some sort of
* macros to allow one to walk it. mer 20 July 1991
* ---------------------------
*/
done = (xidLook->queue.next == end);
lock = (LOCK *) MAKE_PTR(xidLook->tag.lock);
LOCK_PRINT("DeadLockCheck", lock, 0);
/*
* This is our only check to see if we found the lock we want.
*
* The lock we are waiting for is already in MyProc->lockQueue so we
* need to skip it here. We are trying to find it in someone
* else's lockQueue.
*/
if (lock == findlock && !skip_check)
return true;
{
PROC_QUEUE *waitQueue = &(lock->waitProcs);
PROC *proc;
int i;
int j;
proc = (PROC *) MAKE_PTR(waitQueue->links.prev);
for (i = 0; i < waitQueue->size; i++)
{
if (proc != MyProc &&
lock == findlock && /* skip_check also true */
MyNHolding) /* I already hold some lock on it */
{
/*
* For findlock's wait queue, we are interested in
* procs who are blocked waiting for a write-lock on
* the table we are waiting on, and already hold a
* lock on it. We first check to see if there is an
* escalation deadlock, where we hold a readlock and
* want a writelock, and someone else holds readlock
* on the same table, and wants a writelock.
*
* Basically, the test is, "Do we both hold some lock on
* findlock, and we are both waiting in the lock
* queue?"
*/
Assert(skip_check);
Assert(MyProc->prio == 2);
lockMethodTable = LockMethodTable[1];
xidTable = lockMethodTable->xidHash;
MemSet(&item, 0, XID_TAGSIZE);
TransactionIdStore(proc->xid, &item.tag.xid);
item.tag.lock = MAKE_OFFSET(findlock);
#if 0
item.tag.pid = pid;
#endif
if (!(result = (XIDLookupEnt *)
hash_search(xidTable, (Pointer) &item, HASH_FIND, &found)) || !found)
{
elog(NOTICE, "LockAcquire: xid table corrupted");
return true;
}
if (result->nHolding)
return true;
}
/*
* No sense in looking at the wait queue of the lock we
* are looking for. If lock == findlock, and I got here,
* skip_check must be true too.
*/
if (lock != findlock)
{
for (j = 0; j < nprocs; j++)
if (checked_procs[j] == proc)
break;
if (j >= nprocs && lock != findlock)
{
checked_procs[nprocs++] = proc;
Assert(nprocs <= MaxBackendId);
/*
* For non-MyProc entries, we are looking only
* waiters, not necessarily people who already
* hold locks and are waiting. Now we check for
* cases where we have two or more tables in a
* deadlock. We do this by continuing to search
* for someone holding a lock
*/
if (DeadLockCheck(&(proc->lockQueue), findlock, false))
return true;
}
}
proc = (PROC *) MAKE_PTR(proc->links.prev);
}
}
if (done)
break;
SHMQueueFirst(&xidLook->queue, (Pointer *) &tmp, &tmp->queue);
xidLook = tmp;
}
/* if we got here, no deadlock */
return false;
}
#ifdef NOT_USED
/*
* Return an array with the pids of all processes owning a lock.
* This works only for user locks because normal locks have no
* pid information in the corresponding XIDLookupEnt.
*/
ArrayType *
LockOwners(LOCKMETHOD lockmethod, LOCKTAG *locktag)
{
XIDLookupEnt *xidLook = NULL;
SPINLOCK masterLock;
LOCK *lock;
SHMEM_OFFSET lock_offset;
int count = 0;
LOCKMETHODTABLE *lockMethodTable;
HTAB *xidTable;
bool found;
int ndims,
nitems,
hdrlen,
size;
int lbounds[1],
hbounds[1];
ArrayType *array;
int *data_ptr;
/* Assume that no one will modify the result */
static int empty_array[] = {20, 1, 0, 0, 0};
#ifdef USER_LOCKS
int is_user_lock;
is_user_lock = (lockmethod == USER_LOCKMETHOD);
if (is_user_lock)
{
TPRINTF(TRACE_USERLOCKS, "LockOwners: user lock tag [%u,%u]",
locktag->tupleId.ip_posid,
((locktag->tupleId.ip_blkid.bi_hi << 16) +
locktag->tupleId.ip_blkid.bi_lo));
}
#endif
/* This must be changed when short term locks will be used */
locktag->lockmethod = lockmethod;
Assert((lockmethod >= MIN_LOCKMETHOD) && (lockmethod < NumLockMethods));
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
{
elog(NOTICE, "lockMethodTable is null in LockOwners");
return (ArrayType *) &empty_array;
}
if (LockingIsDisabled)
return (ArrayType *) &empty_array;
masterLock = lockMethodTable->ctl->masterLock;
SpinAcquire(masterLock);
/*
* Find a lock with this tag
*/
Assert(lockMethodTable->lockHash->hash == tag_hash);
lock = (LOCK *) hash_search(lockMethodTable->lockHash, (Pointer) locktag,
HASH_FIND, &found);
/*
* let the caller print its own error message, too. Do not elog(WARN).
*/
if (!lock)
{
SpinRelease(masterLock);
elog(NOTICE, "LockOwners: locktable corrupted");
return (ArrayType *) &empty_array;
}
if (!found)
{
SpinRelease(masterLock);
#ifdef USER_LOCKS
if (is_user_lock)
{
TPRINTF(TRACE_USERLOCKS, "LockOwners: no lock with this tag");
return (ArrayType *) &empty_array;
}
#endif
elog(NOTICE, "LockOwners: locktable lookup failed, no lock");
return (ArrayType *) &empty_array;
}
LOCK_PRINT("LockOwners: found", lock, 0);
Assert((lock->nHolding > 0) && (lock->nActive > 0));
Assert(lock->nActive <= lock->nHolding);
lock_offset = MAKE_OFFSET(lock);
/* Construct a 1-dimensional array */
ndims = 1;
hdrlen = ARR_OVERHEAD(ndims);
lbounds[0] = 0;
hbounds[0] = lock->nActive;
size = hdrlen + sizeof(int) * hbounds[0];
array = (ArrayType *) palloc(size);
MemSet(array, 0, size);
memmove((char *) array, (char *) &size, sizeof(int));
memmove((char *) ARR_NDIM_PTR(array), (char *) &ndims, sizeof(int));
memmove((char *) ARR_DIMS(array), (char *) hbounds, ndims * sizeof(int));
memmove((char *) ARR_LBOUND(array), (char *) lbounds, ndims * sizeof(int));
SET_LO_FLAG(false, array);
data_ptr = (int *) ARR_DATA_PTR(array);
xidTable = lockMethodTable->xidHash;
hash_seq(NULL);
nitems = 0;
while ((xidLook = (XIDLookupEnt *) hash_seq(xidTable)) &&
(xidLook != (XIDLookupEnt *) TRUE))
{
if (count++ > 1000)
{
elog(NOTICE, "LockOwners: possible loop, giving up");
break;
}
if (xidLook->tag.pid == 0)
{
XID_PRINT("LockOwners: no pid", xidLook);
continue;
}
if (!xidLook->tag.lock)
{
XID_PRINT("LockOwners: NULL LOCK", xidLook);
continue;
}
if (xidLook->tag.lock != lock_offset)
{
XID_PRINT("LockOwners: different lock", xidLook);
continue;
}
if (LOCK_LOCKMETHOD(*lock) != lockmethod)
{
XID_PRINT("LockOwners: other table", xidLook);
continue;
}
if (xidLook->nHolding <= 0)
{
XID_PRINT("LockOwners: not holding", xidLook);
continue;
}
if (nitems >= hbounds[0])
{
elog(NOTICE, "LockOwners: array size exceeded");
break;
}
/*
* Check that the holding process is still alive by sending him an
* unused (ignored) signal. If the kill fails the process is not
* alive.
*/
if ((xidLook->tag.pid != MyProcPid) \
&&(kill(xidLook->tag.pid, SIGCHLD)) != 0)
{
/* Return a negative pid to signal that process is dead */
data_ptr[nitems++] = -(xidLook->tag.pid);
XID_PRINT("LockOwners: not alive", xidLook);
/* XXX - TODO: remove this entry and update lock stats */
continue;
}
/* Found a process holding the lock */
XID_PRINT("LockOwners: holding", xidLook);
data_ptr[nitems++] = xidLook->tag.pid;
}
SpinRelease(masterLock);
/* Adjust the actual size of the array */
hbounds[0] = nitems;
size = hdrlen + sizeof(int) * hbounds[0];
memmove((char *) array, (char *) &size, sizeof(int));
memmove((char *) ARR_DIMS(array), (char *) hbounds, ndims * sizeof(int));
return array;
}
#endif
#ifdef DEADLOCK_DEBUG
/*
* Dump all locks in the proc->lockQueue. Must have already acquired
* the masterLock.
*/
void
DumpLocks()
{
SHMEM_OFFSET location;
PROC *proc;
SHM_QUEUE *lockQueue;
int done;
XIDLookupEnt *xidLook = NULL;
XIDLookupEnt *tmp = NULL;
SHMEM_OFFSET end;
SPINLOCK masterLock;
int numLockModes;
LOCK *lock;
int count = 0;
int lockmethod = DEFAULT_LOCKMETHOD;
LOCKMETHODTABLE *lockMethodTable;
ShmemPIDLookup(MyProcPid, &location);
if (location == INVALID_OFFSET)
return;
proc = (PROC *) MAKE_PTR(location);
if (proc != MyProc)
return;
lockQueue = &proc->lockQueue;
Assert(lockmethod < NumLockMethods);
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
return;
numLockModes = lockMethodTable->ctl->numLockModes;
masterLock = lockMethodTable->ctl->masterLock;
if (SHMQueueEmpty(lockQueue))
return;
SHMQueueFirst(lockQueue, (Pointer *) &xidLook, &xidLook->queue);
end = MAKE_OFFSET(lockQueue);
if (MyProc->waitLock)
LOCK_PRINT_AUX("DumpLocks: waiting on", MyProc->waitLock, 0);
for (;;)
{
if (count++ > 2000)
{
elog(NOTICE, "DumpLocks: xid loop detected, giving up");
break;
}
/* ---------------------------
* XXX Here we assume the shared memory queue is circular and
* that we know its internal structure. Should have some sort of
* macros to allow one to walk it. mer 20 July 1991
* ---------------------------
*/
done = (xidLook->queue.next == end);
lock = (LOCK *) MAKE_PTR(xidLook->tag.lock);
XID_PRINT_AUX("DumpLocks", xidLook);
LOCK_PRINT_AUX("DumpLocks", lock, 0);
if (done)
break;
SHMQueueFirst(&xidLook->queue, (Pointer *) &tmp, &tmp->queue);
xidLook = tmp;
}
}
/*
* Dump all postgres locks. Must have already acquired the masterLock.
*/
void
DumpAllLocks()
{
SHMEM_OFFSET location;
PROC *proc;
XIDLookupEnt *xidLook = NULL;
LOCK *lock;
int pid;
int count = 0;
int lockmethod = DEFAULT_LOCKMETHOD;
LOCKMETHODTABLE *lockMethodTable;
HTAB *xidTable;
pid = getpid();
ShmemPIDLookup(pid, &location);
if (location == INVALID_OFFSET)
return;
proc = (PROC *) MAKE_PTR(location);
if (proc != MyProc)
return;
Assert(lockmethod < NumLockMethods);
lockMethodTable = LockMethodTable[lockmethod];
if (!lockMethodTable)
return;
xidTable = lockMethodTable->xidHash;
if (MyProc->waitLock)
LOCK_PRINT_AUX("DumpAllLocks: waiting on", MyProc->waitLock, 0);
hash_seq(NULL);
while ((xidLook = (XIDLookupEnt *) hash_seq(xidTable)) &&
(xidLook != (XIDLookupEnt *) TRUE))
{
XID_PRINT_AUX("DumpAllLocks", xidLook);
if (xidLook->tag.lock)
{
lock = (LOCK *) MAKE_PTR(xidLook->tag.lock);
LOCK_PRINT_AUX("DumpAllLocks", lock, 0);
}
else
elog(DEBUG, "DumpAllLocks: xidLook->tag.lock = NULL");
if (count++ > 2000)
{
elog(NOTICE, "DumpAllLocks: possible loop, giving up");
break;
}
}
}
#endif
Reference in New Issue View Git Blame Copy Permalink