/* ----------------------------------------------------------------
* ExecReScanRecursiveUnion
*
* Rescans the relation.
* ----------------------------------------------------------------
*/
void
ExecReScanRecursiveUnion(RecursiveUnionState *node)
{
PlanState *outerPlan = outerPlanState(node);
PlanState *innerPlan = innerPlanState(node);
RecursiveUnion *plan = (RecursiveUnion *) node->ps.plan;
/*
* Set recursive term's chgParam to tell it that we'll modify the working
* table and therefore it has to rescan.
*/
innerPlan->chgParam = bms_add_member(innerPlan->chgParam, plan->wtParam);
/*
* if chgParam of subnode is not null then plan will be re-scanned by
* first ExecProcNode. Because of above, we only have to do this to the
* non-recursive term.
*/
if (outerPlan->chgParam == NULL)
ExecReScan(outerPlan);
/* Release any hashtable storage */
if (node->tableContext)
MemoryContextResetAndDeleteChildren(node->tableContext);
/* And rebuild empty hashtable if needed */
if (plan->numCols > 0)
build_hash_table(node);
/* reset processing state */
node->recursing = false;
node->intermediate_empty = true;
tuplestore_clear(node->working_table);
tuplestore_clear(node->intermediate_table);
}
/*
* Release current scan keys, if any.
*/
void
ginFreeScanKeys(GinScanOpaque so)
{
uint32 i;
if (so->keys == NULL)
return;
for (i = 0; i < so->totalentries; i++)
{
GinScanEntry entry = so->entries[i];
if (entry->buffer != InvalidBuffer)
ReleaseBuffer(entry->buffer);
if (entry->list)
pfree(entry->list);
if (entry->matchIterator)
tbm_end_iterate(entry->matchIterator);
if (entry->matchBitmap)
tbm_free(entry->matchBitmap);
}
MemoryContextResetAndDeleteChildren(so->keyCtx);
so->keys = NULL;
so->nkeys = 0;
so->entries = NULL;
so->totalentries = 0;
}
void
ExecReScanSetOp(SetOpState *node)
{
ExecClearTuple(node->ps.ps_ResultTupleSlot);
node->setop_done = false;
node->numOutput = 0;
if (((SetOp *) node->ps.plan)->strategy == SETOP_HASHED)
{
/*
* In the hashed case, if we haven't yet built the hash table then we
* can just return; nothing done yet, so nothing to undo. If subnode's
* chgParam is not NULL then it will be re-scanned by ExecProcNode,
* else no reason to re-scan it at all.
*/
if (!node->table_filled)
return;
/*
* If we do have the hash table and the subplan does not have any
* parameter changes, then we can just rescan the existing hash table;
* no need to build it again.
*/
if (node->ps.lefttree->chgParam == NULL)
{
ResetTupleHashIterator(node->hashtable, &node->hashiter);
return;
}
}
/* Release first tuple of group, if we have made a copy */
if (node->grp_firstTuple != NULL)
{
heap_freetuple(node->grp_firstTuple);
node->grp_firstTuple = NULL;
}
/* Release any hashtable storage */
if (node->tableContext)
MemoryContextResetAndDeleteChildren(node->tableContext);
/* And rebuild empty hashtable if needed */
if (((SetOp *) node->ps.plan)->strategy == SETOP_HASHED)
{
build_hash_table(node);
node->table_filled = false;
}
/*
* if chgParam of subnode is not null then plan will be re-scanned by
* first ExecProcNode.
*/
if (node->ps.lefttree->chgParam == NULL)
ExecReScan(node->ps.lefttree);
}
/*
* Flush all cache entries when pg_event_trigger is updated.
*
* This should be rare enough that we don't need to be very granular about
* it, so we just blow away everything, which also avoids the possibility of
* memory leaks.
*/
static void
InvalidateEventCacheCallback(Datum arg, int cacheid, uint32 hashvalue)
{
/*
* If the cache isn't valid, then there might be a rebuild in progress,
* so we can't immediately blow it away. But it's advantageous to do
* this when possible, so as to immediately free memory.
*/
if (EventTriggerCacheState == ETCS_VALID)
{
MemoryContextResetAndDeleteChildren(EventTriggerCacheContext);
EventTriggerCache = NULL;
}
/* Mark cache for rebuild. */
EventTriggerCacheState = ETCS_NEEDS_REBUILD;
}
static void
resetTemporaryContext( twopoEssentials *essentials )
{
# ifdef TWOPO_CACHE_PLANS
int i;
# endif
Assert( essentials != NULL );
if ( ! essentials->ctx )
return;
# ifdef TWOPO_DEBUG
# ifdef TWOPO_CACHE_PLANS
if( twopo_cache_plans ){
fprintf(stderr, "TwoPO DEBUG: resetting temporary memory context.\n");
//MemoryContextStats(essentials->ctx->mycontext);
}
# endif
# endif
essentials->root->join_rel_list =
list_truncate(essentials->root->join_rel_list,
essentials->ctx->savelength);
essentials->root->join_rel_hash = NULL;
# ifdef TWOPO_CACHE_PLANS
/*
* Cleaning parent nodes in nodeList deleted by
* MemoryContextResetAndDeleteChildren()
*/
for( i=0; i<essentials->numNodes; i++ ){
essentials->nodeList[i].parents = NULL;
}
# endif
MemoryContextResetAndDeleteChildren(essentials->ctx->mycontext);
}
/*
* Rebuild the event trigger cache.
*/
static void
BuildEventTriggerCache(void)
{
HASHCTL ctl;
HTAB *cache;
MemoryContext oldcontext;
Relation rel;
Relation irel;
SysScanDesc scan;
if (EventTriggerCacheContext != NULL)
{
/*
* Free up any memory already allocated in EventTriggerCacheContext.
* This can happen either because a previous rebuild failed, or
* because an invalidation happened before the rebuild was complete.
*/
MemoryContextResetAndDeleteChildren(EventTriggerCacheContext);
}
else
{
/*
* This is our first time attempting to build the cache, so we need
* to set up the memory context and register a syscache callback to
* capture future invalidation events.
*/
if (CacheMemoryContext == NULL)
CreateCacheMemoryContext();
EventTriggerCacheContext =
AllocSetContextCreate(CacheMemoryContext,
"EventTriggerCache",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
CacheRegisterSyscacheCallback(EVENTTRIGGEROID,
InvalidateEventCacheCallback,
(Datum) 0);
}
/* Switch to correct memory context. */
oldcontext = MemoryContextSwitchTo(EventTriggerCacheContext);
/* Prevent the memory context from being nuked while we're rebuilding. */
EventTriggerCacheState = ETCS_REBUILD_STARTED;
/* Create new hash table. */
MemSet(&ctl, 0, sizeof(ctl));
ctl.keysize = sizeof(EventTriggerEvent);
ctl.entrysize = sizeof(EventTriggerCacheEntry);
ctl.hash = tag_hash;
ctl.hcxt = EventTriggerCacheContext;
cache = hash_create("Event Trigger Cache", 32, &ctl,
HASH_ELEM | HASH_FUNCTION | HASH_CONTEXT);
/*
* Prepare to scan pg_event_trigger in name order. We use an MVCC
* snapshot to avoid getting inconsistent results if the table is
* being concurrently updated.
*/
rel = relation_open(EventTriggerRelationId, AccessShareLock);
irel = index_open(EventTriggerNameIndexId, AccessShareLock);
scan = systable_beginscan_ordered(rel, irel, GetLatestSnapshot(), 0, NULL);
/*
* Build a cache item for each pg_event_trigger tuple, and append each
* one to the appropriate cache entry.
*/
for (;;)
{
HeapTuple tup;
Form_pg_event_trigger form;
char *evtevent;
EventTriggerEvent event;
EventTriggerCacheItem *item;
Datum evttags;
bool evttags_isnull;
EventTriggerCacheEntry *entry;
bool found;
/* Get next tuple. */
tup = systable_getnext_ordered(scan, ForwardScanDirection);
if (!HeapTupleIsValid(tup))
break;
/* Skip trigger if disabled. */
form = (Form_pg_event_trigger) GETSTRUCT(tup);
if (form->evtenabled == TRIGGER_DISABLED)
continue;
/* Decode event name. */
evtevent = NameStr(form->evtevent);
if (strcmp(evtevent, "ddl_command_start") == 0)
event = EVT_DDLCommandStart;
else
continue;
/* Allocate new cache item. */
item = palloc0(sizeof(EventTriggerCacheItem));
item->fnoid = form->evtfoid;
item->enabled = form->evtenabled;
/* Decode and sort tags array. */
evttags = heap_getattr(tup, Anum_pg_event_trigger_evttags,
RelationGetDescr(rel), &evttags_isnull);
if (!evttags_isnull)
{
item->ntags = DecodeTextArrayToCString(evttags, &item->tag);
qsort(item->tag, item->ntags, sizeof(char *), pg_qsort_strcmp);
}
/* Add to cache entry. */
entry = hash_search(cache, &event, HASH_ENTER, &found);
if (found)
entry->triggerlist = lappend(entry->triggerlist, item);
else
entry->triggerlist = list_make1(item);
}
/* Done with pg_event_trigger scan. */
systable_endscan_ordered(scan);
index_close(irel, AccessShareLock);
relation_close(rel, AccessShareLock);
/* Restore previous memory context. */
MemoryContextSwitchTo(oldcontext);
/* Install new cache. */
EventTriggerCache = cache;
/*
* If the cache has been invalidated since we entered this routine, we
* still use and return the cache we just finished constructing, to avoid
* infinite loops, but we leave the cache marked stale so that we'll
* rebuild it again on next access. Otherwise, we mark the cache valid.
*/
if (EventTriggerCacheState == ETCS_REBUILD_STARTED)
EventTriggerCacheState = ETCS_VALID;
}
/*
* Execute the index scan.
*
* This works by reading index TIDs from the revmap, and obtaining the index
* tuples pointed to by them; the summary values in the index tuples are
* compared to the scan keys. We return into the TID bitmap all the pages in
* ranges corresponding to index tuples that match the scan keys.
*
* If a TID from the revmap is read as InvalidTID, we know that range is
* unsummarized. Pages in those ranges need to be returned regardless of scan
* keys.
*/
int64
bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
{
Relation idxRel = scan->indexRelation;
Buffer buf = InvalidBuffer;
BrinDesc *bdesc;
Oid heapOid;
Relation heapRel;
BrinOpaque *opaque;
BlockNumber nblocks;
BlockNumber heapBlk;
int totalpages = 0;
FmgrInfo *consistentFn;
MemoryContext oldcxt;
MemoryContext perRangeCxt;
opaque = (BrinOpaque *) scan->opaque;
bdesc = opaque->bo_bdesc;
pgstat_count_index_scan(idxRel);
/*
* We need to know the size of the table so that we know how long to
* iterate on the revmap.
*/
heapOid = IndexGetRelation(RelationGetRelid(idxRel), false);
heapRel = heap_open(heapOid, AccessShareLock);
nblocks = RelationGetNumberOfBlocks(heapRel);
heap_close(heapRel, AccessShareLock);
/*
* Make room for the consistent support procedures of indexed columns. We
* don't look them up here; we do that lazily the first time we see a scan
* key reference each of them. We rely on zeroing fn_oid to InvalidOid.
*/
consistentFn = palloc0(sizeof(FmgrInfo) * bdesc->bd_tupdesc->natts);
/*
* Setup and use a per-range memory context, which is reset every time we
* loop below. This avoids having to free the tuples within the loop.
*/
perRangeCxt = AllocSetContextCreate(CurrentMemoryContext,
"bringetbitmap cxt",
ALLOCSET_DEFAULT_SIZES);
oldcxt = MemoryContextSwitchTo(perRangeCxt);
/*
* Now scan the revmap. We start by querying for heap page 0,
* incrementing by the number of pages per range; this gives us a full
* view of the table.
*/
for (heapBlk = 0; heapBlk < nblocks; heapBlk += opaque->bo_pagesPerRange)
{
bool addrange;
BrinTuple *tup;
OffsetNumber off;
Size size;
CHECK_FOR_INTERRUPTS();
MemoryContextResetAndDeleteChildren(perRangeCxt);
tup = brinGetTupleForHeapBlock(opaque->bo_rmAccess, heapBlk, &buf,
&off, &size, BUFFER_LOCK_SHARE,
scan->xs_snapshot);
if (tup)
{
tup = brin_copy_tuple(tup, size);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
/*
* For page ranges with no indexed tuple, we must return the whole
* range; otherwise, compare it to the scan keys.
*/
if (tup == NULL)
{
addrange = true;
}
else
{
BrinMemTuple *dtup;
dtup = brin_deform_tuple(bdesc, tup);
if (dtup->bt_placeholder)
{
/*
* Placeholder tuples are always returned, regardless of the
* values stored in them.
*/
addrange = true;
}
else
{
int keyno;
/*
* Compare scan keys with summary values stored for the range.
* If scan keys are matched, the page range must be added to
* the bitmap. We initially assume the range needs to be
* added; in particular this serves the case where there are
* no keys.
*/
addrange = true;
for (keyno = 0; keyno < scan->numberOfKeys; keyno++)
{
ScanKey key = &scan->keyData[keyno];
AttrNumber keyattno = key->sk_attno;
BrinValues *bval = &dtup->bt_columns[keyattno - 1];
Datum add;
/*
* The collation of the scan key must match the collation
* used in the index column (but only if the search is not
* IS NULL/ IS NOT NULL). Otherwise we shouldn't be using
* this index ...
*/
Assert((key->sk_flags & SK_ISNULL) ||
(key->sk_collation ==
bdesc->bd_tupdesc->attrs[keyattno - 1]->attcollation));
/* First time this column? look up consistent function */
if (consistentFn[keyattno - 1].fn_oid == InvalidOid)
{
FmgrInfo *tmp;
tmp = index_getprocinfo(idxRel, keyattno,
BRIN_PROCNUM_CONSISTENT);
fmgr_info_copy(&consistentFn[keyattno - 1], tmp,
CurrentMemoryContext);
}
/*
* Check whether the scan key is consistent with the page
* range values; if so, have the pages in the range added
* to the output bitmap.
*
* When there are multiple scan keys, failure to meet the
* criteria for a single one of them is enough to discard
* the range as a whole, so break out of the loop as soon
* as a false return value is obtained.
*/
add = FunctionCall3Coll(&consistentFn[keyattno - 1],
key->sk_collation,
PointerGetDatum(bdesc),
PointerGetDatum(bval),
PointerGetDatum(key));
addrange = DatumGetBool(add);
if (!addrange)
break;
}
}
}
/* add the pages in the range to the output bitmap, if needed */
if (addrange)
{
BlockNumber pageno;
for (pageno = heapBlk;
pageno <= heapBlk + opaque->bo_pagesPerRange - 1;
pageno++)
{
MemoryContextSwitchTo(oldcxt);
tbm_add_page(tbm, pageno);
totalpages++;
MemoryContextSwitchTo(perRangeCxt);
}
}
}
MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(perRangeCxt);
if (buf != InvalidBuffer)
ReleaseBuffer(buf);
/*
* XXX We have an approximation of the number of *pages* that our scan
* returns, but we don't have a precise idea of the number of heap tuples
* involved.
*/
return totalpages * 10;
}
/*
* Main entry point for walwriter process
*
* This is invoked from BootstrapMain, which has already created the basic
* execution environment, but not enabled signals yet.
*/
void
WalWriterMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext walwriter_context;
/*
* If possible, make this process a group leader, so that the postmaster
* can signal any child processes too. (walwriter probably never has any
* child processes, but for consistency we make all postmaster child
* processes do this.)
*/
#ifdef HAVE_SETSID
if (setsid() < 0)
elog(FATAL, "setsid() failed: %m");
#endif
/*
* Properly accept or ignore signals the postmaster might send us
*
* We have no particular use for SIGINT at the moment, but seems
* reasonable to treat like SIGTERM.
*/
pqsignal(SIGHUP, WalSigHupHandler); /* set flag to read config file */
pqsignal(SIGINT, WalShutdownHandler); /* request shutdown */
pqsignal(SIGTERM, WalShutdownHandler); /* request shutdown */
pqsignal(SIGQUIT, wal_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, SIG_IGN); /* reserve for ProcSignal */
pqsignal(SIGUSR2, SIG_IGN); /* not used */
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
sigdelset(&BlockSig, SIGQUIT);
/*
* Create a resource owner to keep track of our resources (not clear that
* we need this, but may as well have one).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Wal Writer");
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
walwriter_context = AllocSetContextCreate(TopMemoryContext,
"Wal Writer",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(walwriter_context);
/*
* If an exception is encountered, processing resumes here.
*
* This code is heavily based on bgwriter.c, q.v.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in walwriter, but we do have LWLocks, and perhaps buffers?
*/
LWLockReleaseAll();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_Files();
AtEOXact_HashTables(false);
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(walwriter_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(walwriter_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Loop forever
*/
for (;;)
{
long udelay;
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (!PostmasterIsAlive(true))
exit(1);
/*
* Process any requests or signals received recently.
*/
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
}
if (shutdown_requested)
{
/* Normal exit from the walwriter is here */
proc_exit(0); /* done */
}
/*
* Do what we're here for...
*/
XLogBackgroundFlush();
/*
* Delay until time to do something more, but fall out of delay
* reasonably quickly if signaled.
*/
udelay = WalWriterDelay * 1000L;
while (udelay > 999999L)
{
if (got_SIGHUP || shutdown_requested)
break;
pg_usleep(1000000L);
udelay -= 1000000L;
}
if (!(got_SIGHUP || shutdown_requested))
pg_usleep(udelay);
}
}
/*
* main entry pont of pcp worker child process
*/
void
pcp_worker_main(int port)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext PCPMemoryContext;
int authenticated = 0;
char salt[4];
int random_salt = 0;
struct timeval uptime;
char tos;
int rsize;
char *buf = NULL;
ereport(DEBUG1,
(errmsg("I am PCP worker child with pid:%d",getpid())));
/* Identify myself via ps */
init_ps_display("", "", "", "");
gettimeofday(&uptime, NULL);
srandom((unsigned int) (getpid() ^ uptime.tv_usec));
/* set up signal handlers */
signal(SIGTERM, die);
signal(SIGINT, die);
signal(SIGQUIT, die);
signal(SIGCHLD, SIG_DFL);
signal(SIGUSR2, wakeup_handler_child);
signal(SIGUSR1, SIG_IGN);
signal(SIGHUP, SIG_IGN);
signal(SIGPIPE, SIG_IGN);
signal(SIGALRM, SIG_IGN);
/* Create per loop iteration memory context */
PCPMemoryContext = AllocSetContextCreate(TopMemoryContext,
"PCP_worker_main_loop",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(TopMemoryContext);
/*
* install the call back for preparation of pcp worker child exit
*/
on_system_exit(pcp_worker_will_go_down, (Datum)NULL);
/* Initialize my backend status */
pool_initialize_private_backend_status();
/* Initialize process context */
pool_init_process_context();
pcp_frontend = pcp_open(port);
unset_nonblock(pcp_frontend->fd);
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
error_context_stack = NULL;
EmitErrorReport();
MemoryContextSwitchTo(TopMemoryContext);
FlushErrorState();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
for(;;)
{
MemoryContextSwitchTo(PCPMemoryContext);
MemoryContextResetAndDeleteChildren(PCPMemoryContext);
errno = 0;
/* read a PCP packet */
do_pcp_read(pcp_frontend, &tos, 1);
do_pcp_read(pcp_frontend, &rsize, sizeof(int));
rsize = ntohl(rsize);
if ((rsize - sizeof(int)) > 0)
{
buf = (char *)palloc(rsize - sizeof(int));
do_pcp_read(pcp_frontend, buf, rsize - sizeof(int));
}
ereport(DEBUG1,
(errmsg("received PCP packet"),
errdetail("PCP packet type of service '%c'", tos)));
if (tos == 'R') /* authentication */
{
set_ps_display("PCP: processing authentication", false);
process_authentication(pcp_frontend, buf,salt, &random_salt);
authenticated = 1;
continue;
}
if (tos == 'M') /* md5 salt */
{
set_ps_display("PCP: processing authentication", false);
send_md5salt(pcp_frontend, salt);
random_salt = 1;
continue;
}
/* is this connection authenticated? if not disconnect immediately*/
if (!authenticated)
ereport(FATAL,
(errmsg("authentication failed for new PCP connection"),
errdetail("connection not authorized")));
/* process a request */
pcp_process_command(tos, buf, rsize);
}
exit(0);
}
/*
* A tuple in the heap is being inserted. To keep a brin index up to date,
* we need to obtain the relevant index tuple and compare its stored values
* with those of the new tuple. If the tuple values are not consistent with
* the summary tuple, we need to update the index tuple.
*
* If the range is not currently summarized (i.e. the revmap returns NULL for
* it), there's nothing to do.
*/
bool
brininsert(Relation idxRel, Datum *values, bool *nulls,
ItemPointer heaptid, Relation heapRel,
IndexUniqueCheck checkUnique)
{
BlockNumber pagesPerRange;
BrinDesc *bdesc = NULL;
BrinRevmap *revmap;
Buffer buf = InvalidBuffer;
MemoryContext tupcxt = NULL;
MemoryContext oldcxt = NULL;
revmap = brinRevmapInitialize(idxRel, &pagesPerRange, NULL);
for (;;)
{
bool need_insert = false;
OffsetNumber off;
BrinTuple *brtup;
BrinMemTuple *dtup;
BlockNumber heapBlk;
int keyno;
CHECK_FOR_INTERRUPTS();
heapBlk = ItemPointerGetBlockNumber(heaptid);
/* normalize the block number to be the first block in the range */
heapBlk = (heapBlk / pagesPerRange) * pagesPerRange;
brtup = brinGetTupleForHeapBlock(revmap, heapBlk, &buf, &off, NULL,
BUFFER_LOCK_SHARE, NULL);
/* if range is unsummarized, there's nothing to do */
if (!brtup)
break;
/* First time through? */
if (bdesc == NULL)
{
bdesc = brin_build_desc(idxRel);
tupcxt = AllocSetContextCreate(CurrentMemoryContext,
"brininsert cxt",
ALLOCSET_DEFAULT_SIZES);
oldcxt = MemoryContextSwitchTo(tupcxt);
}
dtup = brin_deform_tuple(bdesc, brtup);
/*
* Compare the key values of the new tuple to the stored index values;
* our deformed tuple will get updated if the new tuple doesn't fit
* the original range (note this means we can't break out of the loop
* early). Make a note of whether this happens, so that we know to
* insert the modified tuple later.
*/
for (keyno = 0; keyno < bdesc->bd_tupdesc->natts; keyno++)
{
Datum result;
BrinValues *bval;
FmgrInfo *addValue;
bval = &dtup->bt_columns[keyno];
addValue = index_getprocinfo(idxRel, keyno + 1,
BRIN_PROCNUM_ADDVALUE);
result = FunctionCall4Coll(addValue,
idxRel->rd_indcollation[keyno],
PointerGetDatum(bdesc),
PointerGetDatum(bval),
values[keyno],
nulls[keyno]);
/* if that returned true, we need to insert the updated tuple */
need_insert |= DatumGetBool(result);
}
if (!need_insert)
{
/*
* The tuple is consistent with the new values, so there's nothing
* to do.
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
else
{
Page page = BufferGetPage(buf);
ItemId lp = PageGetItemId(page, off);
Size origsz;
BrinTuple *origtup;
Size newsz;
BrinTuple *newtup;
bool samepage;
/*
* Make a copy of the old tuple, so that we can compare it after
* re-acquiring the lock.
*/
origsz = ItemIdGetLength(lp);
origtup = brin_copy_tuple(brtup, origsz);
/*
* Before releasing the lock, check if we can attempt a same-page
* update. Another process could insert a tuple concurrently in
* the same page though, so downstream we must be prepared to cope
* if this turns out to not be possible after all.
*/
newtup = brin_form_tuple(bdesc, heapBlk, dtup, &newsz);
samepage = brin_can_do_samepage_update(buf, origsz, newsz);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
/*
* Try to update the tuple. If this doesn't work for whatever
* reason, we need to restart from the top; the revmap might be
* pointing at a different tuple for this block now, so we need to
* recompute to ensure both our new heap tuple and the other
* inserter's are covered by the combined tuple. It might be that
* we don't need to update at all.
*/
if (!brin_doupdate(idxRel, pagesPerRange, revmap, heapBlk,
buf, off, origtup, origsz, newtup, newsz,
samepage))
{
/* no luck; start over */
MemoryContextResetAndDeleteChildren(tupcxt);
continue;
}
}
/* success! */
break;
}
brinRevmapTerminate(revmap);
if (BufferIsValid(buf))
ReleaseBuffer(buf);
if (bdesc != NULL)
{
brin_free_desc(bdesc);
MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(tupcxt);
}
return false;
}
/* fork heartbeat sender child */
pid_t
wd_hb_sender(int fork_wait_time, WdHbIf *hb_if)
{
int sock;
pid_t pid = 0;
WdHbPacket pkt;
WdInfo * p = WD_List;
char pack_str[WD_MAX_PACKET_STRING];
int pack_str_len;
sigjmp_buf local_sigjmp_buf;
pid = fork();
if (pid != 0)
{
if (pid == -1)
ereport(PANIC,
(errmsg("failed to fork a heartbeat sender child")));
return pid;
}
on_exit_reset();
processType = PT_HB_SENDER;
if (fork_wait_time > 0)
{
sleep(fork_wait_time);
}
POOL_SETMASK(&UnBlockSig);
signal(SIGTERM, hb_sender_exit);
signal(SIGINT, hb_sender_exit);
signal(SIGQUIT, hb_sender_exit);
signal(SIGCHLD, SIG_IGN);
signal(SIGHUP, SIG_IGN);
signal(SIGUSR1, SIG_IGN);
signal(SIGUSR2, SIG_IGN);
signal(SIGPIPE, SIG_IGN);
signal(SIGALRM, SIG_IGN);
init_ps_display("", "", "", "");
/* Create per loop iteration memory context */
ProcessLoopContext = AllocSetContextCreate(TopMemoryContext,
"wdhb_sender",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(TopMemoryContext);
sock = wd_create_hb_send_socket(hb_if);
set_ps_display("heartbeat sender", false);
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
EmitErrorReport();
MemoryContextSwitchTo(TopMemoryContext);
FlushErrorState();
sleep(pool_config->wd_heartbeat_keepalive);
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
for(;;)
{
MemoryContextSwitchTo(ProcessLoopContext);
MemoryContextResetAndDeleteChildren(ProcessLoopContext);
/* contents of packet */
gettimeofday(&pkt.send_time, NULL);
strlcpy(pkt.from, pool_config->wd_hostname, sizeof(pkt.from));
pkt.from_pgpool_port = pool_config->port;
pkt.status = p->status;
/* authentication key */
if (strlen(pool_config->wd_authkey))
{
/* calculate hash from packet */
pack_str_len = packet_to_string_hb(&pkt, pack_str, sizeof(pack_str));
wd_calc_hash(pack_str, pack_str_len, pkt.hash);
}
/* send heartbeat signal */
wd_hb_send(sock, &pkt, sizeof(pkt), hb_if->addr, hb_if->dest_port);
ereport(DEBUG1,
(errmsg("watchdog heartbeat: send heartbeat signal to %s:%d", hb_if->addr, hb_if->dest_port)));
sleep(pool_config->wd_heartbeat_keepalive);
}
return pid;
}
/*
* ContinuousQueryWorkerStartup
*
* Launches a CQ worker, which continuously generates partial query results to send
* back to the combiner process.
*/
void
ContinuousQueryWorkerRun(Portal portal, ContinuousViewState *state, QueryDesc *queryDesc, ResourceOwner owner)
{
EState *estate = NULL;
DestReceiver *dest;
CmdType operation;
MemoryContext oldcontext;
int timeoutms = state->maxwaitms;
MemoryContext runcontext;
CQProcEntry *entry = GetCQProcEntry(MyCQId);
ResourceOwner cqowner = ResourceOwnerCreate(NULL, "CQResourceOwner");
bool savereadonly = XactReadOnly;
cq_stat_initialize(state->viewid, MyProcPid);
dest = CreateDestReceiver(DestCombiner);
SetCombinerDestReceiverParams(dest, MyCQId);
/* workers only need read-only transactions */
XactReadOnly = true;
runcontext = AllocSetContextCreate(TopMemoryContext, "CQRunContext",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
elog(LOG, "\"%s\" worker %d running", queryDesc->plannedstmt->cq_target->relname, MyProcPid);
MarkWorkerAsRunning(MyCQId, MyWorkerId);
pgstat_report_activity(STATE_RUNNING, queryDesc->sourceText);
TupleBufferInitLatch(WorkerTupleBuffer, MyCQId, MyWorkerId, &MyProc->procLatch);
oldcontext = MemoryContextSwitchTo(runcontext);
retry:
PG_TRY();
{
bool xact_commit = true;
TimestampTz last_process = GetCurrentTimestamp();
TimestampTz last_commit = GetCurrentTimestamp();
start_executor(queryDesc, runcontext, cqowner);
CurrentResourceOwner = cqowner;
estate = queryDesc->estate;
operation = queryDesc->operation;
/*
* Initialize context that lives for the duration of a single iteration
* of the main worker loop
*/
CQExecutionContext = AllocSetContextCreate(estate->es_query_cxt, "CQExecutionContext",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
estate->es_lastoid = InvalidOid;
/*
* Startup combiner receiver
*/
(*dest->rStartup) (dest, operation, queryDesc->tupDesc);
for (;;)
{
if (!TupleBufferHasUnreadSlots())
{
if (TimestampDifferenceExceeds(last_process, GetCurrentTimestamp(), state->emptysleepms))
{
/* force stats flush */
cq_stat_report(true);
pgstat_report_activity(STATE_IDLE, queryDesc->sourceText);
TupleBufferWait(WorkerTupleBuffer, MyCQId, MyWorkerId);
pgstat_report_activity(STATE_RUNNING, queryDesc->sourceText);
}
else
pg_usleep(Min(WAIT_SLEEP_MS, state->emptysleepms) * 1000);
}
TupleBufferResetNotify(WorkerTupleBuffer, MyCQId, MyWorkerId);
if (xact_commit)
StartTransactionCommand();
set_snapshot(estate, cqowner);
CurrentResourceOwner = cqowner;
MemoryContextSwitchTo(estate->es_query_cxt);
estate->es_processed = 0;
estate->es_filtered = 0;
/*
* Run plan on a microbatch
*/
ExecutePlan(estate, queryDesc->planstate, operation,
true, 0, timeoutms, ForwardScanDirection, dest);
IncrementCQExecutions(1);
TupleBufferClearPinnedSlots();
if (state->long_xact)
{
if (TimestampDifferenceExceeds(last_commit, GetCurrentTimestamp(), LONG_RUNNING_XACT_DURATION))
xact_commit = true;
else
xact_commit = false;
}
unset_snapshot(estate, cqowner);
if (xact_commit)
{
CommitTransactionCommand();
last_commit = GetCurrentTimestamp();
}
MemoryContextResetAndDeleteChildren(CQExecutionContext);
MemoryContextSwitchTo(runcontext);
CurrentResourceOwner = cqowner;
if (estate->es_processed || estate->es_filtered)
{
/*
* If the CV query is such that the select does not return any tuples
* ex: select id where id=99; and id=99 does not exist, then this reset
* will fail. What will happen is that the worker will block at the latch for every
* allocated slot, TILL a cv returns a non-zero tuple, at which point
* the worker will resume a simple sleep for the threshold time.
*/
last_process = GetCurrentTimestamp();
/*
* Send stats to the collector
*/
cq_stat_report(false);
}
/* Has the CQ been deactivated? */
if (!entry->active)
{
if (ActiveSnapshotSet())
unset_snapshot(estate, cqowner);
if (IsTransactionState())
CommitTransactionCommand();
break;
}
}
CurrentResourceOwner = cqowner;
/*
* The cleanup functions below expect these things to be registered
*/
RegisterSnapshotOnOwner(estate->es_snapshot, cqowner);
RegisterSnapshotOnOwner(queryDesc->snapshot, cqowner);
RegisterSnapshotOnOwner(queryDesc->crosscheck_snapshot, cqowner);
/* cleanup */
ExecutorFinish(queryDesc);
ExecutorEnd(queryDesc);
FreeQueryDesc(queryDesc);
}
PG_CATCH();
{
EmitErrorReport();
FlushErrorState();
/* Since the worker is read-only, we can simply commit the transaction. */
if (ActiveSnapshotSet())
unset_snapshot(estate, cqowner);
if (IsTransactionState())
CommitTransactionCommand();
TupleBufferUnpinAllPinnedSlots();
TupleBufferClearReaders();
/* This resets the es_query_ctx and in turn the CQExecutionContext */
MemoryContextResetAndDeleteChildren(runcontext);
IncrementCQErrors(1);
if (continuous_query_crash_recovery)
goto retry;
}
PG_END_TRY();
(*dest->rShutdown) (dest);
MemoryContextSwitchTo(oldcontext);
MemoryContextDelete(runcontext);
XactReadOnly = savereadonly;
/*
* Remove proc-level stats
*/
cq_stat_report(true);
cq_stat_send_purge(state->viewid, MyProcPid, CQ_STAT_WORKER);
CurrentResourceOwner = owner;
}
/*
* Main entry point for checkpointer process
*
* This is invoked from AuxiliaryProcessMain, which has already created the
* basic execution environment, but not enabled signals yet.
*/
void
CheckpointerMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext checkpointer_context;
CheckpointerShmem->checkpointer_pid = MyProcPid;
/*
* Properly accept or ignore signals the postmaster might send us
*
* Note: we deliberately ignore SIGTERM, because during a standard Unix
* system shutdown cycle, init will SIGTERM all processes at once. We
* want to wait for the backends to exit, whereupon the postmaster will
* tell us it's okay to shut down (via SIGUSR2).
*/
pqsignal(SIGHUP, ChkptSigHupHandler); /* set flag to read config
* file */
pqsignal(SIGINT, ReqCheckpointHandler); /* request checkpoint */
pqsignal(SIGTERM, SIG_IGN); /* ignore SIGTERM */
pqsignal(SIGQUIT, chkpt_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, chkpt_sigusr1_handler);
pqsignal(SIGUSR2, ReqShutdownHandler); /* request shutdown */
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
sigdelset(&BlockSig, SIGQUIT);
/*
* Initialize so that first time-driven event happens at the correct time.
*/
last_checkpoint_time = last_xlog_switch_time = (pg_time_t) time(NULL);
/*
* Create a resource owner to keep track of our resources (currently only
* buffer pins).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Checkpointer");
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
checkpointer_context = AllocSetContextCreate(TopMemoryContext,
"Checkpointer",
ALLOCSET_DEFAULT_SIZES);
MemoryContextSwitchTo(checkpointer_context);
/*
* If an exception is encountered, processing resumes here.
*
* See notes in postgres.c about the design of this coding.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in checkpointer, but we do have LWLocks, buffers, and temp
* files.
*/
LWLockReleaseAll();
ConditionVariableCancelSleep();
pgstat_report_wait_end();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_SMgr();
AtEOXact_Files();
AtEOXact_HashTables(false);
/* Warn any waiting backends that the checkpoint failed. */
if (ckpt_active)
{
SpinLockAcquire(&CheckpointerShmem->ckpt_lck);
CheckpointerShmem->ckpt_failed++;
CheckpointerShmem->ckpt_done = CheckpointerShmem->ckpt_started;
SpinLockRelease(&CheckpointerShmem->ckpt_lck);
ckpt_active = false;
}
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(checkpointer_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(checkpointer_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Ensure all shared memory values are set correctly for the config. Doing
* this here ensures no race conditions from other concurrent updaters.
*/
UpdateSharedMemoryConfig();
/*
* Advertise our latch that backends can use to wake us up while we're
* sleeping.
*/
ProcGlobal->checkpointerLatch = &MyProc->procLatch;
/*
* Loop forever
*/
for (;;)
{
bool do_checkpoint = false;
int flags = 0;
pg_time_t now;
int elapsed_secs;
int cur_timeout;
int rc;
/* Clear any already-pending wakeups */
ResetLatch(MyLatch);
/*
* Process any requests or signals received recently.
*/
AbsorbFsyncRequests();
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
/*
* Checkpointer is the last process to shut down, so we ask it to
* hold the keys for a range of other tasks required most of which
* have nothing to do with checkpointing at all.
*
* For various reasons, some config values can change dynamically
* so the primary copy of them is held in shared memory to make
* sure all backends see the same value. We make Checkpointer
* responsible for updating the shared memory copy if the
* parameter setting changes because of SIGHUP.
*/
UpdateSharedMemoryConfig();
}
if (checkpoint_requested)
{
checkpoint_requested = false;
do_checkpoint = true;
BgWriterStats.m_requested_checkpoints++;
}
if (shutdown_requested)
{
/*
* From here on, elog(ERROR) should end with exit(1), not send
* control back to the sigsetjmp block above
*/
ExitOnAnyError = true;
/* Close down the database */
ShutdownXLOG(0, 0);
/* Normal exit from the checkpointer is here */
proc_exit(0); /* done */
}
/*
* Force a checkpoint if too much time has elapsed since the last one.
* Note that we count a timed checkpoint in stats only when this
* occurs without an external request, but we set the CAUSE_TIME flag
* bit even if there is also an external request.
*/
now = (pg_time_t) time(NULL);
elapsed_secs = now - last_checkpoint_time;
if (elapsed_secs >= CheckPointTimeout)
{
if (!do_checkpoint)
BgWriterStats.m_timed_checkpoints++;
do_checkpoint = true;
flags |= CHECKPOINT_CAUSE_TIME;
}
/*
* Do a checkpoint if requested.
*/
if (do_checkpoint)
{
bool ckpt_performed = false;
bool do_restartpoint;
/*
* Check if we should perform a checkpoint or a restartpoint. As a
* side-effect, RecoveryInProgress() initializes TimeLineID if
* it's not set yet.
*/
do_restartpoint = RecoveryInProgress();
/*
* Atomically fetch the request flags to figure out what kind of a
* checkpoint we should perform, and increase the started-counter
* to acknowledge that we've started a new checkpoint.
*/
SpinLockAcquire(&CheckpointerShmem->ckpt_lck);
flags |= CheckpointerShmem->ckpt_flags;
CheckpointerShmem->ckpt_flags = 0;
CheckpointerShmem->ckpt_started++;
SpinLockRelease(&CheckpointerShmem->ckpt_lck);
/*
* The end-of-recovery checkpoint is a real checkpoint that's
* performed while we're still in recovery.
*/
if (flags & CHECKPOINT_END_OF_RECOVERY)
do_restartpoint = false;
/*
* We will warn if (a) too soon since last checkpoint (whatever
* caused it) and (b) somebody set the CHECKPOINT_CAUSE_XLOG flag
* since the last checkpoint start. Note in particular that this
* implementation will not generate warnings caused by
* CheckPointTimeout < CheckPointWarning.
*/
if (!do_restartpoint &&
(flags & CHECKPOINT_CAUSE_XLOG) &&
elapsed_secs < CheckPointWarning)
ereport(LOG,
(errmsg_plural("checkpoints are occurring too frequently (%d second apart)",
"checkpoints are occurring too frequently (%d seconds apart)",
elapsed_secs,
elapsed_secs),
errhint("Consider increasing the configuration parameter \"max_wal_size\".")));
/*
* Initialize checkpointer-private variables used during
* checkpoint.
*/
ckpt_active = true;
if (do_restartpoint)
ckpt_start_recptr = GetXLogReplayRecPtr(NULL);
else
ckpt_start_recptr = GetInsertRecPtr();
ckpt_start_time = now;
ckpt_cached_elapsed = 0;
/*
* Do the checkpoint.
*/
if (!do_restartpoint)
{
CreateCheckPoint(flags);
ckpt_performed = true;
}
else
ckpt_performed = CreateRestartPoint(flags);
/*
* After any checkpoint, close all smgr files. This is so we
* won't hang onto smgr references to deleted files indefinitely.
*/
smgrcloseall();
/*
* Indicate checkpoint completion to any waiting backends.
*/
SpinLockAcquire(&CheckpointerShmem->ckpt_lck);
CheckpointerShmem->ckpt_done = CheckpointerShmem->ckpt_started;
SpinLockRelease(&CheckpointerShmem->ckpt_lck);
if (ckpt_performed)
{
/*
* Note we record the checkpoint start time not end time as
* last_checkpoint_time. This is so that time-driven
* checkpoints happen at a predictable spacing.
*/
last_checkpoint_time = now;
}
else
{
/*
* We were not able to perform the restartpoint (checkpoints
* throw an ERROR in case of error). Most likely because we
* have not received any new checkpoint WAL records since the
* last restartpoint. Try again in 15 s.
*/
last_checkpoint_time = now - CheckPointTimeout + 15;
}
ckpt_active = false;
}
/* Check for archive_timeout and switch xlog files if necessary. */
CheckArchiveTimeout();
/*
* Send off activity statistics to the stats collector. (The reason
* why we re-use bgwriter-related code for this is that the bgwriter
* and checkpointer used to be just one process. It's probably not
* worth the trouble to split the stats support into two independent
* stats message types.)
*/
pgstat_send_bgwriter();
/*
* Sleep until we are signaled or it's time for another checkpoint or
* xlog file switch.
*/
now = (pg_time_t) time(NULL);
elapsed_secs = now - last_checkpoint_time;
if (elapsed_secs >= CheckPointTimeout)
continue; /* no sleep for us ... */
cur_timeout = CheckPointTimeout - elapsed_secs;
if (XLogArchiveTimeout > 0 && !RecoveryInProgress())
{
elapsed_secs = now - last_xlog_switch_time;
if (elapsed_secs >= XLogArchiveTimeout)
continue; /* no sleep for us ... */
cur_timeout = Min(cur_timeout, XLogArchiveTimeout - elapsed_secs);
}
rc = WaitLatch(MyLatch,
WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
cur_timeout * 1000L /* convert to ms */,
WAIT_EVENT_CHECKPOINTER_MAIN);
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (rc & WL_POSTMASTER_DEATH)
exit(1);
}
}
void
ContQuerySchedulerMain(int argc, char *argv[])
{
sigjmp_buf local_sigjmp_buf;
List *dbs = NIL;
/* we are a postmaster subprocess now */
IsUnderPostmaster = true;
am_cont_scheduler = true;
/* reset MyProcPid */
MyProcPid = getpid();
MyPMChildSlot = AssignPostmasterChildSlot();
/* record Start Time for logging */
MyStartTime = time(NULL);
/* Identify myself via ps */
init_ps_display("continuous query scheduler process", "", "", "");
ereport(LOG, (errmsg("continuous query scheduler started")));
if (PostAuthDelay)
pg_usleep(PostAuthDelay * 1000000L);
SetProcessingMode(InitProcessing);
/*
* If possible, make this process a group leader, so that the postmaster
* can signal any child processes too. This is only for consistency sake, we
* never fork the scheduler process. Instead dynamic bgworkers are used.
*/
#ifdef HAVE_SETSID
if (setsid() < 0)
elog(FATAL, "setsid() failed: %m");
#endif
/*
* Set up signal handlers. We operate on databases much like a regular
* backend, so we use the same signal handling. See equivalent code in
* tcop/postgres.c.
*/
pqsignal(SIGHUP, sighup_handler);
pqsignal(SIGINT, sigint_handler);
pqsignal(SIGTERM, sigterm_handler);
pqsignal(SIGQUIT, quickdie);
InitializeTimeouts(); /* establishes SIGALRM handler */
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, procsignal_sigusr1_handler);
pqsignal(SIGUSR2, sigusr2_handler);
pqsignal(SIGFPE, FloatExceptionHandler);
pqsignal(SIGCHLD, SIG_DFL);
#define BACKTRACE_SEGFAULTS
#ifdef BACKTRACE_SEGFAULTS
pqsignal(SIGSEGV, debug_segfault);
#endif
/* Early initialization */
BaseInit();
/*
* Create a per-backend PGPROC struct in shared memory, except in the
* EXEC_BACKEND case where this was done in SubPostmasterMain. We must do
* this before we can use LWLocks (and in the EXEC_BACKEND case we already
* had to do some stuff with LWLocks).
*/
#ifndef EXEC_BACKEND
InitProcess();
#endif
InitPostgres(NULL, InvalidOid, NULL, NULL);
SetProcessingMode(NormalProcessing);
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks.
*/
ContQuerySchedulerMemCxt = AllocSetContextCreate(TopMemoryContext,
"ContQuerySchedulerCtx",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(ContQuerySchedulerMemCxt);
/*
* If an exception is encountered, processing resumes here.
*
* This code is a stripped down version of PostgresMain error recovery.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevents interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Forget any pending QueryCancel or timeout request */
disable_all_timeouts(false);
QueryCancelPending = false; /* second to avoid race condition */
/* Report the error to the server log */
EmitErrorReport();
/* Abort the current transaction in order to recover */
AbortCurrentTransaction();
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(ContQuerySchedulerMemCxt);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(ContQuerySchedulerMemCxt);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. We don't want to be
* filling the error logs as fast as we can.
*/
pg_usleep(1000000L);
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/* must unblock signals before calling rebuild_database_list */
PG_SETMASK(&UnBlockSig);
ContQuerySchedulerShmem->scheduler_pid = MyProcPid;
dbs = get_database_list();
/* Loop forever */
for (;;)
{
ListCell *lc;
int rc;
foreach(lc, dbs)
{
DatabaseEntry *db_entry = lfirst(lc);
bool found;
ContQueryProcGroup *grp = hash_search(ContQuerySchedulerShmem->proc_table, &db_entry->oid, HASH_ENTER, &found);
/* If we don't have an entry for this dboid, initialize a new one and fire off bg procs */
if (!found)
{
grp->db_oid = db_entry->oid;
namestrcpy(&grp->db_name, NameStr(db_entry->name));
start_group(grp);
}
}
/* Allow sinval catchup interrupts while sleeping */
EnableCatchupInterrupt();
/*
* Wait until naptime expires or we get some type of signal (all the
* signal handlers will wake us by calling SetLatch).
*/
rc = WaitLatch(&MyProc->procLatch, WL_LATCH_SET | WL_POSTMASTER_DEATH, 0);
ResetLatch(&MyProc->procLatch);
DisableCatchupInterrupt();
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (rc & WL_POSTMASTER_DEATH)
proc_exit(1);
/* the normal shutdown case */
if (got_SIGTERM)
break;
/* update config? */
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
/* update tuning parameters, so that they can be read downstream by background processes */
update_tuning_params();
}
/* terminate a proc group? */
if (got_SIGUSR2)
{
HASH_SEQ_STATUS status;
ContQueryProcGroup *grp;
got_SIGUSR2 = false;
hash_seq_init(&status, ContQuerySchedulerShmem->proc_table);
while ((grp = (ContQueryProcGroup *) hash_seq_search(&status)) != NULL)
{
ListCell *lc;
if (!grp->terminate)
continue;
foreach(lc, dbs)
{
DatabaseEntry *entry = lfirst(lc);
if (entry->oid == grp->db_oid)
{
dbs = list_delete(dbs, entry);
break;
}
}
terminate_group(grp);
}
}
/*
* Main entry point for bgwriter process
*
* This is invoked from BootstrapMain, which has already created the basic
* execution environment, but not enabled signals yet.
*/
void
BackgroundWriterMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext bgwriter_context;
BgWriterShmem->bgwriter_pid = MyProcPid;
am_bg_writer = true;
/*
* If possible, make this process a group leader, so that the postmaster
* can signal any child processes too. (bgwriter probably never has any
* child processes, but for consistency we make all postmaster child
* processes do this.)
*/
#ifdef HAVE_SETSID
if (setsid() < 0)
elog(FATAL, "setsid() failed: %m");
#endif
/*
* Properly accept or ignore signals the postmaster might send us
*
* Note: we deliberately ignore SIGTERM, because during a standard Unix
* system shutdown cycle, init will SIGTERM all processes at once. We
* want to wait for the backends to exit, whereupon the postmaster will
* tell us it's okay to shut down (via SIGUSR2).
*
* SIGUSR1 is presently unused; keep it spare in case someday we want this
* process to participate in sinval messaging.
*/
pqsignal(SIGHUP, BgSigHupHandler); /* set flag to read config file */
pqsignal(SIGINT, ReqCheckpointHandler); /* request checkpoint */
pqsignal(SIGTERM, SIG_IGN); /* ignore SIGTERM */
pqsignal(SIGQUIT, bg_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, SIG_IGN); /* reserve for sinval */
pqsignal(SIGUSR2, ReqShutdownHandler); /* request shutdown */
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
#ifdef HAVE_SIGPROCMASK
sigdelset(&BlockSig, SIGQUIT);
#else
BlockSig &= ~(sigmask(SIGQUIT));
#endif
/*
* Initialize so that first time-driven event happens at the correct time.
*/
last_checkpoint_time = last_xlog_switch_time = time(NULL);
/*
* Create a resource owner to keep track of our resources (currently only
* buffer pins).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Background Writer");
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
bgwriter_context = AllocSetContextCreate(TopMemoryContext,
"Background Writer",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(bgwriter_context);
/*
* If an exception is encountered, processing resumes here.
*
* See notes in postgres.c about the design of this coding.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in bgwriter, but we do have LWLocks, buffers, and temp files.
*/
LWLockReleaseAll();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_Files();
AtEOXact_HashTables(false);
/* Warn any waiting backends that the checkpoint failed. */
if (ckpt_active)
{
/* use volatile pointer to prevent code rearrangement */
volatile BgWriterShmemStruct *bgs = BgWriterShmem;
SpinLockAcquire(&bgs->ckpt_lck);
bgs->ckpt_failed++;
bgs->ckpt_done = bgs->ckpt_started;
SpinLockRelease(&bgs->ckpt_lck);
ckpt_active = false;
}
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(bgwriter_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(bgwriter_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Loop forever
*/
for (;;)
{
bool do_checkpoint = false;
int flags = 0;
time_t now;
int elapsed_secs;
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (!PostmasterIsAlive(true))
exit(1);
/*
* Process any requests or signals received recently.
*/
AbsorbFsyncRequests();
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
}
if (checkpoint_requested)
{
checkpoint_requested = false;
do_checkpoint = true;
BgWriterStats.m_requested_checkpoints++;
}
if (shutdown_requested)
{
/*
* From here on, elog(ERROR) should end with exit(1), not send
* control back to the sigsetjmp block above
*/
ExitOnAnyError = true;
/* Close down the database */
ShutdownXLOG(0, 0);
DumpFreeSpaceMap(0, 0);
/* Normal exit from the bgwriter is here */
proc_exit(0); /* done */
}
/*
* Force a checkpoint if too much time has elapsed since the last one.
* Note that we count a timed checkpoint in stats only when this
* occurs without an external request, but we set the CAUSE_TIME flag
* bit even if there is also an external request.
*/
now = time(NULL);
elapsed_secs = now - last_checkpoint_time;
if (elapsed_secs >= CheckPointTimeout)
{
if (!do_checkpoint)
BgWriterStats.m_timed_checkpoints++;
do_checkpoint = true;
flags |= CHECKPOINT_CAUSE_TIME;
}
/*
* Do a checkpoint if requested, otherwise do one cycle of
* dirty-buffer writing.
*/
if (do_checkpoint)
{
/* use volatile pointer to prevent code rearrangement */
volatile BgWriterShmemStruct *bgs = BgWriterShmem;
/*
* Atomically fetch the request flags to figure out what kind of a
* checkpoint we should perform, and increase the started-counter
* to acknowledge that we've started a new checkpoint.
*/
SpinLockAcquire(&bgs->ckpt_lck);
flags |= bgs->ckpt_flags;
bgs->ckpt_flags = 0;
bgs->ckpt_started++;
SpinLockRelease(&bgs->ckpt_lck);
/*
* We will warn if (a) too soon since last checkpoint (whatever
* caused it) and (b) somebody set the CHECKPOINT_CAUSE_XLOG flag
* since the last checkpoint start. Note in particular that this
* implementation will not generate warnings caused by
* CheckPointTimeout < CheckPointWarning.
*/
if ((flags & CHECKPOINT_CAUSE_XLOG) &&
elapsed_secs < CheckPointWarning)
ereport(LOG,
(errmsg("checkpoints are occurring too frequently (%d seconds apart)",
elapsed_secs),
errhint("Consider increasing the configuration parameter \"checkpoint_segments\".")));
/*
* Initialize bgwriter-private variables used during checkpoint.
*/
ckpt_active = true;
ckpt_start_recptr = GetInsertRecPtr();
ckpt_start_time = now;
ckpt_cached_elapsed = 0;
/*
* Do the checkpoint.
*/
CreateCheckPoint(flags);
/*
* After any checkpoint, close all smgr files. This is so we
* won't hang onto smgr references to deleted files indefinitely.
*/
smgrcloseall();
/*
* Indicate checkpoint completion to any waiting backends.
*/
SpinLockAcquire(&bgs->ckpt_lck);
bgs->ckpt_done = bgs->ckpt_started;
SpinLockRelease(&bgs->ckpt_lck);
ckpt_active = false;
/*
* Note we record the checkpoint start time not end time as
* last_checkpoint_time. This is so that time-driven checkpoints
* happen at a predictable spacing.
*/
last_checkpoint_time = now;
}
else
BgBufferSync();
/* Check for archive_timeout and switch xlog files if necessary. */
CheckArchiveTimeout();
/* Nap for the configured time. */
BgWriterNap();
}
}
pid_t
wd_child(int fork_wait_time)
{
int sock;
volatile int fd;
int rtn;
pid_t pid = 0;
sigjmp_buf local_sigjmp_buf;
pid = fork();
if (pid != 0)
{
if (pid == -1)
ereport(PANIC,
(errmsg("failed to fork a watchdog process")));
return pid;
}
on_exit_reset();
processType = PT_WATCHDOG;
if (fork_wait_time > 0)
{
sleep(fork_wait_time);
}
POOL_SETMASK(&UnBlockSig);
signal(SIGTERM, wd_child_exit);
signal(SIGINT, wd_child_exit);
signal(SIGQUIT, wd_child_exit);
signal(SIGCHLD, SIG_IGN);
signal(SIGHUP, SIG_IGN);
signal(SIGUSR1, SIG_IGN);
signal(SIGUSR2, SIG_IGN);
signal(SIGPIPE, SIG_IGN);
signal(SIGALRM, SIG_IGN);
init_ps_display("", "", "", "");
if (WD_List == NULL)
{
/* memory allocate is not ready */
wd_child_exit(15);
}
/* Create per loop iteration memory context */
ProcessLoopContext = AllocSetContextCreate(TopMemoryContext,
"wd_child_main_loop",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(TopMemoryContext);
sock = wd_create_recv_socket(WD_MYSELF->wd_port);
if (sock < 0)
{
/* socket create failed */
wd_child_exit(15);
}
set_ps_display("watchdog", false);
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
if(fd > 0)
close(fd);
error_context_stack = NULL;
EmitErrorReport();
MemoryContextSwitchTo(TopMemoryContext);
FlushErrorState();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/* child loop */
for(;;)
{
MemoryContextSwitchTo(ProcessLoopContext);
MemoryContextResetAndDeleteChildren(ProcessLoopContext);
fd = -1;
WdPacket buf;
fd = wd_accept(sock);
if (fd < 0)
{
continue;
}
rtn = wd_recv_packet(fd, &buf);
if (rtn == WD_OK)
{
wd_send_response(fd, &buf);
}
close(fd);
}
return pid;
}
/*
* A tuple in the heap is being inserted. To keep a brin index up to date,
* we need to obtain the relevant index tuple and compare its stored values
* with those of the new tuple. If the tuple values are not consistent with
* the summary tuple, we need to update the index tuple.
*
* If autosummarization is enabled, check if we need to summarize the previous
* page range.
*
* If the range is not currently summarized (i.e. the revmap returns NULL for
* it), there's nothing to do for this tuple.
*/
bool
brininsert(Relation idxRel, Datum *values, bool *nulls,
ItemPointer heaptid, Relation heapRel,
IndexUniqueCheck checkUnique,
IndexInfo *indexInfo)
{
BlockNumber pagesPerRange;
BlockNumber origHeapBlk;
BlockNumber heapBlk;
BrinDesc *bdesc = (BrinDesc *) indexInfo->ii_AmCache;
BrinRevmap *revmap;
Buffer buf = InvalidBuffer;
MemoryContext tupcxt = NULL;
MemoryContext oldcxt = CurrentMemoryContext;
bool autosummarize = BrinGetAutoSummarize(idxRel);
revmap = brinRevmapInitialize(idxRel, &pagesPerRange, NULL);
/*
* origHeapBlk is the block number where the insertion occurred. heapBlk
* is the first block in the corresponding page range.
*/
origHeapBlk = ItemPointerGetBlockNumber(heaptid);
heapBlk = (origHeapBlk / pagesPerRange) * pagesPerRange;
for (;;)
{
bool need_insert = false;
OffsetNumber off;
BrinTuple *brtup;
BrinMemTuple *dtup;
int keyno;
CHECK_FOR_INTERRUPTS();
/*
* If auto-summarization is enabled and we just inserted the first
* tuple into the first block of a new non-first page range, request a
* summarization run of the previous range.
*/
if (autosummarize &&
heapBlk > 0 &&
heapBlk == origHeapBlk &&
ItemPointerGetOffsetNumber(heaptid) == FirstOffsetNumber)
{
BlockNumber lastPageRange = heapBlk - 1;
BrinTuple *lastPageTuple;
lastPageTuple =
brinGetTupleForHeapBlock(revmap, lastPageRange, &buf, &off,
NULL, BUFFER_LOCK_SHARE, NULL);
if (!lastPageTuple)
{
bool recorded;
recorded = AutoVacuumRequestWork(AVW_BRINSummarizeRange,
RelationGetRelid(idxRel),
lastPageRange);
if (!recorded)
ereport(LOG,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("request for BRIN range summarization for index \"%s\" page %u was not recorded",
RelationGetRelationName(idxRel),
lastPageRange)));
}
else
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
brtup = brinGetTupleForHeapBlock(revmap, heapBlk, &buf, &off,
NULL, BUFFER_LOCK_SHARE, NULL);
/* if range is unsummarized, there's nothing to do */
if (!brtup)
break;
/* First time through in this statement? */
if (bdesc == NULL)
{
MemoryContextSwitchTo(indexInfo->ii_Context);
bdesc = brin_build_desc(idxRel);
indexInfo->ii_AmCache = (void *) bdesc;
MemoryContextSwitchTo(oldcxt);
}
/* First time through in this brininsert call? */
if (tupcxt == NULL)
{
tupcxt = AllocSetContextCreate(CurrentMemoryContext,
"brininsert cxt",
ALLOCSET_DEFAULT_SIZES);
MemoryContextSwitchTo(tupcxt);
}
dtup = brin_deform_tuple(bdesc, brtup, NULL);
/*
* Compare the key values of the new tuple to the stored index values;
* our deformed tuple will get updated if the new tuple doesn't fit
* the original range (note this means we can't break out of the loop
* early). Make a note of whether this happens, so that we know to
* insert the modified tuple later.
*/
for (keyno = 0; keyno < bdesc->bd_tupdesc->natts; keyno++)
{
Datum result;
BrinValues *bval;
FmgrInfo *addValue;
bval = &dtup->bt_columns[keyno];
addValue = index_getprocinfo(idxRel, keyno + 1,
BRIN_PROCNUM_ADDVALUE);
result = FunctionCall4Coll(addValue,
idxRel->rd_indcollation[keyno],
PointerGetDatum(bdesc),
PointerGetDatum(bval),
values[keyno],
nulls[keyno]);
/* if that returned true, we need to insert the updated tuple */
need_insert |= DatumGetBool(result);
}
if (!need_insert)
{
/*
* The tuple is consistent with the new values, so there's nothing
* to do.
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
else
{
Page page = BufferGetPage(buf);
ItemId lp = PageGetItemId(page, off);
Size origsz;
BrinTuple *origtup;
Size newsz;
BrinTuple *newtup;
bool samepage;
/*
* Make a copy of the old tuple, so that we can compare it after
* re-acquiring the lock.
*/
origsz = ItemIdGetLength(lp);
origtup = brin_copy_tuple(brtup, origsz, NULL, NULL);
/*
* Before releasing the lock, check if we can attempt a same-page
* update. Another process could insert a tuple concurrently in
* the same page though, so downstream we must be prepared to cope
* if this turns out to not be possible after all.
*/
newtup = brin_form_tuple(bdesc, heapBlk, dtup, &newsz);
samepage = brin_can_do_samepage_update(buf, origsz, newsz);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
/*
* Try to update the tuple. If this doesn't work for whatever
* reason, we need to restart from the top; the revmap might be
* pointing at a different tuple for this block now, so we need to
* recompute to ensure both our new heap tuple and the other
* inserter's are covered by the combined tuple. It might be that
* we don't need to update at all.
*/
if (!brin_doupdate(idxRel, pagesPerRange, revmap, heapBlk,
buf, off, origtup, origsz, newtup, newsz,
samepage))
{
/* no luck; start over */
MemoryContextResetAndDeleteChildren(tupcxt);
continue;
}
}
/* success! */
break;
}
brinRevmapTerminate(revmap);
if (BufferIsValid(buf))
ReleaseBuffer(buf);
MemoryContextSwitchTo(oldcxt);
if (tupcxt != NULL)
MemoryContextDelete(tupcxt);
return false;
}
/*
* Fetch dictionary cache entry
*/
TSDictionaryCacheEntry *
lookup_ts_dictionary_cache(Oid dictId)
{
TSDictionaryCacheEntry *entry;
if (TSDictionaryCacheHash == NULL)
{
/* First time through: initialize the hash table */
HASHCTL ctl;
MemSet(&ctl, 0, sizeof(ctl));
ctl.keysize = sizeof(Oid);
ctl.entrysize = sizeof(TSDictionaryCacheEntry);
TSDictionaryCacheHash = hash_create("Tsearch dictionary cache", 8,
&ctl, HASH_ELEM | HASH_BLOBS);
/* Flush cache on pg_ts_dict and pg_ts_template changes */
CacheRegisterSyscacheCallback(TSDICTOID, InvalidateTSCacheCallBack,
PointerGetDatum(TSDictionaryCacheHash));
CacheRegisterSyscacheCallback(TSTEMPLATEOID, InvalidateTSCacheCallBack,
PointerGetDatum(TSDictionaryCacheHash));
/* Also make sure CacheMemoryContext exists */
if (!CacheMemoryContext)
CreateCacheMemoryContext();
}
/* Check single-entry cache */
if (lastUsedDictionary && lastUsedDictionary->dictId == dictId &&
lastUsedDictionary->isvalid)
return lastUsedDictionary;
/* Try to look up an existing entry */
entry = (TSDictionaryCacheEntry *) hash_search(TSDictionaryCacheHash,
(void *) &dictId,
HASH_FIND, NULL);
if (entry == NULL || !entry->isvalid)
{
/*
* If we didn't find one, we want to make one. But first look up the
* object to be sure the OID is real.
*/
HeapTuple tpdict,
tptmpl;
Form_pg_ts_dict dict;
Form_pg_ts_template ctemplate;
MemoryContext saveCtx;
tpdict = SearchSysCache1(TSDICTOID, ObjectIdGetDatum(dictId));
if (!HeapTupleIsValid(tpdict))
elog(ERROR, "cache lookup failed for text search dictionary %u",
dictId);
dict = (Form_pg_ts_dict) GETSTRUCT(tpdict);
/*
* Sanity checks
*/
if (!OidIsValid(dict->dicttemplate))
elog(ERROR, "text search dictionary %u has no ctemplate", dictId);
/*
* Retrieve dictionary's ctemplate
*/
tptmpl = SearchSysCache1(TSTEMPLATEOID,
ObjectIdGetDatum(dict->dicttemplate));
if (!HeapTupleIsValid(tptmpl))
elog(ERROR, "cache lookup failed for text search ctemplate %u",
dict->dicttemplate);
ctemplate = (Form_pg_ts_template) GETSTRUCT(tptmpl);
/*
* Sanity checks
*/
if (!OidIsValid(ctemplate->tmpllexize))
elog(ERROR, "text search ctemplate %u has no lexize method",
ctemplate->tmpllexize);
if (entry == NULL)
{
bool found;
/* Now make the cache entry */
entry = (TSDictionaryCacheEntry *)
hash_search(TSDictionaryCacheHash,
(void *) &dictId,
HASH_ENTER, &found);
Assert(!found); /* it wasn't there a moment ago */
/* Create private___ memory context the first time through */
saveCtx = AllocSetContextCreate(CacheMemoryContext,
NameStr(dict->dictname),
ALLOCSET_SMALL_MINSIZE,
ALLOCSET_SMALL_INITSIZE,
ALLOCSET_SMALL_MAXSIZE);
}
else
{
/* Clear the existing entry's private___ context */
saveCtx = entry->dictCtx;
MemoryContextResetAndDeleteChildren(saveCtx);
}
MemSet(entry, 0, sizeof(TSDictionaryCacheEntry));
entry->dictId = dictId;
entry->dictCtx = saveCtx;
entry->lexizeOid = ctemplate->tmpllexize;
if (OidIsValid(ctemplate->tmplinit))
{
List *dictoptions;
Datum opt;
bool isnull;
MemoryContext oldcontext;
/*
* Init method runs in dictionary's private___ memory context, and we
* make sure the options are stored there too
*/
oldcontext = MemoryContextSwitchTo(entry->dictCtx);
opt = SysCacheGetAttr(TSDICTOID, tpdict,
Anum_pg_ts_dict_dictinitoption,
&isnull);
if (isnull)
dictoptions = NIL;
else
dictoptions = deserialize_deflist(opt);
entry->dictData =
DatumGetPointer(OidFunctionCall1(ctemplate->tmplinit,
PointerGetDatum(dictoptions)));
MemoryContextSwitchTo(oldcontext);
}
ReleaseSysCache(tptmpl);
ReleaseSysCache(tpdict);
fmgr_info_cxt(entry->lexizeOid, &entry->lexize, entry->dictCtx);
entry->isvalid = true;
}
lastUsedDictionary = entry;
return entry;
}
/*
* Main entry point for walwriter process
*
* This is invoked from AuxiliaryProcessMain, which has already created the
* basic execution environment, but not enabled signals yet.
*/
void
WalWriterMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext walwriter_context;
int left_till_hibernate;
bool hibernating;
/*
* Properly accept or ignore signals the postmaster might send us
*
* We have no particular use for SIGINT at the moment, but seems
* reasonable to treat like SIGTERM.
*/
pqsignal(SIGHUP, WalSigHupHandler); /* set flag to read config file */
pqsignal(SIGINT, WalShutdownHandler); /* request shutdown */
pqsignal(SIGTERM, WalShutdownHandler); /* request shutdown */
pqsignal(SIGQUIT, wal_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, walwriter_sigusr1_handler);
pqsignal(SIGUSR2, SIG_IGN); /* not used */
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
sigdelset(&BlockSig, SIGQUIT);
/*
* Create a resource owner to keep track of our resources (not clear that
* we need this, but may as well have one).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Wal Writer");
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
walwriter_context = AllocSetContextCreate(TopMemoryContext,
"Wal Writer",
ALLOCSET_DEFAULT_SIZES);
MemoryContextSwitchTo(walwriter_context);
/*
* If an exception is encountered, processing resumes here.
*
* This code is heavily based on bgwriter.c, q.v.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in walwriter, but we do have LWLocks, and perhaps buffers?
*/
LWLockReleaseAll();
ConditionVariableCancelSleep();
pgstat_report_wait_end();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_SMgr();
AtEOXact_Files();
AtEOXact_HashTables(false);
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(walwriter_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(walwriter_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Reset hibernation state after any error.
*/
left_till_hibernate = LOOPS_UNTIL_HIBERNATE;
hibernating = false;
SetWalWriterSleeping(false);
/*
* Advertise our latch that backends can use to wake us up while we're
* sleeping.
*/
ProcGlobal->walwriterLatch = &MyProc->procLatch;
/*
* Loop forever
*/
for (;;)
{
long cur_timeout;
int rc;
/*
* Advertise whether we might hibernate in this cycle. We do this
* before resetting the latch to ensure that any async commits will
* see the flag set if they might possibly need to wake us up, and
* that we won't miss any signal they send us. (If we discover work
* to do in the last cycle before we would hibernate, the global flag
* will be set unnecessarily, but little harm is done.) But avoid
* touching the global flag if it doesn't need to change.
*/
if (hibernating != (left_till_hibernate <= 1))
{
hibernating = (left_till_hibernate <= 1);
SetWalWriterSleeping(hibernating);
}
/* Clear any already-pending wakeups */
ResetLatch(MyLatch);
/*
* Process any requests or signals received recently.
*/
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
}
if (shutdown_requested)
{
/* Normal exit from the walwriter is here */
proc_exit(0); /* done */
}
/*
* Do what we're here for; then, if XLogBackgroundFlush() found useful
* work to do, reset hibernation counter.
*/
if (XLogBackgroundFlush())
left_till_hibernate = LOOPS_UNTIL_HIBERNATE;
else if (left_till_hibernate > 0)
left_till_hibernate--;
/*
* Sleep until we are signaled or WalWriterDelay has elapsed. If we
* haven't done anything useful for quite some time, lengthen the
* sleep time so as to reduce the server's idle power consumption.
*/
if (left_till_hibernate > 0)
cur_timeout = WalWriterDelay; /* in ms */
else
cur_timeout = WalWriterDelay * HIBERNATE_FACTOR;
rc = WaitLatch(MyLatch,
WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
cur_timeout,
WAIT_EVENT_WAL_WRITER_MAIN);
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (rc & WL_POSTMASTER_DEATH)
exit(1);
}
}
/*
* Main entry point for bgwriter process
*
* This is invoked from BootstrapMain, which has already created the basic
* execution environment, but not enabled signals yet.
*/
void
BackgroundWriterMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext bgwriter_context;
am_bg_writer = true;
/*
* If possible, make this process a group leader, so that the postmaster
* can signal any child processes too. (bgwriter probably never has any
* child processes, but for consistency we make all postmaster child
* processes do this.)
*/
#ifdef HAVE_SETSID
if (setsid() < 0)
elog(FATAL, "setsid() failed: %m");
#endif
/*
* Properly accept or ignore signals the postmaster might send us
*
* SIGUSR1 is presently unused; keep it spare in case someday we want this
* process to participate in ProcSignal signalling.
*/
pqsignal(SIGHUP, BgSigHupHandler); /* set flag to read config file */
pqsignal(SIGINT, SIG_IGN); /* as of 9.2 no longer requests checkpoint */
pqsignal(SIGTERM, ReqShutdownHandler); /* shutdown */
pqsignal(SIGQUIT, bg_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, SIG_IGN); /* reserve for ProcSignal */
pqsignal(SIGUSR2, SIG_IGN);
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
sigdelset(&BlockSig, SIGQUIT);
/*
* Create a resource owner to keep track of our resources (currently only
* buffer pins).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Background Writer");
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
bgwriter_context = AllocSetContextCreate(TopMemoryContext,
"Background Writer",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(bgwriter_context);
/*
* If an exception is encountered, processing resumes here.
*
* See notes in postgres.c about the design of this coding.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in bgwriter, but we do have LWLocks, buffers, and temp files.
*/
LWLockReleaseAll();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_Files();
AtEOXact_HashTables(false);
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(bgwriter_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(bgwriter_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Use the recovery target timeline ID during recovery
*/
if (RecoveryInProgress())
ThisTimeLineID = GetRecoveryTargetTLI();
/*
* Loop forever
*/
for (;;)
{
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (!PostmasterIsAlive())
exit(1);
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
/* update global shmem state for sync rep */
}
if (shutdown_requested)
{
/*
* From here on, elog(ERROR) should end with exit(1), not send
* control back to the sigsetjmp block above
*/
ExitOnAnyError = true;
/* Normal exit from the bgwriter is here */
proc_exit(0); /* done */
}
/*
* Do one cycle of dirty-buffer writing.
*/
BgBufferSync();
/* Nap for the configured time. */
BgWriterNap();
}
}
/*
* Main entry point for bgwriter process
*
* This is invoked from AuxiliaryProcessMain, which has already created the
* basic execution environment, but not enabled signals yet.
*/
void
BackgroundWriterMain(void)
{
sigjmp_buf local_sigjmp_buf;
MemoryContext bgwriter_context;
bool prev_hibernate;
/*
* Properly accept or ignore signals the postmaster might send us.
*
* bgwriter doesn't participate in ProcSignal signalling, but a SIGUSR1
* handler is still needed for latch wakeups.
*/
pqsignal(SIGHUP, BgSigHupHandler); /* set flag to read config file */
pqsignal(SIGINT, SIG_IGN);
pqsignal(SIGTERM, ReqShutdownHandler); /* shutdown */
pqsignal(SIGQUIT, bg_quickdie); /* hard crash time */
pqsignal(SIGALRM, SIG_IGN);
pqsignal(SIGPIPE, SIG_IGN);
pqsignal(SIGUSR1, bgwriter_sigusr1_handler);
pqsignal(SIGUSR2, SIG_IGN);
/*
* Reset some signals that are accepted by postmaster but not here
*/
pqsignal(SIGCHLD, SIG_DFL);
pqsignal(SIGTTIN, SIG_DFL);
pqsignal(SIGTTOU, SIG_DFL);
pqsignal(SIGCONT, SIG_DFL);
pqsignal(SIGWINCH, SIG_DFL);
/* We allow SIGQUIT (quickdie) at all times */
sigdelset(&BlockSig, SIGQUIT);
/*
* Create a resource owner to keep track of our resources (currently only
* buffer pins).
*/
CurrentResourceOwner = ResourceOwnerCreate(NULL, "Background Writer");
/*
* We just started, assume there has been either a shutdown or
* end-of-recovery snapshot.
*/
last_snapshot_ts = GetCurrentTimestamp();
/*
* Create a memory context that we will do all our work in. We do this so
* that we can reset the context during error recovery and thereby avoid
* possible memory leaks. Formerly this code just ran in
* TopMemoryContext, but resetting that would be a really bad idea.
*/
bgwriter_context = AllocSetContextCreate(TopMemoryContext,
"Background Writer",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(bgwriter_context);
/*
* If an exception is encountered, processing resumes here.
*
* See notes in postgres.c about the design of this coding.
*/
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
/* Prevent interrupts while cleaning up */
HOLD_INTERRUPTS();
/* Report the error to the server log */
EmitErrorReport();
/*
* These operations are really just a minimal subset of
* AbortTransaction(). We don't have very many resources to worry
* about in bgwriter, but we do have LWLocks, buffers, and temp files.
*/
LWLockReleaseAll();
AbortBufferIO();
UnlockBuffers();
/* buffer pins are released here: */
ResourceOwnerRelease(CurrentResourceOwner,
RESOURCE_RELEASE_BEFORE_LOCKS,
false, true);
/* we needn't bother with the other ResourceOwnerRelease phases */
AtEOXact_Buffers(false);
AtEOXact_SMgr();
AtEOXact_Files();
AtEOXact_HashTables(false);
/*
* Now return to normal top-level context and clear ErrorContext for
* next time.
*/
MemoryContextSwitchTo(bgwriter_context);
FlushErrorState();
/* Flush any leaked data in the top-level context */
MemoryContextResetAndDeleteChildren(bgwriter_context);
/* Now we can allow interrupts again */
RESUME_INTERRUPTS();
/*
* Sleep at least 1 second after any error. A write error is likely
* to be repeated, and we don't want to be filling the error logs as
* fast as we can.
*/
pg_usleep(1000000L);
/*
* Close all open files after any error. This is helpful on Windows,
* where holding deleted files open causes various strange errors.
* It's not clear we need it elsewhere, but shouldn't hurt.
*/
smgrcloseall();
/* Report wait end here, when there is no further possibility of wait */
pgstat_report_wait_end();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/*
* Unblock signals (they were blocked when the postmaster forked us)
*/
PG_SETMASK(&UnBlockSig);
/*
* Reset hibernation state after any error.
*/
prev_hibernate = false;
/*
* Loop forever
*/
for (;;)
{
bool can_hibernate;
int rc;
/* Clear any already-pending wakeups */
ResetLatch(MyLatch);
if (got_SIGHUP)
{
got_SIGHUP = false;
ProcessConfigFile(PGC_SIGHUP);
}
if (shutdown_requested)
{
/*
* From here on, elog(ERROR) should end with exit(1), not send
* control back to the sigsetjmp block above
*/
ExitOnAnyError = true;
/* Normal exit from the bgwriter is here */
proc_exit(0); /* done */
}
/*
* Do one cycle of dirty-buffer writing.
*/
can_hibernate = BgBufferSync();
/*
* Send off activity statistics to the stats collector
*/
pgstat_send_bgwriter();
if (FirstCallSinceLastCheckpoint())
{
/*
* After any checkpoint, close all smgr files. This is so we
* won't hang onto smgr references to deleted files indefinitely.
*/
smgrcloseall();
}
/*
* Log a new xl_running_xacts every now and then so replication can
* get into a consistent state faster (think of suboverflowed
* snapshots) and clean up resources (locks, KnownXids*) more
* frequently. The costs of this are relatively low, so doing it 4
* times (LOG_SNAPSHOT_INTERVAL_MS) a minute seems fine.
*
* We assume the interval for writing xl_running_xacts is
* significantly bigger than BgWriterDelay, so we don't complicate the
* overall timeout handling but just assume we're going to get called
* often enough even if hibernation mode is active. It's not that
* important that log_snap_interval_ms is met strictly. To make sure
* we're not waking the disk up unnecessarily on an idle system we
* check whether there has been any WAL inserted since the last time
* we've logged a running xacts.
*
* We do this logging in the bgwriter as its the only process that is
* run regularly and returns to its mainloop all the time. E.g.
* Checkpointer, when active, is barely ever in its mainloop and thus
* makes it hard to log regularly.
*/
if (XLogStandbyInfoActive() && !RecoveryInProgress())
{
TimestampTz timeout = 0;
TimestampTz now = GetCurrentTimestamp();
timeout = TimestampTzPlusMilliseconds(last_snapshot_ts,
LOG_SNAPSHOT_INTERVAL_MS);
/*
* only log if enough time has passed and some xlog record has
* been inserted.
*/
if (now >= timeout &&
last_snapshot_lsn != GetXLogInsertRecPtr())
{
last_snapshot_lsn = LogStandbySnapshot();
last_snapshot_ts = now;
}
}
/*
* Sleep until we are signaled or BgWriterDelay has elapsed.
*
* Note: the feedback control loop in BgBufferSync() expects that we
* will call it every BgWriterDelay msec. While it's not critical for
* correctness that that be exact, the feedback loop might misbehave
* if we stray too far from that. Hence, avoid loading this process
* down with latch events that are likely to happen frequently during
* normal operation.
*/
rc = WaitLatch(MyLatch,
WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
BgWriterDelay /* ms */ );
/*
* If no latch event and BgBufferSync says nothing's happening, extend
* the sleep in "hibernation" mode, where we sleep for much longer
* than bgwriter_delay says. Fewer wakeups save electricity. When a
* backend starts using buffers again, it will wake us up by setting
* our latch. Because the extra sleep will persist only as long as no
* buffer allocations happen, this should not distort the behavior of
* BgBufferSync's control loop too badly; essentially, it will think
* that the system-wide idle interval didn't exist.
*
* There is a race condition here, in that a backend might allocate a
* buffer between the time BgBufferSync saw the alloc count as zero
* and the time we call StrategyNotifyBgWriter. While it's not
* critical that we not hibernate anyway, we try to reduce the odds of
* that by only hibernating when BgBufferSync says nothing's happening
* for two consecutive cycles. Also, we mitigate any possible
* consequences of a missed wakeup by not hibernating forever.
*/
if (rc == WL_TIMEOUT && can_hibernate && prev_hibernate)
{
/* Ask for notification at next buffer allocation */
StrategyNotifyBgWriter(MyProc->pgprocno);
/* Sleep ... */
rc = WaitLatch(MyLatch,
WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
BgWriterDelay * HIBERNATE_FACTOR);
/* Reset the notification request in case we timed out */
StrategyNotifyBgWriter(-1);
}
/*
* Emergency bailout if postmaster has died. This is to avoid the
* necessity for manual cleanup of all postmaster children.
*/
if (rc & WL_POSTMASTER_DEATH)
exit(1);
prev_hibernate = can_hibernate;
}
}
/* fork lifecheck process*/
static pid_t
fork_a_lifecheck(int fork_wait_time)
{
pid_t pid;
sigjmp_buf local_sigjmp_buf;
pid = fork();
if (pid != 0)
{
if (pid == -1)
ereport(ERROR,
(errmsg("failed to fork a lifecheck process")));
return pid;
}
on_exit_reset();
processType = PT_LIFECHECK;
if (fork_wait_time > 0) {
sleep(fork_wait_time);
}
POOL_SETMASK(&UnBlockSig);
init_ps_display("", "", "", "");
signal(SIGTERM, wd_exit);
signal(SIGINT, wd_exit);
signal(SIGQUIT, wd_exit);
signal(SIGCHLD, SIG_DFL);
signal(SIGHUP, SIG_IGN);
signal(SIGPIPE, SIG_IGN);
/* Create per loop iteration memory context */
ProcessLoopContext = AllocSetContextCreate(TopMemoryContext,
"wd_lifecheck_main_loop",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(TopMemoryContext);
set_ps_display("lifecheck",false);
/* wait until ready to go */
while (WD_OK != is_wd_lifecheck_ready())
{
sleep(pool_config->wd_interval * 10);
}
ereport(LOG,
(errmsg("watchdog: lifecheck started")));
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
EmitErrorReport();
MemoryContextSwitchTo(TopMemoryContext);
FlushErrorState();
sleep(pool_config->wd_heartbeat_keepalive);
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
/* watchdog loop */
for (;;)
{
MemoryContextSwitchTo(ProcessLoopContext);
MemoryContextResetAndDeleteChildren(ProcessLoopContext);
/* pgpool life check */
wd_lifecheck();
sleep(pool_config->wd_interval);
}
return pid;
}
/*
* A tuple in the heap is being inserted. To keep a brin index up to date,
* we need to obtain the relevant index tuple and compare its stored values
* with those of the new tuple. If the tuple values are not consistent with
* the summary tuple, we need to update the index tuple.
*
* If the range is not currently summarized (i.e. the revmap returns NULL for
* it), there's nothing to do.
*/
Datum
brininsert(PG_FUNCTION_ARGS)
{
Relation idxRel = (Relation) PG_GETARG_POINTER(0);
Datum *values = (Datum *) PG_GETARG_POINTER(1);
bool *nulls = (bool *) PG_GETARG_POINTER(2);
ItemPointer heaptid = (ItemPointer) PG_GETARG_POINTER(3);
/* we ignore the rest of our arguments */
BlockNumber pagesPerRange;
BrinDesc *bdesc = NULL;
BrinRevmap *revmap;
Buffer buf = InvalidBuffer;
MemoryContext tupcxt = NULL;
MemoryContext oldcxt = NULL;
revmap = brinRevmapInitialize(idxRel, &pagesPerRange);
for (;;)
{
bool need_insert = false;
OffsetNumber off;
BrinTuple *brtup;
BrinMemTuple *dtup;
BlockNumber heapBlk;
int keyno;
#ifdef USE_ASSERT_CHECKING
BrinTuple *tmptup;
BrinMemTuple *tmpdtup;
Size tmpsiz;
#endif
CHECK_FOR_INTERRUPTS();
heapBlk = ItemPointerGetBlockNumber(heaptid);
/* normalize the block number to be the first block in the range */
heapBlk = (heapBlk / pagesPerRange) * pagesPerRange;
brtup = brinGetTupleForHeapBlock(revmap, heapBlk, &buf, &off, NULL,
BUFFER_LOCK_SHARE);
/* if range is unsummarized, there's nothing to do */
if (!brtup)
break;
/* First time through? */
if (bdesc == NULL)
{
bdesc = brin_build_desc(idxRel);
tupcxt = AllocSetContextCreate(CurrentMemoryContext,
"brininsert cxt",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
oldcxt = MemoryContextSwitchTo(tupcxt);
}
dtup = brin_deform_tuple(bdesc, brtup);
#ifdef USE_ASSERT_CHECKING
{
/*
* When assertions are enabled, we use this as an opportunity to
* test the "union" method, which would otherwise be used very
* rarely: first create a placeholder tuple, and addValue the
* value we just got into it. Then union the existing index tuple
* with the updated placeholder tuple. The tuple resulting from
* that union should be identical to the one resulting from the
* regular operation (straight addValue) below.
*
* Here we create the tuple to compare with; the actual comparison
* is below.
*/
tmptup = brin_form_placeholder_tuple(bdesc, heapBlk, &tmpsiz);
tmpdtup = brin_deform_tuple(bdesc, tmptup);
for (keyno = 0; keyno < bdesc->bd_tupdesc->natts; keyno++)
{
BrinValues *bval;
FmgrInfo *addValue;
bval = &tmpdtup->bt_columns[keyno];
addValue = index_getprocinfo(idxRel, keyno + 1,
BRIN_PROCNUM_ADDVALUE);
FunctionCall4Coll(addValue,
idxRel->rd_indcollation[keyno],
PointerGetDatum(bdesc),
PointerGetDatum(bval),
values[keyno],
nulls[keyno]);
}
union_tuples(bdesc, tmpdtup, brtup);
tmpdtup->bt_placeholder = dtup->bt_placeholder;
tmptup = brin_form_tuple(bdesc, heapBlk, tmpdtup, &tmpsiz);
}
#endif
/*
* Compare the key values of the new tuple to the stored index values;
* our deformed tuple will get updated if the new tuple doesn't fit
* the original range (note this means we can't break out of the loop
* early). Make a note of whether this happens, so that we know to
* insert the modified tuple later.
*/
for (keyno = 0; keyno < bdesc->bd_tupdesc->natts; keyno++)
{
Datum result;
BrinValues *bval;
FmgrInfo *addValue;
bval = &dtup->bt_columns[keyno];
addValue = index_getprocinfo(idxRel, keyno + 1,
BRIN_PROCNUM_ADDVALUE);
result = FunctionCall4Coll(addValue,
idxRel->rd_indcollation[keyno],
PointerGetDatum(bdesc),
PointerGetDatum(bval),
values[keyno],
nulls[keyno]);
/* if that returned true, we need to insert the updated tuple */
need_insert |= DatumGetBool(result);
}
#ifdef USE_ASSERT_CHECKING
{
/*
* Now we can compare the tuple produced by the union function
* with the one from plain addValue.
*/
BrinTuple *cmptup;
Size cmpsz;
cmptup = brin_form_tuple(bdesc, heapBlk, dtup, &cmpsz);
Assert(brin_tuples_equal(tmptup, tmpsiz, cmptup, cmpsz));
}
#endif
if (!need_insert)
{
/*
* The tuple is consistent with the new values, so there's nothing
* to do.
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
else
{
Page page = BufferGetPage(buf);
ItemId lp = PageGetItemId(page, off);
Size origsz;
BrinTuple *origtup;
Size newsz;
BrinTuple *newtup;
bool samepage;
/*
* Make a copy of the old tuple, so that we can compare it after
* re-acquiring the lock.
*/
origsz = ItemIdGetLength(lp);
origtup = brin_copy_tuple(brtup, origsz);
/*
* Before releasing the lock, check if we can attempt a same-page
* update. Another process could insert a tuple concurrently in
* the same page though, so downstream we must be prepared to cope
* if this turns out to not be possible after all.
*/
newtup = brin_form_tuple(bdesc, heapBlk, dtup, &newsz);
samepage = brin_can_do_samepage_update(buf, origsz, newsz);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
/*
* Try to update the tuple. If this doesn't work for whatever
* reason, we need to restart from the top; the revmap might be
* pointing at a different tuple for this block now, so we need to
* recompute to ensure both our new heap tuple and the other
* inserter's are covered by the combined tuple. It might be that
* we don't need to update at all.
*/
if (!brin_doupdate(idxRel, pagesPerRange, revmap, heapBlk,
buf, off, origtup, origsz, newtup, newsz,
samepage))
{
/* no luck; start over */
MemoryContextResetAndDeleteChildren(tupcxt);
continue;
}
}
/* success! */
break;
}
brinRevmapTerminate(revmap);
if (BufferIsValid(buf))
ReleaseBuffer(buf);
if (bdesc != NULL)
{
brin_free_desc(bdesc);
MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(tupcxt);
}
return BoolGetDatum(false);
}
/*
* Clean up SPI state at subtransaction commit or abort.
*
* During commit, there shouldn't be any unclosed entries remaining from
* the current subtransaction; we emit a warning if any are found.
*/
void
AtEOSubXact_SPI(bool isCommit, SubTransactionId mySubid)
{
bool found = false;
while (_SPI_connected >= 0)
{
_SPI_connection *connection = &(_SPI_stack[_SPI_connected]);
if (connection->connectSubid != mySubid)
break; /* couldn't be any underneath it either */
found = true;
/*
* Release procedure memory explicitly (see note in SPI_connect)
*/
if (connection->execCxt)
{
MemoryContextDelete(connection->execCxt);
connection->execCxt = NULL;
}
if (connection->procCxt)
{
MemoryContextDelete(connection->procCxt);
connection->procCxt = NULL;
}
/*
* Pop the stack entry and reset global variables. Unlike
* SPI_finish(), we don't risk switching to memory contexts that might
* be already gone.
*/
_SPI_connected--;
_SPI_curid = _SPI_connected;
if (_SPI_connected == -1)
_SPI_current = NULL;
else
_SPI_current = &(_SPI_stack[_SPI_connected]);
SPI_processed = 0;
SPI_lastoid = InvalidOid;
SPI_tuptable = NULL;
}
if (found && isCommit)
ereport(WARNING,
(errcode(ERRCODE_WARNING),
errmsg("subtransaction left non-empty SPI stack"),
errhint("Check for missing \"SPI_finish\" calls.")));
/*
* If we are aborting a subtransaction and there is an open SPI context
* surrounding the subxact, clean up to prevent memory leakage.
*/
if (_SPI_current && !isCommit)
{
/* free Executor memory the same as _SPI_end_call would do */
MemoryContextResetAndDeleteChildren(_SPI_current->execCxt);
/* throw away any partially created tuple-table */
SPI_freetuptable(_SPI_current->tuptable);
_SPI_current->tuptable = NULL;
}
}
/* fork heartbeat receiver child */
pid_t
wd_hb_receiver(int fork_wait_time, WdHbIf *hb_if)
{
int sock;
pid_t pid = 0;
WdHbPacket pkt;
struct timeval tv;
char from[WD_MAX_HOST_NAMELEN];
int from_pgpool_port;
char buf[(MD5_PASSWD_LEN+1)*2];
char pack_str[WD_MAX_PACKET_STRING];
int pack_str_len;
sigjmp_buf local_sigjmp_buf;
WdInfo * p;
pid = fork();
if (pid != 0)
{
if (pid == -1)
ereport(PANIC,
(errmsg("failed to fork a heartbeat receiver child")));
return pid;
}
on_exit_reset();
processType = PT_HB_RECEIVER;
if (fork_wait_time > 0)
{
sleep(fork_wait_time);
}
POOL_SETMASK(&UnBlockSig);
signal(SIGTERM, hb_receiver_exit);
signal(SIGINT, hb_receiver_exit);
signal(SIGQUIT, hb_receiver_exit);
signal(SIGCHLD, SIG_IGN);
signal(SIGHUP, SIG_IGN);
signal(SIGUSR1, SIG_IGN);
signal(SIGUSR2, SIG_IGN);
signal(SIGPIPE, SIG_IGN);
signal(SIGALRM, SIG_IGN);
init_ps_display("", "", "", "");
/* Create per loop iteration memory context */
ProcessLoopContext = AllocSetContextCreate(TopMemoryContext,
"wdhb_hb_receiver",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
MemoryContextSwitchTo(TopMemoryContext);
sock = wd_create_hb_recv_socket(hb_if);
set_ps_display("heartbeat receiver", false);
if (sigsetjmp(local_sigjmp_buf, 1) != 0)
{
/* Since not using PG_TRY, must reset error stack by hand */
error_context_stack = NULL;
EmitErrorReport();
MemoryContextSwitchTo(TopMemoryContext);
FlushErrorState();
}
/* We can now handle ereport(ERROR) */
PG_exception_stack = &local_sigjmp_buf;
for(;;)
{
MemoryContextSwitchTo(ProcessLoopContext);
MemoryContextResetAndDeleteChildren(ProcessLoopContext);
/* receive heartbeat signal */
wd_hb_recv(sock, &pkt);
/* authentication */
if (strlen(pool_config->wd_authkey))
{
/* calculate hash from packet */
pack_str_len = packet_to_string_hb(&pkt, pack_str, sizeof(pack_str));
wd_calc_hash(pack_str, pack_str_len, buf);
if (strcmp(pkt.hash, buf))
ereport(ERROR,
(errmsg("watchdog heartbeat receive"),
errdetail("authentication failed")));
}
/* get current time */
gettimeofday(&tv, NULL);
/* who send this packet? */
strlcpy(from, pkt.from, sizeof(from));
from_pgpool_port = pkt.from_pgpool_port;
p = WD_List;
while (p->status != WD_END)
{
if (!strcmp(p->hostname, from) && p->pgpool_port == from_pgpool_port)
{
/* ignore the packet from down pgpool */
if (pkt.status == WD_DOWN)
{
ereport(DEBUG1,
(errmsg("watchdog heartbeat: received heartbeat signal from \"%s:%d\" whose status is down. ignored",
from, from_pgpool_port)));
break;
}
/* this is the first packet or the latest packet */
if (!WD_TIME_ISSET(p->hb_send_time) ||
WD_TIME_BEFORE(p->hb_send_time, pkt.send_time))
{
ereport(DEBUG1,
(errmsg("watchdog heartbeat: received heartbeat signal from \"%s:%d\"",
from, from_pgpool_port)));
p->hb_send_time = pkt.send_time;
p->hb_last_recv_time = tv;
}
else
{
ereport(DEBUG1,
(errmsg("watchdog heartbeat: received heartbeat signal is older than the latest, ignored")));
}
break;
}
p++;
}
}
return pid;
}
