From 3761fe3c20bb040b15f0e8da58d824631da00caa Mon Sep 17 00:00:00 2001 From: Robert Haas Date: Fri, 16 Dec 2016 11:29:23 -0500 Subject: [PATCH] Simplify LWLock tranche machinery by removing array_base/array_stride. array_base and array_stride were added so that we could identify the offset of an LWLock within a tranche, but this facility is only very marginally used apart from the main tranche. So, give every lock in the main tranche its own tranche ID and get rid of array_base, array_stride, and all that's attached. For debugging facilities (Trace_lwlocks and LWLOCK_STATS) print the pointer address of the LWLock using %p instead of the offset. This is arguably more useful, and certainly a lot cheaper. Drop the offset-within-tranche from the information reported to dtrace and from one can't-happen message inside lwlock.c. The main user-visible impact of this change is that pg_stat_activity will now report all waits for LWLocks as "LWLock" rather than reporting some as "LWLockTranche" and others as "LWLockNamed". The main motivation for this change is that the need to specify an array_base and an array_stride is awkward for parallel query. There is only a very limited supply of tranche IDs so we can't just keep allocating new ones, and if we try to use the same tranche IDs every time then we run into trouble when multiple parallel contexts are use simultaneously. So if we didn't get rid of this mechanism we'd have to make it even more complicated. By simplifying it in this way, we instead reduce the size of the generated code for lwlock.c by about 5%. Discussion: http://postgr.es/m/CA+TgmoYsFn6NUW1x0AZtupJGUAs1UDY4dJtCN47_Q6D0sP80PA@mail.gmail.com --- doc/src/sgml/monitoring.sgml | 52 +++---- src/backend/access/transam/slru.c | 6 +- src/backend/access/transam/xlog.c | 9 +- src/backend/postmaster/pgstat.c | 10 +- src/backend/replication/logical/origin.c | 8 +- src/backend/replication/slot.c | 8 +- src/backend/storage/buffer/buf_init.c | 16 +-- src/backend/storage/ipc/procarray.c | 9 +- src/backend/storage/lmgr/lwlock.c | 175 ++++++++--------------- src/backend/utils/mmgr/dsa.c | 15 +- src/backend/utils/probes.d | 16 +-- src/include/access/slru.h | 1 - src/include/pgstat.h | 3 +- src/include/storage/lwlock.h | 45 ++---- 14 files changed, 112 insertions(+), 261 deletions(-) diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml index 128ee13b5f..5b58d2e84d 100644 --- a/doc/src/sgml/monitoring.sgml +++ b/doc/src/sgml/monitoring.sgml @@ -646,18 +646,11 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser - LWLockNamed: The backend is waiting for a specific named - lightweight lock. Each such lock protects a particular data - structure in shared memory. wait_event will contain - the name of the lightweight lock. - - - - - LWLockTranche: The backend is waiting for one of a - group of related lightweight locks. All locks in the group perform - a similar function; wait_event will identify the general - purpose of locks in that group. + LWLock: The backend is waiting for a lightweight lock. + Each such lock protects a particular data structure in shared memory. + wait_event will contain a name identifying the purpose + of the lightweight lock. (Some locks have specific names; others + are part of a group of locks each with a similar purpose.) @@ -825,7 +818,7 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser - LWLockNamed + LWLock ShmemIndexLock Waiting to find or allocate space in shared memory. @@ -1011,7 +1004,6 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser Waiting to read or update old snapshot control information. - LWLockTranche clog Waiting for I/O on a clog (transaction status) buffer. @@ -1279,7 +1271,7 @@ SELECT pid, wait_event_type, wait_event FROM pg_stat_activity WHERE wait_event i pid | wait_event_type | wait_event ------+-----------------+--------------- 2540 | Lock | relation - 6644 | LWLockNamed | ProcArrayLock + 6644 | LWLock | ProcArrayLock (2 rows) @@ -3347,55 +3339,49 @@ SELECT pg_stat_get_backend_pid(s.backendid) AS pid, lwlock-acquire - (char *, int, LWLockMode) + (char *, LWLockMode) Probe that fires when an LWLock has been acquired. arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. - arg2 is the requested lock mode, either exclusive or shared. + arg1 is the requested lock mode, either exclusive or shared. lwlock-release - (char *, int) + (char *) Probe that fires when an LWLock has been released (but note that any released waiters have not yet been awakened). - arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. + arg0 is the LWLock's tranche. lwlock-wait-start - (char *, int, LWLockMode) + (char *, LWLockMode) Probe that fires when an LWLock was not immediately available and a server process has begun to wait for the lock to become available. arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. - arg2 is the requested lock mode, either exclusive or shared. + arg1 is the requested lock mode, either exclusive or shared. lwlock-wait-done - (char *, int, LWLockMode) + (char *, LWLockMode) Probe that fires when a server process has been released from its wait for an LWLock (it does not actually have the lock yet). arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. - arg2 is the requested lock mode, either exclusive or shared. + arg1 is the requested lock mode, either exclusive or shared. lwlock-condacquire - (char *, int, LWLockMode) + (char *, LWLockMode) Probe that fires when an LWLock was successfully acquired when the caller specified no waiting. arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. - arg2 is the requested lock mode, either exclusive or shared. + arg1 is the requested lock mode, either exclusive or shared. lwlock-condacquire-fail - (char *, int, LWLockMode) + (char *, LWLockMode) Probe that fires when an LWLock was not successfully acquired when the caller specified no waiting. arg0 is the LWLock's tranche. - arg1 is the LWLock's offset within its tranche. - arg2 is the requested lock mode, either exclusive or shared. + arg1 is the requested lock mode, either exclusive or shared. lock-wait-start diff --git a/src/backend/access/transam/slru.c b/src/backend/access/transam/slru.c index bbae5847f2..8b95f35dba 100644 --- a/src/backend/access/transam/slru.c +++ b/src/backend/access/transam/slru.c @@ -216,9 +216,6 @@ SimpleLruInit(SlruCtl ctl, const char *name, int nslots, int nlsns, Assert(strlen(name) + 1 < SLRU_MAX_NAME_LENGTH); strlcpy(shared->lwlock_tranche_name, name, SLRU_MAX_NAME_LENGTH); shared->lwlock_tranche_id = tranche_id; - shared->lwlock_tranche.name = shared->lwlock_tranche_name; - shared->lwlock_tranche.array_base = shared->buffer_locks; - shared->lwlock_tranche.array_stride = sizeof(LWLockPadded); ptr += BUFFERALIGN(offset); for (slotno = 0; slotno < nslots; slotno++) @@ -237,7 +234,8 @@ SimpleLruInit(SlruCtl ctl, const char *name, int nslots, int nlsns, Assert(found); /* Register SLRU tranche in the main tranches array */ - LWLockRegisterTranche(shared->lwlock_tranche_id, &shared->lwlock_tranche); + LWLockRegisterTranche(shared->lwlock_tranche_id, + shared->lwlock_tranche_name); /* * Initialize the unshared control struct, including directory path. We diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c index 084401d2f2..aa9ee5a0dd 100644 --- a/src/backend/access/transam/xlog.c +++ b/src/backend/access/transam/xlog.c @@ -517,7 +517,6 @@ typedef struct XLogCtlInsert * WAL insertion locks. */ WALInsertLockPadded *WALInsertLocks; - LWLockTranche WALInsertLockTranche; } XLogCtlInsert; /* @@ -4688,7 +4687,7 @@ XLOGShmemInit(void) /* Initialize local copy of WALInsertLocks and register the tranche */ WALInsertLocks = XLogCtl->Insert.WALInsertLocks; LWLockRegisterTranche(LWTRANCHE_WAL_INSERT, - &XLogCtl->Insert.WALInsertLockTranche); + "wal_insert"); return; } memset(XLogCtl, 0, sizeof(XLogCtlData)); @@ -4711,11 +4710,7 @@ XLOGShmemInit(void) (WALInsertLockPadded *) allocptr; allocptr += sizeof(WALInsertLockPadded) * NUM_XLOGINSERT_LOCKS; - XLogCtl->Insert.WALInsertLockTranche.name = "wal_insert"; - XLogCtl->Insert.WALInsertLockTranche.array_base = WALInsertLocks; - XLogCtl->Insert.WALInsertLockTranche.array_stride = sizeof(WALInsertLockPadded); - - LWLockRegisterTranche(LWTRANCHE_WAL_INSERT, &XLogCtl->Insert.WALInsertLockTranche); + LWLockRegisterTranche(LWTRANCHE_WAL_INSERT, "wal_insert"); for (i = 0; i < NUM_XLOGINSERT_LOCKS; i++) { LWLockInitialize(&WALInsertLocks[i].l.lock, LWTRANCHE_WAL_INSERT); diff --git a/src/backend/postmaster/pgstat.c b/src/backend/postmaster/pgstat.c index c7584cb1d3..61e6a2cd33 100644 --- a/src/backend/postmaster/pgstat.c +++ b/src/backend/postmaster/pgstat.c @@ -3152,11 +3152,8 @@ pgstat_get_wait_event_type(uint32 wait_event_info) switch (classId) { - case PG_WAIT_LWLOCK_NAMED: - event_type = "LWLockNamed"; - break; - case PG_WAIT_LWLOCK_TRANCHE: - event_type = "LWLockTranche"; + case PG_WAIT_LWLOCK: + event_type = "LWLock"; break; case PG_WAIT_LOCK: event_type = "Lock"; @@ -3209,8 +3206,7 @@ pgstat_get_wait_event(uint32 wait_event_info) switch (classId) { - case PG_WAIT_LWLOCK_NAMED: - case PG_WAIT_LWLOCK_TRANCHE: + case PG_WAIT_LWLOCK: event_name = GetLWLockIdentifier(classId, eventId); break; case PG_WAIT_LOCK: diff --git a/src/backend/replication/logical/origin.c b/src/backend/replication/logical/origin.c index cc2b513236..0deec75dc2 100644 --- a/src/backend/replication/logical/origin.c +++ b/src/backend/replication/logical/origin.c @@ -143,7 +143,6 @@ typedef struct ReplicationStateOnDisk typedef struct ReplicationStateCtl { int tranche_id; - LWLockTranche tranche; ReplicationState states[FLEXIBLE_ARRAY_MEMBER]; } ReplicationStateCtl; @@ -474,11 +473,6 @@ ReplicationOriginShmemInit(void) int i; replication_states_ctl->tranche_id = LWTRANCHE_REPLICATION_ORIGIN; - replication_states_ctl->tranche.name = "replication_origin"; - replication_states_ctl->tranche.array_base = - &replication_states[0].lock; - replication_states_ctl->tranche.array_stride = - sizeof(ReplicationState); MemSet(replication_states, 0, ReplicationOriginShmemSize()); @@ -488,7 +482,7 @@ ReplicationOriginShmemInit(void) } LWLockRegisterTranche(replication_states_ctl->tranche_id, - &replication_states_ctl->tranche); + "replication_origin"); } /* --------------------------------------------------------------------------- diff --git a/src/backend/replication/slot.c b/src/backend/replication/slot.c index d8ed005e7e..cf814d11a2 100644 --- a/src/backend/replication/slot.c +++ b/src/backend/replication/slot.c @@ -98,8 +98,6 @@ ReplicationSlot *MyReplicationSlot = NULL; int max_replication_slots = 0; /* the maximum number of replication * slots */ -static LWLockTranche ReplSlotIOLWLockTranche; - static void ReplicationSlotDropAcquired(void); static void ReplicationSlotDropPtr(ReplicationSlot *slot); @@ -141,12 +139,8 @@ ReplicationSlotsShmemInit(void) ShmemInitStruct("ReplicationSlot Ctl", ReplicationSlotsShmemSize(), &found); - ReplSlotIOLWLockTranche.name = "replication_slot_io"; - ReplSlotIOLWLockTranche.array_base = - ((char *) ReplicationSlotCtl) + offsetof(ReplicationSlotCtlData, replication_slots) +offsetof(ReplicationSlot, io_in_progress_lock); - ReplSlotIOLWLockTranche.array_stride = sizeof(ReplicationSlot); LWLockRegisterTranche(LWTRANCHE_REPLICATION_SLOT_IO_IN_PROGRESS, - &ReplSlotIOLWLockTranche); + "replication_slot_io"); if (!found) { diff --git a/src/backend/storage/buffer/buf_init.c b/src/backend/storage/buffer/buf_init.c index a4163cf717..c507389a05 100644 --- a/src/backend/storage/buffer/buf_init.c +++ b/src/backend/storage/buffer/buf_init.c @@ -21,8 +21,6 @@ BufferDescPadded *BufferDescriptors; char *BufferBlocks; LWLockMinimallyPadded *BufferIOLWLockArray = NULL; -LWLockTranche BufferIOLWLockTranche; -LWLockTranche BufferContentLWLockTranche; WritebackContext BackendWritebackContext; CkptSortItem *CkptBufferIds; @@ -90,18 +88,8 @@ InitBufferPool(void) NBuffers * (Size) sizeof(LWLockMinimallyPadded), &foundIOLocks); - BufferIOLWLockTranche.name = "buffer_io"; - BufferIOLWLockTranche.array_base = BufferIOLWLockArray; - BufferIOLWLockTranche.array_stride = sizeof(LWLockMinimallyPadded); - LWLockRegisterTranche(LWTRANCHE_BUFFER_IO_IN_PROGRESS, - &BufferIOLWLockTranche); - - BufferContentLWLockTranche.name = "buffer_content"; - BufferContentLWLockTranche.array_base = - ((char *) BufferDescriptors) + offsetof(BufferDesc, content_lock); - BufferContentLWLockTranche.array_stride = sizeof(BufferDescPadded); - LWLockRegisterTranche(LWTRANCHE_BUFFER_CONTENT, - &BufferContentLWLockTranche); + LWLockRegisterTranche(LWTRANCHE_BUFFER_IO_IN_PROGRESS, "buffer_io"); + LWLockRegisterTranche(LWTRANCHE_BUFFER_CONTENT, "buffer_content"); /* * The array used to sort to-be-checkpointed buffer ids is located in diff --git a/src/backend/storage/ipc/procarray.c b/src/backend/storage/ipc/procarray.c index bf38470f01..0f637556cc 100644 --- a/src/backend/storage/ipc/procarray.c +++ b/src/backend/storage/ipc/procarray.c @@ -106,9 +106,6 @@ static TransactionId *KnownAssignedXids; static bool *KnownAssignedXidsValid; static TransactionId latestObservedXid = InvalidTransactionId; -/* LWLock tranche for backend locks */ -static LWLockTranche ProcLWLockTranche; - /* * If we're in STANDBY_SNAPSHOT_PENDING state, standbySnapshotPendingXmin is * the highest xid that might still be running that we don't have in @@ -266,11 +263,7 @@ CreateSharedProcArray(void) } /* Register and initialize fields of ProcLWLockTranche */ - ProcLWLockTranche.name = "proc"; - ProcLWLockTranche.array_base = (char *) (ProcGlobal->allProcs) + - offsetof(PGPROC, backendLock); - ProcLWLockTranche.array_stride = sizeof(PGPROC); - LWLockRegisterTranche(LWTRANCHE_PROC, &ProcLWLockTranche); + LWLockRegisterTranche(LWTRANCHE_PROC, "proc"); } /* diff --git a/src/backend/storage/lmgr/lwlock.c b/src/backend/storage/lmgr/lwlock.c index 03c4c7825e..4b381e4e0c 100644 --- a/src/backend/storage/lmgr/lwlock.c +++ b/src/backend/storage/lmgr/lwlock.c @@ -108,18 +108,14 @@ extern slock_t *ShmemLock; #define LW_SHARED_MASK ((uint32) ((1 << 24)-1)) /* - * This is indexed by tranche ID and stores metadata for all tranches known + * This is indexed by tranche ID and stores the names of all tranches known * to the current backend. */ -static LWLockTranche **LWLockTrancheArray = NULL; +static char **LWLockTrancheArray = NULL; static int LWLockTranchesAllocated = 0; #define T_NAME(lock) \ - (LWLockTrancheArray[(lock)->tranche]->name) -#define T_ID(lock) \ - ((int) ((((char *) lock) - \ - ((char *) LWLockTrancheArray[(lock)->tranche]->array_base)) / \ - LWLockTrancheArray[(lock)->tranche]->array_stride)) + (LWLockTrancheArray[(lock)->tranche]) /* * This points to the main array of LWLocks in shared memory. Backends inherit @@ -127,10 +123,6 @@ static int LWLockTranchesAllocated = 0; * where we have special measures to pass it down). */ LWLockPadded *MainLWLockArray = NULL; -static LWLockTranche MainLWLockTranche; -static LWLockTranche BufMappingLWLockTranche; -static LWLockTranche LockManagerLWLockTranche; -static LWLockTranche PredicateLockManagerLWLockTranche; /* * We use this structure to keep track of locked LWLocks for release @@ -175,7 +167,7 @@ static inline void LWLockReportWaitEnd(void); typedef struct lwlock_stats_key { int tranche; - int instance; + void *instance; } lwlock_stats_key; typedef struct lwlock_stats @@ -202,32 +194,18 @@ PRINT_LWDEBUG(const char *where, LWLock *lock, LWLockMode mode) if (Trace_lwlocks) { uint32 state = pg_atomic_read_u32(&lock->state); - int id = T_ID(lock); - - if (lock->tranche == 0 && id < NUM_INDIVIDUAL_LWLOCKS) - ereport(LOG, - (errhidestmt(true), - errhidecontext(true), - errmsg_internal("%d: %s(%s): excl %u shared %u haswaiters %u waiters %u rOK %d", - MyProcPid, - where, MainLWLockNames[id], - (state & LW_VAL_EXCLUSIVE) != 0, - state & LW_SHARED_MASK, - (state & LW_FLAG_HAS_WAITERS) != 0, - pg_atomic_read_u32(&lock->nwaiters), - (state & LW_FLAG_RELEASE_OK) != 0))); - else - ereport(LOG, - (errhidestmt(true), - errhidecontext(true), - errmsg_internal("%d: %s(%s %d): excl %u shared %u haswaiters %u waiters %u rOK %d", - MyProcPid, - where, T_NAME(lock), id, - (state & LW_VAL_EXCLUSIVE) != 0, - state & LW_SHARED_MASK, - (state & LW_FLAG_HAS_WAITERS) != 0, - pg_atomic_read_u32(&lock->nwaiters), - (state & LW_FLAG_RELEASE_OK) != 0))); + + ereport(LOG, + (errhidestmt(true), + errhidecontext(true), + errmsg_internal("%d: %s(%s %p): excl %u shared %u haswaiters %u waiters %u rOK %d", + MyProcPid, + where, T_NAME(lock), lock, + (state & LW_VAL_EXCLUSIVE) != 0, + state & LW_SHARED_MASK, + (state & LW_FLAG_HAS_WAITERS) != 0, + pg_atomic_read_u32(&lock->nwaiters), + (state & LW_FLAG_RELEASE_OK) != 0))); } } @@ -237,20 +215,11 @@ LOG_LWDEBUG(const char *where, LWLock *lock, const char *msg) /* hide statement & context here, otherwise the log is just too verbose */ if (Trace_lwlocks) { - int id = T_ID(lock); - - if (lock->tranche == 0 && id < NUM_INDIVIDUAL_LWLOCKS) - ereport(LOG, - (errhidestmt(true), - errhidecontext(true), - errmsg_internal("%s(%s): %s", where, - MainLWLockNames[id], msg))); - else - ereport(LOG, - (errhidestmt(true), - errhidecontext(true), - errmsg_internal("%s(%s %d): %s", where, - T_NAME(lock), id, msg))); + ereport(LOG, + (errhidestmt(true), + errhidecontext(true), + errmsg_internal("%s(%s %p): %s", where, + T_NAME(lock), lock, msg))); } } @@ -315,8 +284,8 @@ print_lwlock_stats(int code, Datum arg) while ((lwstats = (lwlock_stats *) hash_seq_search(&scan)) != NULL) { fprintf(stderr, - "PID %d lwlock %s %d: shacq %u exacq %u blk %u spindelay %u dequeue self %u\n", - MyProcPid, LWLockTrancheArray[lwstats->key.tranche]->name, + "PID %d lwlock %s %p: shacq %u exacq %u blk %u spindelay %u dequeue self %u\n", + MyProcPid, LWLockTrancheArray[lwstats->key.tranche], lwstats->key.instance, lwstats->sh_acquire_count, lwstats->ex_acquire_count, lwstats->block_count, lwstats->spin_delay_count, lwstats->dequeue_self_count); @@ -342,7 +311,7 @@ get_lwlock_stats_entry(LWLock *lock) /* Fetch or create the entry. */ key.tranche = lock->tranche; - key.instance = T_ID(lock); + key.instance = lock; lwstats = hash_search(lwlock_stats_htab, &key, HASH_ENTER, &found); if (!found) { @@ -464,7 +433,7 @@ InitializeLWLocks(void) /* Initialize all individual LWLocks in main array */ for (id = 0, lock = MainLWLockArray; id < NUM_INDIVIDUAL_LWLOCKS; id++, lock++) - LWLockInitialize(&lock->lock, LWTRANCHE_MAIN); + LWLockInitialize(&lock->lock, id); /* Initialize buffer mapping LWLocks in main array */ lock = MainLWLockArray + NUM_INDIVIDUAL_LWLOCKS; @@ -506,10 +475,8 @@ InitializeLWLocks(void) name = trancheNames; trancheNames += strlen(request->tranche_name) + 1; strcpy(name, request->tranche_name); - tranche->lwLockTranche.name = name; tranche->trancheId = LWLockNewTrancheId(); - tranche->lwLockTranche.array_base = lock; - tranche->lwLockTranche.array_stride = sizeof(LWLockPadded); + tranche->trancheName = name; for (j = 0; j < request->num_lwlocks; j++, lock++) LWLockInitialize(&lock->lock, tranche->trancheId); @@ -527,39 +494,25 @@ RegisterLWLockTranches(void) if (LWLockTrancheArray == NULL) { - LWLockTranchesAllocated = 32; - LWLockTrancheArray = (LWLockTranche **) + LWLockTranchesAllocated = 64; + LWLockTrancheArray = (char **) MemoryContextAllocZero(TopMemoryContext, - LWLockTranchesAllocated * sizeof(LWLockTranche *)); + LWLockTranchesAllocated * sizeof(char *)); Assert(LWLockTranchesAllocated >= LWTRANCHE_FIRST_USER_DEFINED); } - MainLWLockTranche.name = "main"; - MainLWLockTranche.array_base = MainLWLockArray; - MainLWLockTranche.array_stride = sizeof(LWLockPadded); - LWLockRegisterTranche(LWTRANCHE_MAIN, &MainLWLockTranche); + for (i = 0; i < NUM_INDIVIDUAL_LWLOCKS; ++i) + LWLockRegisterTranche(i, MainLWLockNames[i]); - BufMappingLWLockTranche.name = "buffer_mapping"; - BufMappingLWLockTranche.array_base = MainLWLockArray + NUM_INDIVIDUAL_LWLOCKS; - BufMappingLWLockTranche.array_stride = sizeof(LWLockPadded); - LWLockRegisterTranche(LWTRANCHE_BUFFER_MAPPING, &BufMappingLWLockTranche); - - LockManagerLWLockTranche.name = "lock_manager"; - LockManagerLWLockTranche.array_base = MainLWLockArray + NUM_INDIVIDUAL_LWLOCKS + - NUM_BUFFER_PARTITIONS; - LockManagerLWLockTranche.array_stride = sizeof(LWLockPadded); - LWLockRegisterTranche(LWTRANCHE_LOCK_MANAGER, &LockManagerLWLockTranche); - - PredicateLockManagerLWLockTranche.name = "predicate_lock_manager"; - PredicateLockManagerLWLockTranche.array_base = MainLWLockArray + NUM_INDIVIDUAL_LWLOCKS + - NUM_BUFFER_PARTITIONS + NUM_LOCK_PARTITIONS; - PredicateLockManagerLWLockTranche.array_stride = sizeof(LWLockPadded); - LWLockRegisterTranche(LWTRANCHE_PREDICATE_LOCK_MANAGER, &PredicateLockManagerLWLockTranche); + LWLockRegisterTranche(LWTRANCHE_BUFFER_MAPPING, "buffer_mapping"); + LWLockRegisterTranche(LWTRANCHE_LOCK_MANAGER, "lock_manager"); + LWLockRegisterTranche(LWTRANCHE_PREDICATE_LOCK_MANAGER, + "predicate_lock_manager"); /* Register named tranches. */ for (i = 0; i < NamedLWLockTrancheRequests; i++) LWLockRegisterTranche(NamedLWLockTrancheArray[i].trancheId, - &NamedLWLockTrancheArray[i].lwLockTranche); + NamedLWLockTrancheArray[i].trancheName); } /* @@ -633,7 +586,7 @@ LWLockNewTrancheId(void) * (TopMemoryContext, static variable, or similar). */ void -LWLockRegisterTranche(int tranche_id, LWLockTranche *tranche) +LWLockRegisterTranche(int tranche_id, char *tranche_name) { Assert(LWLockTrancheArray != NULL); @@ -645,15 +598,14 @@ LWLockRegisterTranche(int tranche_id, LWLockTranche *tranche) while (i <= tranche_id) i *= 2; - LWLockTrancheArray = (LWLockTranche **) - repalloc(LWLockTrancheArray, - i * sizeof(LWLockTranche *)); + LWLockTrancheArray = (char **) + repalloc(LWLockTrancheArray, i * sizeof(char *)); LWLockTranchesAllocated = i; while (j < LWLockTranchesAllocated) LWLockTrancheArray[j++] = NULL; } - LWLockTrancheArray[tranche_id] = tranche; + LWLockTrancheArray[tranche_id] = tranche_name; } /* @@ -729,12 +681,7 @@ LWLockInitialize(LWLock *lock, int tranche_id) static inline void LWLockReportWaitStart(LWLock *lock) { - int lockId = T_ID(lock); - - if (lock->tranche == 0) - pgstat_report_wait_start(PG_WAIT_LWLOCK_NAMED | (uint16) lockId); - else - pgstat_report_wait_start(PG_WAIT_LWLOCK_TRANCHE | lock->tranche); + pgstat_report_wait_start(PG_WAIT_LWLOCK | lock->tranche); } /* @@ -752,10 +699,7 @@ LWLockReportWaitEnd(void) const char * GetLWLockIdentifier(uint32 classId, uint16 eventId) { - if (classId == PG_WAIT_LWLOCK_NAMED) - return MainLWLockNames[eventId]; - - Assert(classId == PG_WAIT_LWLOCK_TRANCHE); + Assert(classId == PG_WAIT_LWLOCK); /* * It is quite possible that user has registered tranche in one of the @@ -763,10 +707,10 @@ GetLWLockIdentifier(uint32 classId, uint16 eventId) * all of them, so we can't assume the tranche is registered here. */ if (eventId >= LWLockTranchesAllocated || - LWLockTrancheArray[eventId]->name == NULL) + LWLockTrancheArray[eventId] == NULL) return "extension"; - return LWLockTrancheArray[eventId]->name; + return LWLockTrancheArray[eventId]; } /* @@ -1279,7 +1223,7 @@ LWLockAcquire(LWLock *lock, LWLockMode mode) #endif LWLockReportWaitStart(lock); - TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), mode); for (;;) { @@ -1301,7 +1245,7 @@ LWLockAcquire(LWLock *lock, LWLockMode mode) } #endif - TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), mode); LWLockReportWaitEnd(); LOG_LWDEBUG("LWLockAcquire", lock, "awakened"); @@ -1310,7 +1254,7 @@ LWLockAcquire(LWLock *lock, LWLockMode mode) result = false; } - TRACE_POSTGRESQL_LWLOCK_ACQUIRE(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_ACQUIRE(T_NAME(lock), mode); /* Add lock to list of locks held by this backend */ held_lwlocks[num_held_lwlocks].lock = lock; @@ -1361,14 +1305,14 @@ LWLockConditionalAcquire(LWLock *lock, LWLockMode mode) RESUME_INTERRUPTS(); LOG_LWDEBUG("LWLockConditionalAcquire", lock, "failed"); - TRACE_POSTGRESQL_LWLOCK_CONDACQUIRE_FAIL(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_CONDACQUIRE_FAIL(T_NAME(lock), mode); } else { /* Add lock to list of locks held by this backend */ held_lwlocks[num_held_lwlocks].lock = lock; held_lwlocks[num_held_lwlocks++].mode = mode; - TRACE_POSTGRESQL_LWLOCK_CONDACQUIRE(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_CONDACQUIRE(T_NAME(lock), mode); } return !mustwait; } @@ -1440,7 +1384,7 @@ LWLockAcquireOrWait(LWLock *lock, LWLockMode mode) #endif LWLockReportWaitStart(lock); - TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), mode); for (;;) { @@ -1458,7 +1402,7 @@ LWLockAcquireOrWait(LWLock *lock, LWLockMode mode) Assert(nwaiters < MAX_BACKENDS); } #endif - TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), mode); LWLockReportWaitEnd(); LOG_LWDEBUG("LWLockAcquireOrWait", lock, "awakened"); @@ -1488,8 +1432,7 @@ LWLockAcquireOrWait(LWLock *lock, LWLockMode mode) /* Failed to get lock, so release interrupt holdoff */ RESUME_INTERRUPTS(); LOG_LWDEBUG("LWLockAcquireOrWait", lock, "failed"); - TRACE_POSTGRESQL_LWLOCK_ACQUIRE_OR_WAIT_FAIL(T_NAME(lock), T_ID(lock), - mode); + TRACE_POSTGRESQL_LWLOCK_ACQUIRE_OR_WAIT_FAIL(T_NAME(lock), mode); } else { @@ -1497,7 +1440,7 @@ LWLockAcquireOrWait(LWLock *lock, LWLockMode mode) /* Add lock to list of locks held by this backend */ held_lwlocks[num_held_lwlocks].lock = lock; held_lwlocks[num_held_lwlocks++].mode = mode; - TRACE_POSTGRESQL_LWLOCK_ACQUIRE_OR_WAIT(T_NAME(lock), T_ID(lock), mode); + TRACE_POSTGRESQL_LWLOCK_ACQUIRE_OR_WAIT(T_NAME(lock), mode); } return !mustwait; @@ -1657,8 +1600,7 @@ LWLockWaitForVar(LWLock *lock, uint64 *valptr, uint64 oldval, uint64 *newval) #endif LWLockReportWaitStart(lock); - TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), T_ID(lock), - LW_EXCLUSIVE); + TRACE_POSTGRESQL_LWLOCK_WAIT_START(T_NAME(lock), LW_EXCLUSIVE); for (;;) { @@ -1677,8 +1619,7 @@ LWLockWaitForVar(LWLock *lock, uint64 *valptr, uint64 oldval, uint64 *newval) } #endif - TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), T_ID(lock), - LW_EXCLUSIVE); + TRACE_POSTGRESQL_LWLOCK_WAIT_DONE(T_NAME(lock), LW_EXCLUSIVE); LWLockReportWaitEnd(); LOG_LWDEBUG("LWLockWaitForVar", lock, "awakened"); @@ -1686,7 +1627,7 @@ LWLockWaitForVar(LWLock *lock, uint64 *valptr, uint64 oldval, uint64 *newval) /* Now loop back and check the status of the lock again. */ } - TRACE_POSTGRESQL_LWLOCK_ACQUIRE(T_NAME(lock), T_ID(lock), LW_EXCLUSIVE); + TRACE_POSTGRESQL_LWLOCK_ACQUIRE(T_NAME(lock), LW_EXCLUSIVE); /* * Fix the process wait semaphore's count for any absorbed wakeups. @@ -1784,7 +1725,7 @@ LWLockRelease(LWLock *lock) break; if (i < 0) - elog(ERROR, "lock %s %d is not held", T_NAME(lock), T_ID(lock)); + elog(ERROR, "lock %s is not held", T_NAME(lock)); mode = held_lwlocks[i].mode; @@ -1829,7 +1770,7 @@ LWLockRelease(LWLock *lock) LWLockWakeup(lock); } - TRACE_POSTGRESQL_LWLOCK_RELEASE(T_NAME(lock), T_ID(lock)); + TRACE_POSTGRESQL_LWLOCK_RELEASE(T_NAME(lock)); /* * Now okay to allow cancel/die interrupts. diff --git a/src/backend/utils/mmgr/dsa.c b/src/backend/utils/mmgr/dsa.c index abe3f1a63d..988a2970b2 100644 --- a/src/backend/utils/mmgr/dsa.c +++ b/src/backend/utils/mmgr/dsa.c @@ -362,9 +362,6 @@ struct dsa_area /* Pointer to the control object in shared memory. */ dsa_area_control *control; - /* The lock tranche for this process. */ - LWLockTranche lwlock_tranche; - /* Has the mapping been pinned? */ bool mapping_pinned; @@ -1207,10 +1204,8 @@ create_internal(void *place, size_t size, area->mapping_pinned = false; memset(area->segment_maps, 0, sizeof(dsa_segment_map) * DSA_MAX_SEGMENTS); area->high_segment_index = 0; - area->lwlock_tranche.array_base = &area->control->pools[0]; - area->lwlock_tranche.array_stride = sizeof(dsa_area_pool); - area->lwlock_tranche.name = control->lwlock_tranche_name; - LWLockRegisterTranche(control->lwlock_tranche_id, &area->lwlock_tranche); + LWLockRegisterTranche(control->lwlock_tranche_id, + control->lwlock_tranche_name); LWLockInitialize(&control->lock, control->lwlock_tranche_id); for (i = 0; i < DSA_NUM_SIZE_CLASSES; ++i) LWLockInitialize(DSA_SCLASS_LOCK(area, i), @@ -1267,10 +1262,8 @@ attach_internal(void *place, dsm_segment *segment, dsa_handle handle) memset(&area->segment_maps[0], 0, sizeof(dsa_segment_map) * DSA_MAX_SEGMENTS); area->high_segment_index = 0; - area->lwlock_tranche.array_base = &area->control->pools[0]; - area->lwlock_tranche.array_stride = sizeof(dsa_area_pool); - area->lwlock_tranche.name = control->lwlock_tranche_name; - LWLockRegisterTranche(control->lwlock_tranche_id, &area->lwlock_tranche); + LWLockRegisterTranche(control->lwlock_tranche_id, + control->lwlock_tranche_name); /* Set up the segment map for this process's mapping. */ segment_map = &area->segment_maps[0]; diff --git a/src/backend/utils/probes.d b/src/backend/utils/probes.d index 2f92dfa9ad..adcebe2e05 100644 --- a/src/backend/utils/probes.d +++ b/src/backend/utils/probes.d @@ -28,14 +28,14 @@ provider postgresql { probe transaction__commit(LocalTransactionId); probe transaction__abort(LocalTransactionId); - probe lwlock__acquire(const char *, int, LWLockMode); - probe lwlock__release(const char *, int); - probe lwlock__wait__start(const char *, int, LWLockMode); - probe lwlock__wait__done(const char *, int, LWLockMode); - probe lwlock__condacquire(const char *, int, LWLockMode); - probe lwlock__condacquire__fail(const char *, int, LWLockMode); - probe lwlock__acquire__or__wait(const char *, int, LWLockMode); - probe lwlock__acquire__or__wait__fail(const char *, int, LWLockMode); + probe lwlock__acquire(const char *, LWLockMode); + probe lwlock__release(const char *); + probe lwlock__wait__start(const char *, LWLockMode); + probe lwlock__wait__done(const char *, LWLockMode); + probe lwlock__condacquire(const char *, LWLockMode); + probe lwlock__condacquire__fail(const char *, LWLockMode); + probe lwlock__acquire__or__wait(const char *, LWLockMode); + probe lwlock__acquire__or__wait__fail(const char *, LWLockMode); probe lock__wait__start(unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, LOCKMODE); probe lock__wait__done(unsigned int, unsigned int, unsigned int, unsigned int, unsigned int, LOCKMODE); diff --git a/src/include/access/slru.h b/src/include/access/slru.h index 5fcebc52fb..ca8ab7ba52 100644 --- a/src/include/access/slru.h +++ b/src/include/access/slru.h @@ -104,7 +104,6 @@ typedef struct SlruSharedData /* LWLocks */ int lwlock_tranche_id; - LWLockTranche lwlock_tranche; char lwlock_tranche_name[SLRU_MAX_NAME_LENGTH]; LWLockPadded *buffer_locks; } SlruSharedData; diff --git a/src/include/pgstat.h b/src/include/pgstat.h index 152ff06208..282f8aeccd 100644 --- a/src/include/pgstat.h +++ b/src/include/pgstat.h @@ -715,8 +715,7 @@ typedef enum BackendState * Wait Classes * ---------- */ -#define PG_WAIT_LWLOCK_NAMED 0x01000000U -#define PG_WAIT_LWLOCK_TRANCHE 0x02000000U +#define PG_WAIT_LWLOCK 0x01000000U #define PG_WAIT_LOCK 0x03000000U #define PG_WAIT_BUFFER_PIN 0x04000000U #define PG_WAIT_ACTIVITY 0x05000000U diff --git a/src/include/storage/lwlock.h b/src/include/storage/lwlock.h index 9a2d86975c..db1c687e21 100644 --- a/src/include/storage/lwlock.h +++ b/src/include/storage/lwlock.h @@ -24,32 +24,6 @@ struct PGPROC; -/* - * Prior to PostgreSQL 9.4, every lightweight lock in the system was stored - * in a single array. For convenience and for compatibility with past - * releases, we still have a main array, but it's now also permissible to - * store LWLocks elsewhere in the main shared memory segment or in a dynamic - * shared memory segment. Each array of lwlocks forms a separate "tranche". - * - * It's occasionally necessary to identify a particular LWLock "by name"; e.g. - * because we wish to report the lock to dtrace. We could store a name or - * other identifying information in the lock itself, but since it's common - * to have many nearly-identical locks (e.g. one per buffer) this would end - * up wasting significant amounts of memory. Instead, each lwlock stores a - * tranche ID which tells us which array it's part of. Based on that, we can - * figure out where the lwlock lies within the array using the data structure - * shown below; the lock is then identified based on the tranche name and - * computed array index. We need the array stride because the array might not - * be an array of lwlocks, but rather some larger data structure that includes - * one or more lwlocks per element. - */ -typedef struct LWLockTranche -{ - const char *name; - void *array_base; - Size array_stride; -} LWLockTranche; - /* * Code outside of lwlock.c should not manipulate the contents of this * structure directly, but we have to declare it here to allow LWLocks to be @@ -118,8 +92,8 @@ extern char *MainLWLockNames[]; /* struct for storing named tranche information */ typedef struct NamedLWLockTranche { - LWLockTranche lwLockTranche; int trancheId; + char *trancheName; } NamedLWLockTranche; extern PGDLLIMPORT NamedLWLockTranche *NamedLWLockTrancheArray; @@ -199,9 +173,9 @@ extern LWLockPadded *GetNamedLWLockTranche(const char *tranche_name); * There is another, more flexible method of obtaining lwlocks. First, call * LWLockNewTrancheId just once to obtain a tranche ID; this allocates from * a shared counter. Next, each individual process using the tranche should - * call LWLockRegisterTranche() to associate that tranche ID with appropriate - * metadata. Finally, LWLockInitialize should be called just once per lwlock, - * passing the tranche ID as an argument. + * call LWLockRegisterTranche() to associate that tranche ID with a name. + * Finally, LWLockInitialize should be called just once per lwlock, passing + * the tranche ID as an argument. * * It may seem strange that each process using the tranche must register it * separately, but dynamic shared memory segments aren't guaranteed to be @@ -209,17 +183,18 @@ extern LWLockPadded *GetNamedLWLockTranche(const char *tranche_name); * registration in the main shared memory segment wouldn't work for that case. */ extern int LWLockNewTrancheId(void); -extern void LWLockRegisterTranche(int tranche_id, LWLockTranche *tranche); +extern void LWLockRegisterTranche(int tranche_id, char *tranche_name); extern void LWLockInitialize(LWLock *lock, int tranche_id); /* - * We reserve a few predefined tranche IDs. A call to LWLockNewTrancheId - * will never return a value less than LWTRANCHE_FIRST_USER_DEFINED. + * Every tranche ID less than NUM_INDIVIDUAL_LWLOCKS is reserved; also, + * we reserve additional tranche IDs for builtin tranches not included in + * the set of individual LWLocks. A call to LWLockNewTrancheId will never + * return a value less than LWTRANCHE_FIRST_USER_DEFINED. */ typedef enum BuiltinTrancheIds { - LWTRANCHE_MAIN, - LWTRANCHE_CLOG_BUFFERS, + LWTRANCHE_CLOG_BUFFERS = NUM_INDIVIDUAL_LWLOCKS, LWTRANCHE_COMMITTS_BUFFERS, LWTRANCHE_SUBTRANS_BUFFERS, LWTRANCHE_MXACTOFFSET_BUFFERS, -- 2.40.0