/*
* leadlag_common
* common operation of lead() and lag()
* For lead() forward is true, whereas for lag() it is false.
* withoffset indicates we have an offset second argument.
* withdefault indicates we have a default third argument.
*/
static Datum
leadlag_common(FunctionCallInfo fcinfo,
bool forward, bool withoffset, bool withdefault)
{
WindowObject winobj = PG_WINDOW_OBJECT();
int32 offset;
bool const_offset;
Datum result;
bool isnull;
bool isout;
if (withoffset)
{
offset = DatumGetInt32(WinGetFuncArgCurrent(winobj, 1, &isnull));
if (isnull)
PG_RETURN_NULL();
const_offset = get_fn_expr_arg_stable(fcinfo->flinfo, 1);
}
else
{
offset = 1;
const_offset = true;
}
result = WinGetFuncArgInPartition(winobj, 0,
(forward ? offset : -offset),
WINDOW_SEEK_CURRENT,
const_offset,
&isnull, &isout);
if (isout)
{
/*
* target row is out of the partition; supply default value if
* provided. otherwise it'll stay NULL
*/
if (withdefault)
result = WinGetFuncArgCurrent(winobj, 2, &isnull);
}
if (isnull)
PG_RETURN_NULL();
PG_RETURN_DATUM(result);
}
Datum
c_overpaid(PG_FUNCTION_ARGS)
{
HeapTupleHeader t = PG_GETARG_HEAPTUPLEHEADER(0);
int32 limit = PG_GETARG_INT32(1);
bool isnull;
int32 salary;
salary = DatumGetInt32(GetAttributeByName(t, "salary", &isnull));
if (isnull)
PG_RETURN_BOOL(false);
/*
* Alternatively, we might prefer to do PG_RETURN_NULL() for null salary
*/
PG_RETURN_BOOL(salary > limit);
}
/*
* Compare two keys of the same index column
*/
int
ginCompareEntries(GinState *ginstate, OffsetNumber attnum,
Datum a, GinNullCategory categorya,
Datum b, GinNullCategory categoryb)
{
/* if not of same null category, sort by that first */
if (categorya != categoryb)
return (categorya < categoryb) ? -1 : 1;
/* all null items in same category are equal */
if (categorya != GIN_CAT_NORM_KEY)
return 0;
/* both not null, so safe to call the compareFn */
return DatumGetInt32(FunctionCall2Coll(&ginstate->compareFn[attnum - 1],
ginstate->supportCollation[attnum - 1],
a, b));
}
Datum
on_partitions_updated(PG_FUNCTION_ARGS)
{
Oid relid;
PartRelationInfo *prel;
/* Parent relation oid */
relid = DatumGetInt32(PG_GETARG_DATUM(0));
prel = get_pathman_relation_info(relid, NULL);
if (prel != NULL)
{
LWLockAcquire(pmstate->load_config_lock, LW_EXCLUSIVE);
remove_relation_info(relid);
load_relations_hashtable(false);
LWLockRelease(pmstate->load_config_lock);
}
PG_RETURN_NULL();
}
/*
* CompareShardIntervals acts as a helper function to compare two shard intervals
* by their minimum values, using the value's type comparison function.
*
* If a shard interval does not have min/max value, it's treated as being greater
* than the other.
*/
int
CompareShardIntervals(const void *leftElement, const void *rightElement,
FmgrInfo *typeCompareFunction)
{
ShardInterval *leftShardInterval = *((ShardInterval **) leftElement);
ShardInterval *rightShardInterval = *((ShardInterval **) rightElement);
Datum leftDatum = 0;
Datum rightDatum = 0;
Datum comparisonDatum = 0;
int comparisonResult = 0;
Assert(typeCompareFunction != NULL);
/*
* Left element should be treated as the greater element in case it doesn't
* have min or max values.
*/
if (!leftShardInterval->minValueExists || !leftShardInterval->maxValueExists)
{
comparisonResult = 1;
return comparisonResult;
}
/*
* Right element should be treated as the greater element in case it doesn't
* have min or max values.
*/
if (!rightShardInterval->minValueExists || !rightShardInterval->maxValueExists)
{
comparisonResult = -1;
return comparisonResult;
}
/* if both shard interval have min/max values, calculate the comparison result */
leftDatum = leftShardInterval->minValue;
rightDatum = rightShardInterval->minValue;
comparisonDatum = CompareCall2(typeCompareFunction, leftDatum, rightDatum);
comparisonResult = DatumGetInt32(comparisonDatum);
return comparisonResult;
}
void GpRelationNode_GetValues(
Datum *values,
Oid *relfilenodeOid,
int32 *segmentFileNum,
int64 *createMirrorDataLossTrackingSessionNum,
ItemPointer persistentTid,
int64 *persistentSerialNum)
{
*relfilenodeOid = DatumGetObjectId(values[Anum_gp_relation_node_relfilenode_oid - 1]);
*segmentFileNum = DatumGetInt32(values[Anum_gp_relation_node_segment_file_num - 1]);
*createMirrorDataLossTrackingSessionNum = DatumGetInt64(values[Anum_gp_relation_node_create_mirror_data_loss_tracking_session_num - 1]);
*persistentTid = *((ItemPointer) DatumGetPointer(values[Anum_gp_relation_node_persistent_tid - 1]));
*persistentSerialNum = DatumGetInt64(values[Anum_gp_relation_node_persistent_serial_num - 1]);
}
//cs3223 fnv1 32bit
void hashMeth2(uint32 keyval, HashJoinTable hashtable) {
uint32 hashkey = 8192*bitvector_size;
long keyV;
unsigned int hash = OFFSET32;
int i=0;
//printf("size of keyval: %d\n", sizeof(keyval));
int numOctal = sizeof(keyval);
if (numOctal < 8)
keyV = DatumGetInt32(keyval);
else
keyV = DatumGetInt64(keyval);
for (i=0;i<numOctal;i++){
hash = hash ^ (keyV & 0x000000ff);
hash = hash * PRIME32;
keyV = keyV >> 8;
}
// Mapping method
hash = hash % hashkey;
setKbit(hashtable->bitvector, hash);
}
int HdfsFreeFileInfo(FsysName protocol, hdfsFileInfo * info, int numEntries)
{
FunctionCallInfoData fcinfo;
FileSystemUdfData fsysUdf;
FmgrInfo *fsysFunc = FsysInterfaceGetFunc(protocol, FSYS_FUNC_FREEFILEINFO);
fsysUdf.type = T_FileSystemFunctionData;
fsysUdf.fsys_fileinfo = info;
fsysUdf.fsys_fileinfonum = numEntries;
InitFunctionCallInfoData(/* FunctionCallInfoData */ fcinfo,
/* FmgrInfo */ fsysFunc,
/* nArgs */ 0,
/* Call Context */ (Node *) (&fsysUdf),
/* ResultSetInfo */ NULL);
Datum d = FunctionCallInvoke(&fcinfo);
return DatumGetInt32(d);
}
/*
* Checks if range overlaps with existing partitions.
* Returns TRUE if overlaps and FALSE otherwise.
*/
Datum
check_overlap(PG_FUNCTION_ARGS)
{
int parent_oid = DatumGetInt32(PG_GETARG_DATUM(0));
Datum p1 = PG_GETARG_DATUM(1);
Oid p1_type = get_fn_expr_argtype(fcinfo->flinfo, 1);
Datum p2 = PG_GETARG_DATUM(2);
Oid p2_type = get_fn_expr_argtype(fcinfo->flinfo, 2);
PartRelationInfo *prel;
RangeRelation *rangerel;
RangeEntry *ranges;
FmgrInfo cmp_func_1;
FmgrInfo cmp_func_2;
int i;
bool byVal;
prel = get_pathman_relation_info(parent_oid, NULL);
rangerel = get_pathman_range_relation(parent_oid, NULL);
if (!prel || !rangerel || prel->parttype != PT_RANGE)
PG_RETURN_NULL();
/* comparison functions */
cmp_func_1 = *get_cmp_func(p1_type, prel->atttype);
cmp_func_2 = *get_cmp_func(p2_type, prel->atttype);
byVal = rangerel->by_val;
ranges = (RangeEntry *) dsm_array_get_pointer(&rangerel->ranges);
for (i=0; i<rangerel->ranges.length; i++)
{
int c1 = FunctionCall2(&cmp_func_1, p1,
PATHMAN_GET_DATUM(ranges[i].max, byVal));
int c2 = FunctionCall2(&cmp_func_2, p2,
PATHMAN_GET_DATUM(ranges[i].min, byVal));
if (c1 < 0 && c2 > 0)
PG_RETURN_BOOL(true);
}
PG_RETURN_BOOL(false);
}
long getDocumentCount()
{
int spiResultCode;
Datum countDatum;
bool isNull;
long count;
SPI_connect();
spiResultCode = SPI_execute("SELECT COUNT(*)::integer FROM CardVectors", 1, 1);
if(spiResultCode == SPI_OK_SELECT)
{
countDatum = SPI_getbinval(SPI_tuptable->vals[0], SPI_tuptable->tupdesc, 1, &isNull);
count = DatumGetInt32(countDatum);
} else {
ereport(ERROR, (errmsg("Unable to query CardVectors table. Does it exist?")));
count = -1;
}
SPI_finish();
return count;
}
/*
* AuxiliaryProcKill() -- Cut-down version of ProcKill for auxiliary
* processes (bgwriter, etc). The PGPROC and sema are not released, only
* marked as not-in-use.
*/
static void
AuxiliaryProcKill(int code, Datum arg)
{
int proctype = DatumGetInt32(arg);
PGPROC *auxproc;
Assert(proctype >= 0 && proctype < NUM_AUXILIARY_PROCS);
auxproc = &AuxiliaryProcs[proctype];
Assert(MyProc == auxproc);
/* Release any LW locks I am holding (see notes above) */
LWLockReleaseAll();
/* Update shared estimate of spins_per_delay */
update_spins_per_delay();
if (code == 0 || code == 1)
{
MyProc->postmasterResetRequired = false;
}
/*
* If the parent process of this auxiliary process does not exist,
* we want to set the proc array entry free here. The postmaster may
* not own this process, so that it can't set the entry free. This
* could happen to the filerep subprocesses when the filerep main
* process dies unexpectedly.
*/
if (!ParentProcIsAlive())
{
MyProc->pid = 0;
MyProc->postmasterResetRequired = true;
}
/* PGPROC struct isn't mine anymore */
MyProc = NULL;
lockHolderProcPtr = NULL;
}
size_t append_datum(char* buf, Datum val, bool isnull, Oid typeoid)
{
HeapTuple typeTup;
Form_pg_type typeStruct;
FmgrInfo tmp_flinfo;
char *str;
size_t len;
typeTup = SearchSysCache(TYPEOID, ObjectIdGetDatum(typeoid), 0, 0, 0);
if (!HeapTupleIsValid(typeTup)) {
elog(ERROR, "Cache lookup failed for %u", typeoid);
}
typeStruct = (Form_pg_type)GETSTRUCT(typeTup);
if (typeStruct->typtype != 'b') {
// Non-basic type
elog(ERROR, "Don't support non-basic types (%s)",
format_type_be(typeoid));
}
fmgr_info_cxt(typeStruct->typoutput, &tmp_flinfo, CurTransactionContext);
ReleaseSysCache(typeTup);
if (!isnull) {
if (typeoid == INT4OID) {
*((int*)buf) = DatumGetInt32(val);
return 4;
}
SPI_push();
str = OutputFunctionCall(&tmp_flinfo, val);
SPI_pop();
len = strlen(str);
strncpy(buf, str, len);
return len;
}
return 0;
}
static void
fetch_restrict(HeapTuple *tuple, TupleDesc *tupdesc,
restrict_columns_t *restrict_columns, restrict_t *rest)
{
Datum binval;
bool isnull;
int t;
for(t=0; t<MAX_RULE_LENGTH;++t)
rest->via[t] = -1;
binval = SPI_getbinval(*tuple, *tupdesc, restrict_columns->target_id, &isnull);
if (isnull)
elog(ERROR, "target_id contains a null value");
rest->target_id = DatumGetInt32(binval);
binval = SPI_getbinval(*tuple, *tupdesc, restrict_columns->to_cost, &isnull);
if (isnull)
elog(ERROR, "to_cost contains a null value");
rest->to_cost = DatumGetFloat8(binval);
char *str = DatumGetCString(SPI_getvalue(*tuple, *tupdesc, restrict_columns->via_path));
//PGR_DBG("restriction: %f, %i, %s", rest->to_cost, rest->target_id, str);
if (str != NULL) {
char* pch = NULL;
int ci = 0;
pch = (char *)strtok (str," ,");
while (pch != NULL && ci < MAX_RULE_LENGTH)
{
rest->via[ci] = atoi(pch);
//PGR_DBG(" rest->via[%i]=%i", ci, rest->via[ci]);
ci++;
pch = (char *)strtok (NULL, " ,");
}
}
}
void
ParquetFetchSegFileInfo(AppendOnlyEntry *aoEntry, List *segfileinfos, Snapshot parquetMetaDataSnapshot)
{
Relation pg_parquetseg_rel;
TupleDesc pg_parquetseg_dsc;
HeapTuple tuple;
SysScanDesc parquetscan;
/*
* Since this function is called for insert operation,
* here we use RowExclusiveLock.
*/
pg_parquetseg_rel = heap_open(aoEntry->segrelid, RowExclusiveLock);
pg_parquetseg_dsc = RelationGetDescr(pg_parquetseg_rel);
parquetscan = systable_beginscan(pg_parquetseg_rel, InvalidOid, FALSE,
parquetMetaDataSnapshot, 0, NULL);
while (HeapTupleIsValid(tuple = systable_getnext(parquetscan)))
{
ListCell *lc;
int segno = DatumGetInt32(fastgetattr(tuple, Anum_pg_parquetseg_segno, pg_parquetseg_dsc, NULL));
foreach (lc, segfileinfos)
{
ResultRelSegFileInfo *segfileinfo = (ResultRelSegFileInfo *)lfirst(lc);
Assert(segfileinfo != NULL);
if (segfileinfo->segno == segno)
{
segfileinfo->numfiles = 1;
segfileinfo->tupcount = (int64)DatumGetFloat8(fastgetattr(tuple, Anum_pg_parquetseg_tupcount, pg_parquetseg_dsc, NULL));
segfileinfo->varblock = 0;
segfileinfo->eof = (int64 *)palloc(sizeof(int64));
segfileinfo->eof[0] = (int64)DatumGetFloat8(fastgetattr(tuple, Anum_pg_parquetseg_eof, pg_parquetseg_dsc, NULL));
segfileinfo->uncompressed_eof = (int64 *)palloc(sizeof(int64));
segfileinfo->uncompressed_eof[0] = (int64)DatumGetFloat8(fastgetattr(tuple, Anum_pg_parquetseg_eofuncompressed, pg_parquetseg_dsc, NULL));
break;
}
}
}
/*
* CleanupProcSignalState
* Remove current process from ProcSignalSlots
*
* This function is called via on_shmem_exit() during backend shutdown.
*/
static void
CleanupProcSignalState(int status, Datum arg)
{
int pss_idx = DatumGetInt32(arg);
volatile ProcSignalSlot *slot;
slot = &ProcSignalSlots[pss_idx - 1];
Assert(slot == MyProcSignalSlot);
/* sanity check */
if (slot->pss_pid != MyProcPid)
{
/*
* don't ERROR here. We're exiting anyway, and don't want to get into
* infinite loop trying to exit
*/
elog(LOG, "process %d releasing ProcSignal slot %d, but it contains %d",
MyProcPid, pss_idx, (int) slot->pss_pid);
return; /* XXX better to zero the slot anyway? */
}
slot->pss_pid = 0;
}
/*
* AuxiliaryProcKill() -- Cut-down version of ProcKill for auxiliary
* processes (bgwriter, etc). The PGPROC and sema are not released, only
* marked as not-in-use.
*/
static void
AuxiliaryProcKill(int code, Datum arg)
{
int proctype = DatumGetInt32(arg);
PGPROC *auxproc PG_USED_FOR_ASSERTS_ONLY;
PGPROC *proc;
Assert(proctype >= 0 && proctype < NUM_AUXILIARY_PROCS);
auxproc = &AuxiliaryProcs[proctype];
Assert(MyProc == auxproc);
/* Release any LW locks I am holding (see notes above) */
LWLockReleaseAll();
/*
* Reset MyLatch to the process local one. This is so that signal
* handlers et al can continue using the latch after the shared latch
* isn't ours anymore. After that clear MyProc and disown the shared
* latch.
*/
SwitchBackToLocalLatch();
proc = MyProc;
MyProc = NULL;
DisownLatch(&proc->procLatch);
SpinLockAcquire(ProcStructLock);
/* Mark auxiliary proc no longer in use */
proc->pid = 0;
/* Update shared estimate of spins_per_delay */
ProcGlobal->spins_per_delay = update_spins_per_delay(ProcGlobal->spins_per_delay);
SpinLockRelease(ProcStructLock);
}
static int
get_nnode(PlxFn *plx_fn, FunctionCallInfo fcinfo)
{
PlxQuery *plx_q = plx_fn->hash_query;
int err;
SPIPlanPtr plan;
Oid types[FUNC_MAX_ARGS];
Datum values[FUNC_MAX_ARGS];
char arg_nulls[FUNC_MAX_ARGS];
Datum val;
bool isnull;
int i;
if ((err = SPI_connect()) != SPI_OK_CONNECT)
plx_error(plx_fn, "SPI_connect: %s", SPI_result_code_string(err));
for (i = 0; i < plx_q->nargs; i++)
{
int idx = plx_q->plx_fn_arg_indexes[i];
types[i] = plx_fn->arg_types[idx]->oid;
values[i] = PG_GETARG_DATUM(idx);
}
plan = SPI_prepare(plx_q->sql->data, plx_q->nargs, types);
err = SPI_execute_plan(plan, values, arg_nulls, true, 0);
if (err != SPI_OK_SELECT)
plx_error(plx_fn,
"query '%s' failed: %s",
plx_q->sql->data,
SPI_result_code_string(err));
val = SPI_getbinval(SPI_tuptable->vals[0], SPI_tuptable->tupdesc, 1, &isnull);
err = SPI_finish();
if (err != SPI_OK_FINISH)
plx_error(plx_fn, "SPI_finish: %s", SPI_result_code_string(err));
return DatumGetInt32(val);
}
/*
* compare_keys
*/
static int
compare_keys(KeyedAggState *state, KeyValue *kv, Datum incoming, bool incoming_null, Oid collation, bool *result_null)
{
TypeCacheEntry *type = state->key_type;
FunctionCallInfoData cmp_fcinfo;
int result;
if (incoming_null || KV_KEY_IS_NULL(kv))
{
*result_null = true;
return 0;
}
InitFunctionCallInfoData(cmp_fcinfo, &type->cmp_proc_finfo, 2, collation, NULL, NULL);
cmp_fcinfo.arg[0] = kv->key;
cmp_fcinfo.argnull[0] = KV_KEY_IS_NULL(kv);
cmp_fcinfo.arg[1] = incoming;
cmp_fcinfo.argnull[1] = incoming_null;
result = DatumGetInt32(FunctionCallInvoke(&cmp_fcinfo));
*result_null = cmp_fcinfo.isnull;
return result;
}
int checkMeth2(uint32 keyval, HashJoinTable hashtable){
uint32 hashkey = 8192*bitvector_size;
long keyV;
unsigned int hash = OFFSET32;
int i=0;
//printf("size of keyval: %d\n", sizeof(keyval));
int numOctal = sizeof(keyval);
if (numOctal < 8)
keyV = DatumGetInt32(keyval);
else
keyV = DatumGetInt64(keyval);
for (i=0;i<numOctal;i++){
hash = hash ^ (keyV & 0x000000ff);
hash = hash * PRIME32;
keyV = keyV >> 8;
}
// Mapping method
hash = hash % hashkey;
if (checkKbit(hashtable->bitvector, hash) == 0){
//printf(" Filtered value: %d\n",GET_4_BYTES(keyval));
return 0;
}
return 1;
}
Datum
gin_numeric_cmp(PG_FUNCTION_ARGS)
{
Numeric a = (Numeric) PG_GETARG_POINTER(0);
Numeric b = (Numeric) PG_GETARG_POINTER(1);
int res = 0;
if (NUMERIC_IS_LEFTMOST(a))
{
res = (NUMERIC_IS_LEFTMOST(b)) ? 0 : -1;
}
else if (NUMERIC_IS_LEFTMOST(b))
{
res = 1;
}
else
{
res = DatumGetInt32(DirectFunctionCall2(numeric_cmp,
NumericGetDatum(a),
NumericGetDatum(b)));
}
PG_RETURN_INT32(res);
}
int HdfsDisconnect(FsysName protocol, hdfsFS fileSystem)
{
FunctionCallInfoData fcinfo;
FileSystemUdfData fsysUdf;
FmgrInfo *fsysFunc = FsysInterfaceGetFunc(protocol, FSYS_FUNC_DISCONNECT);
#ifdef USE_ASSERT_CHECKING
if (testmode_fault(gp_fsys_fault_inject_percent))
return -1;
#endif
fsysUdf.type = T_FileSystemFunctionData;
fsysUdf.fsys_hdfs = fileSystem;
InitFunctionCallInfoData(/* FunctionCallInfoData */ fcinfo,
/* FmgrInfo */ fsysFunc,
/* nArgs */ 0,
/* Call Context */ (Node *) (&fsysUdf),
/* ResultSetInfo */ NULL);
Datum d = FunctionCallInvoke(&fcinfo);
return DatumGetInt32(d);
}
/*
* For a newly inserted heap tid, check if an entry with this tid
* already exists in a unique index. If it does, abort the inserting
* transaction.
*/
static void
_bt_validate_tid(Relation irel, ItemPointer h_tid)
{
MIRROREDLOCK_BUFMGR_DECLARE;
BlockNumber blkno;
BlockNumber num_pages;
Buffer buf;
Page page;
BTPageOpaque opaque;
IndexTuple itup;
OffsetNumber maxoff,
minoff,
offnum;
elog(DEBUG1, "validating tid (%d,%d) for index (%s)",
ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
RelationGetRelationName(irel));
blkno = BTREE_METAPAGE + 1;
num_pages = RelationGetNumberOfBlocks(irel);
MIRROREDLOCK_BUFMGR_LOCK;
for (; blkno < num_pages; blkno++)
{
buf = ReadBuffer(irel, blkno);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (!PageIsNew(page))
_bt_checkpage(irel, buf);
if (P_ISLEAF(opaque))
{
minoff = P_FIRSTDATAKEY(opaque);
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = minoff;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
itup = (IndexTuple) PageGetItem(page,
PageGetItemId(page, offnum));
if (ItemPointerEquals(&itup->t_tid, h_tid))
{
Form_pg_attribute key_att = RelationGetDescr(irel)->attrs[0];
Oid key = InvalidOid;
bool isnull;
if (key_att->atttypid == OIDOID)
{
key = DatumGetInt32(
index_getattr(itup, 1, RelationGetDescr(irel), &isnull));
elog(ERROR, "found tid (%d,%d), %s (%d) already in index (%s)",
ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
NameStr(key_att->attname), key, RelationGetRelationName(irel));
}
else
{
elog(ERROR, "found tid (%d,%d) already in index (%s)",
ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
RelationGetRelationName(irel));
}
}
}
}
ReleaseBuffer(buf);
}
MIRROREDLOCK_BUFMGR_UNLOCK;
}
Datum
gp_inject_fault(PG_FUNCTION_ARGS)
{
char *faultName = TextDatumGetCString(PG_GETARG_DATUM(0));
char *type = TextDatumGetCString(PG_GETARG_DATUM(1));
char *ddlStatement = TextDatumGetCString(PG_GETARG_DATUM(2));
char *databaseName = TextDatumGetCString(PG_GETARG_DATUM(3));
char *tableName = TextDatumGetCString(PG_GETARG_DATUM(4));
int numOccurrences = PG_GETARG_INT32(5);
int sleepTimeSeconds = PG_GETARG_INT32(6);
int dbid = PG_GETARG_INT32(7);
StringInfo faultmsg = makeStringInfo();
/* Fast path if injecting fault in our postmaster. */
if (GpIdentity.dbid == dbid)
{
appendStringInfo(faultmsg, "%s\n%s\n%s\n%s\n%s\n%d\n%d\n",
faultName, type, ddlStatement, databaseName,
tableName, numOccurrences, sleepTimeSeconds);
int offset = 0;
char *response =
processTransitionRequest_faultInject(
faultmsg->data, &offset, faultmsg->len);
if (!response)
elog(ERROR, "failed to inject fault locally (dbid %d)", dbid);
if (strncmp(response, "Success:", strlen("Success:")) != 0)
elog(ERROR, "%s", response);
elog(NOTICE, "%s", response);
PG_RETURN_DATUM(true);
}
/* Obtain host and port of the requested dbid */
HeapTuple tuple;
Relation rel = heap_open(GpSegmentConfigRelationId, AccessShareLock);
ScanKeyData scankey;
SysScanDesc sscan;
ScanKeyInit(&scankey,
Anum_gp_segment_configuration_dbid,
BTEqualStrategyNumber, F_INT2EQ,
Int16GetDatum((int16) dbid));
sscan = systable_beginscan(rel, GpSegmentConfigDbidIndexId, true,
GetTransactionSnapshot(), 1, &scankey);
tuple = systable_getnext(sscan);
if (!HeapTupleIsValid(tuple))
elog(ERROR, "cannot find dbid %d", dbid);
bool isnull;
Datum datum = heap_getattr(tuple, Anum_gp_segment_configuration_hostname,
RelationGetDescr(rel), &isnull);
char *hostname;
if (!isnull)
hostname =
DatumGetCString(DirectFunctionCall1(textout, datum));
else
elog(ERROR, "hostname is null for dbid %d", dbid);
int port = DatumGetInt32(heap_getattr(tuple,
Anum_gp_segment_configuration_port,
RelationGetDescr(rel), &isnull));
systable_endscan(sscan);
heap_close(rel, NoLock);
struct addrinfo *addrList = NULL;
struct addrinfo hint;
int ret;
/* Initialize hint structure */
MemSet(&hint, 0, sizeof(hint));
hint.ai_socktype = SOCK_STREAM;
hint.ai_family = AF_UNSPEC;
char portStr[100];
if (snprintf(portStr, sizeof(portStr), "%d", port) >= sizeof(portStr))
elog(ERROR, "port number too long for dbid %d", dbid);
/* Use pg_getaddrinfo_all() to resolve the address */
ret = pg_getaddrinfo_all(hostname, portStr, &hint, &addrList);
if (ret || !addrList)
{
if (addrList)
pg_freeaddrinfo_all(hint.ai_family, addrList);
elog(ERROR, "could not translate host name \"%s\" to address: %s\n",
hostname, gai_strerror(ret));
}
PrimaryMirrorTransitionClientInfo client;
client.receivedDataCallbackFn = transitionReceivedDataFn;
client.errorLogFn = transitionErrorLogFn;
client.checkForNeedToExitFn = checkForNeedToExitFn;
transitionMsgErrors = makeStringInfo();
appendStringInfo(faultmsg, "%s\n%s\n%s\n%s\n%s\n%s\n%d\n%d\n",
"faultInject", faultName, type, ddlStatement,
databaseName, tableName, numOccurrences,
sleepTimeSeconds);
if (sendTransitionMessage(&client, addrList, faultmsg->data, faultmsg->len,
1 /* retries */, 60 /* timeout */) !=
TRANS_ERRCODE_SUCCESS)
{
pg_freeaddrinfo_all(hint.ai_family, addrList);
ereport(ERROR, (errmsg("failed to inject %s fault in dbid %d",
faultName, dbid),
errdetail("%s", transitionMsgErrors->data)));
}
pg_freeaddrinfo_all(hint.ai_family, addrList);
PG_RETURN_DATUM(BoolGetDatum(true));
}
/* ----------
* toast_fetch_datum_slice -
*
* Reconstruct a segment of a Datum from the chunks saved
* in the toast relation
* ----------
*/
static struct varlena *
toast_fetch_datum_slice(struct varlena * attr, int32 sliceoffset, int32 length)
{
Relation toastrel;
Relation toastidx;
ScanKeyData toastkey[3];
int nscankeys;
SysScanDesc toastscan;
HeapTuple ttup;
TupleDesc toasttupDesc;
struct varlena *result;
struct varatt_external toast_pointer;
int32 attrsize;
int32 residx;
int32 nextidx;
int numchunks;
int startchunk;
int endchunk;
int32 startoffset;
int32 endoffset;
int totalchunks;
Pointer chunk;
bool isnull;
char *chunkdata;
int32 chunksize;
int32 chcpystrt;
int32 chcpyend;
Assert(VARATT_IS_EXTERNAL(attr));
/* Must copy to access aligned fields */
VARATT_EXTERNAL_GET_POINTER(toast_pointer, attr);
/*
* It's nonsense to fetch slices of a compressed datum -- this isn't lo_*
* we can't return a compressed datum which is meaningful to toast later
*/
Assert(!VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer));
attrsize = toast_pointer.va_extsize;
totalchunks = ((attrsize - 1) / TOAST_MAX_CHUNK_SIZE) + 1;
if (sliceoffset >= attrsize)
{
sliceoffset = 0;
length = 0;
}
if (((sliceoffset + length) > attrsize) || length < 0)
length = attrsize - sliceoffset;
result = (struct varlena *) palloc(length + VARHDRSZ);
if (VARATT_EXTERNAL_IS_COMPRESSED(toast_pointer))
SET_VARSIZE_COMPRESSED(result, length + VARHDRSZ);
else
SET_VARSIZE(result, length + VARHDRSZ);
if (length == 0)
return result; /* Can save a lot of work at this point! */
startchunk = sliceoffset / TOAST_MAX_CHUNK_SIZE;
endchunk = (sliceoffset + length - 1) / TOAST_MAX_CHUNK_SIZE;
numchunks = (endchunk - startchunk) + 1;
startoffset = sliceoffset % TOAST_MAX_CHUNK_SIZE;
endoffset = (sliceoffset + length - 1) % TOAST_MAX_CHUNK_SIZE;
/*
* Open the toast relation and its index
*/
toastrel = heap_open(toast_pointer.va_toastrelid, AccessShareLock);
toasttupDesc = toastrel->rd_att;
toastidx = index_open(toastrel->rd_rel->reltoastidxid, AccessShareLock);
/*
* Setup a scan key to fetch from the index. This is either two keys or
* three depending on the number of chunks.
*/
ScanKeyInit(&toastkey[0],
(AttrNumber) 1,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(toast_pointer.va_valueid));
/*
* Use equality condition for one chunk, a range condition otherwise:
*/
if (numchunks == 1)
{
ScanKeyInit(&toastkey[1],
(AttrNumber) 2,
BTEqualStrategyNumber, F_INT4EQ,
Int32GetDatum(startchunk));
nscankeys = 2;
}
else
{
ScanKeyInit(&toastkey[1],
(AttrNumber) 2,
BTGreaterEqualStrategyNumber, F_INT4GE,
Int32GetDatum(startchunk));
ScanKeyInit(&toastkey[2],
(AttrNumber) 2,
BTLessEqualStrategyNumber, F_INT4LE,
Int32GetDatum(endchunk));
nscankeys = 3;
}
/*
* Read the chunks by index
*
* The index is on (valueid, chunkidx) so they will come in order
*/
nextidx = startchunk;
toastscan = systable_beginscan_ordered(toastrel, toastidx,
SnapshotToast, nscankeys, toastkey);
while ((ttup = systable_getnext_ordered(toastscan, ForwardScanDirection)) != NULL)
{
/*
* Have a chunk, extract the sequence number and the data
*/
residx = DatumGetInt32(fastgetattr(ttup, 2, toasttupDesc, &isnull));
Assert(!isnull);
chunk = DatumGetPointer(fastgetattr(ttup, 3, toasttupDesc, &isnull));
Assert(!isnull);
if (!VARATT_IS_EXTENDED(chunk))
{
chunksize = VARSIZE(chunk) - VARHDRSZ;
chunkdata = VARDATA(chunk);
}
else if (VARATT_IS_SHORT(chunk))
{
/* could happen due to heap_form_tuple doing its thing */
chunksize = VARSIZE_SHORT(chunk) - VARHDRSZ_SHORT;
chunkdata = VARDATA_SHORT(chunk);
}
else
{
/* should never happen */
elog(ERROR, "found toasted toast chunk for toast value %u in %s",
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
chunksize = 0; /* keep compiler quiet */
chunkdata = NULL;
}
/*
* Some checks on the data we've found
*/
if ((residx != nextidx) || (residx > endchunk) || (residx < startchunk))
elog(ERROR, "unexpected chunk number %d (expected %d) for toast value %u in %s",
residx, nextidx,
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
if (residx < totalchunks - 1)
{
if (chunksize != TOAST_MAX_CHUNK_SIZE)
elog(ERROR, "unexpected chunk size %d (expected %d) in chunk %d of %d for toast value %u in %s when fetching slice",
chunksize, (int) TOAST_MAX_CHUNK_SIZE,
residx, totalchunks,
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
}
else if (residx == totalchunks - 1)
{
if ((residx * TOAST_MAX_CHUNK_SIZE + chunksize) != attrsize)
elog(ERROR, "unexpected chunk size %d (expected %d) in final chunk %d for toast value %u in %s when fetching slice",
chunksize,
(int) (attrsize - residx * TOAST_MAX_CHUNK_SIZE),
residx,
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
}
else
elog(ERROR, "unexpected chunk number %d (out of range %d..%d) for toast value %u in %s",
residx,
0, totalchunks - 1,
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
/*
* Copy the data into proper place in our result
*/
chcpystrt = 0;
chcpyend = chunksize - 1;
if (residx == startchunk)
chcpystrt = startoffset;
if (residx == endchunk)
chcpyend = endoffset;
memcpy(VARDATA(result) +
(residx * TOAST_MAX_CHUNK_SIZE - sliceoffset) + chcpystrt,
chunkdata + chcpystrt,
(chcpyend - chcpystrt) + 1);
nextidx++;
}
/*
* Final checks that we successfully fetched the datum
*/
if (nextidx != (endchunk + 1))
elog(ERROR, "missing chunk number %d for toast value %u in %s",
nextidx,
toast_pointer.va_valueid,
RelationGetRelationName(toastrel));
/*
* End scan and close relations
*/
systable_endscan_ordered(toastscan);
index_close(toastidx, AccessShareLock);
heap_close(toastrel, AccessShareLock);
return result;
}
static int32
gbt_byteacmp(const bytea *a, const bytea *b)
{
return
(DatumGetInt32(DirectFunctionCall2(byteacmp, PointerGetDatum(a), PointerGetDatum(b))));
}
/*----------
* _bt_compare() -- Compare scankey to a particular tuple on the page.
*
* The passed scankey must be an insertion-type scankey (see nbtree/README),
* but it can omit the rightmost column(s) of the index.
*
* keysz: number of key conditions to be checked (might be less than the
* number of index columns!)
* page/offnum: location of btree item to be compared to.
*
* This routine returns:
* <0 if scankey < tuple at offnum;
* 0 if scankey == tuple at offnum;
* >0 if scankey > tuple at offnum.
* NULLs in the keys are treated as sortable values. Therefore
* "equality" does not necessarily mean that the item should be
* returned to the caller as a matching key!
*
* CRUCIAL NOTE: on a non-leaf page, the first data key is assumed to be
* "minus infinity": this routine will always claim it is less than the
* scankey. The actual key value stored (if any, which there probably isn't)
* does not matter. This convention allows us to implement the Lehman and
* Yao convention that the first down-link pointer is before the first key.
* See backend/access/nbtree/README for details.
*----------
*/
int32
_bt_compare(Relation rel,
int keysz,
ScanKey scankey,
Page page,
OffsetNumber offnum)
{
TupleDesc itupdesc = RelationGetDescr(rel);
BTPageOpaque opaque = (BTPageOpaque) PageGetSpecialPointer(page);
IndexTuple itup;
int i;
/*
* Force result ">" if target item is first data item on an internal page
* --- see NOTE above.
*/
if (!P_ISLEAF(opaque) && offnum == P_FIRSTDATAKEY(opaque))
return 1;
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offnum));
/*
* The scan key is set up with the attribute number associated with each
* term in the key. It is important that, if the index is multi-key, the
* scan contain the first k key attributes, and that they be in order. If
* you think about how multi-key ordering works, you'll understand why
* this is.
*
* We don't test for violation of this condition here, however. The
* initial setup for the index scan had better have gotten it right (see
* _bt_first).
*/
for (i = 1; i <= keysz; i++)
{
Datum datum;
bool isNull;
int32 result;
datum = index_getattr(itup, scankey->sk_attno, itupdesc, &isNull);
/* see comments about NULLs handling in btbuild */
if (scankey->sk_flags & SK_ISNULL) /* key is NULL */
{
if (isNull)
result = 0; /* NULL "=" NULL */
else if (scankey->sk_flags & SK_BT_NULLS_FIRST)
result = -1; /* NULL "<" NOT_NULL */
else
result = 1; /* NULL ">" NOT_NULL */
}
else if (isNull) /* key is NOT_NULL and item is NULL */
{
if (scankey->sk_flags & SK_BT_NULLS_FIRST)
result = 1; /* NOT_NULL ">" NULL */
else
result = -1; /* NOT_NULL "<" NULL */
}
else
{
/*
* The sk_func needs to be passed the index value as left arg and
* the sk_argument as right arg (they might be of different
* types). Since it is convenient for callers to think of
* _bt_compare as comparing the scankey to the index item, we have
* to flip the sign of the comparison result. (Unless it's a DESC
* column, in which case we *don't* flip the sign.)
*/
result = DatumGetInt32(FunctionCall2Coll(&scankey->sk_func,
scankey->sk_collation,
datum,
scankey->sk_argument));
if (!(scankey->sk_flags & SK_BT_DESC))
result = -result;
}
/* if the keys are unequal, return the difference */
if (result != 0)
return result;
scankey++;
}
/* if we get here, the keys are equal */
return 0;
}
/*
* record_cmp()
* Internal comparison function for records.
*
* Returns -1, 0 or 1
*
* Do not assume that the two inputs are exactly the same record type;
* for instance we might be comparing an anonymous ROW() construct against a
* named composite type. We will compare as long as they have the same number
* of non-dropped columns of the same types.
*/
static int
record_cmp(FunctionCallInfo fcinfo)
{
HeapTupleHeader record1 = PG_GETARG_HEAPTUPLEHEADER(0);
HeapTupleHeader record2 = PG_GETARG_HEAPTUPLEHEADER(1);
int result = 0;
Oid tupType1;
Oid tupType2;
int32 tupTypmod1;
int32 tupTypmod2;
TupleDesc tupdesc1;
TupleDesc tupdesc2;
HeapTupleData tuple1;
HeapTupleData tuple2;
int ncolumns1;
int ncolumns2;
RecordCompareData *my_extra;
int ncols;
Datum *values1;
Datum *values2;
bool *nulls1;
bool *nulls2;
int i1;
int i2;
int j;
/* Extract type info from the tuples */
tupType1 = HeapTupleHeaderGetTypeId(record1);
tupTypmod1 = HeapTupleHeaderGetTypMod(record1);
tupdesc1 = lookup_rowtype_tupdesc(tupType1, tupTypmod1);
ncolumns1 = tupdesc1->natts;
tupType2 = HeapTupleHeaderGetTypeId(record2);
tupTypmod2 = HeapTupleHeaderGetTypMod(record2);
tupdesc2 = lookup_rowtype_tupdesc(tupType2, tupTypmod2);
ncolumns2 = tupdesc2->natts;
/* Build temporary HeapTuple control structures */
tuple1.t_len = HeapTupleHeaderGetDatumLength(record1);
ItemPointerSetInvalid(&(tuple1.t_self));
tuple1.t_tableOid = InvalidOid;
tuple1.t_data = record1;
tuple2.t_len = HeapTupleHeaderGetDatumLength(record2);
ItemPointerSetInvalid(&(tuple2.t_self));
tuple2.t_tableOid = InvalidOid;
tuple2.t_data = record2;
/*
* We arrange to look up the needed comparison info just once per series
* of calls, assuming the record types don't change underneath us.
*/
ncols = Max(ncolumns1, ncolumns2);
my_extra = (RecordCompareData *) fcinfo->flinfo->fn_extra;
if (my_extra == NULL ||
my_extra->ncolumns < ncols)
{
fcinfo->flinfo->fn_extra =
MemoryContextAlloc(fcinfo->flinfo->fn_mcxt,
sizeof(RecordCompareData) - sizeof(ColumnCompareData)
+ ncols * sizeof(ColumnCompareData));
my_extra = (RecordCompareData *) fcinfo->flinfo->fn_extra;
my_extra->ncolumns = ncols;
my_extra->record1_type = InvalidOid;
my_extra->record1_typmod = 0;
my_extra->record2_type = InvalidOid;
my_extra->record2_typmod = 0;
}
if (my_extra->record1_type != tupType1 ||
my_extra->record1_typmod != tupTypmod1 ||
my_extra->record2_type != tupType2 ||
my_extra->record2_typmod != tupTypmod2)
{
MemSet(my_extra->columns, 0, ncols * sizeof(ColumnCompareData));
my_extra->record1_type = tupType1;
my_extra->record1_typmod = tupTypmod1;
my_extra->record2_type = tupType2;
my_extra->record2_typmod = tupTypmod2;
}
/* Break down the tuples into fields */
values1 = (Datum *) palloc(ncolumns1 * sizeof(Datum));
nulls1 = (bool *) palloc(ncolumns1 * sizeof(bool));
heap_deform_tuple(&tuple1, tupdesc1, values1, nulls1);
values2 = (Datum *) palloc(ncolumns2 * sizeof(Datum));
nulls2 = (bool *) palloc(ncolumns2 * sizeof(bool));
heap_deform_tuple(&tuple2, tupdesc2, values2, nulls2);
/*
* Scan corresponding columns, allowing for dropped columns in different
* places in the two rows. i1 and i2 are physical column indexes, j is
* the logical column index.
*/
i1 = i2 = j = 0;
while (i1 < ncolumns1 || i2 < ncolumns2)
{
TypeCacheEntry *typentry;
Oid collation;
FunctionCallInfoData locfcinfo;
int32 cmpresult;
/*
* Skip dropped columns
*/
if (i1 < ncolumns1 && tupdesc1->attrs[i1]->attisdropped)
{
i1++;
continue;
}
if (i2 < ncolumns2 && tupdesc2->attrs[i2]->attisdropped)
{
i2++;
continue;
}
if (i1 >= ncolumns1 || i2 >= ncolumns2)
break; /* we'll deal with mismatch below loop */
/*
* Have two matching columns, they must be same type
*/
if (tupdesc1->attrs[i1]->atttypid !=
tupdesc2->attrs[i2]->atttypid)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("cannot compare dissimilar column types %s and %s at record column %d",
format_type_be(tupdesc1->attrs[i1]->atttypid),
format_type_be(tupdesc2->attrs[i2]->atttypid),
j + 1)));
/*
* If they're not same collation, we don't complain here, but the
* comparison function might.
*/
collation = tupdesc1->attrs[i1]->attcollation;
if (collation != tupdesc2->attrs[i2]->attcollation)
collation = InvalidOid;
/*
* Lookup the comparison function if not done already
*/
typentry = my_extra->columns[j].typentry;
if (typentry == NULL ||
typentry->type_id != tupdesc1->attrs[i1]->atttypid)
{
typentry = lookup_type_cache(tupdesc1->attrs[i1]->atttypid,
TYPECACHE_CMP_PROC_FINFO);
if (!OidIsValid(typentry->cmp_proc_finfo.fn_oid))
ereport(ERROR,
(errcode(ERRCODE_UNDEFINED_FUNCTION),
errmsg("could not identify a comparison function for type %s",
format_type_be(typentry->type_id))));
my_extra->columns[j].typentry = typentry;
}
/*
* We consider two NULLs equal; NULL > not-NULL.
*/
if (!nulls1[i1] || !nulls2[i2])
{
if (nulls1[i1])
{
/* arg1 is greater than arg2 */
result = 1;
break;
}
if (nulls2[i2])
{
/* arg1 is less than arg2 */
result = -1;
break;
}
/* Compare the pair of elements */
InitFunctionCallInfoData(locfcinfo, &typentry->cmp_proc_finfo, 2,
collation, NULL, NULL);
locfcinfo.arg[0] = values1[i1];
locfcinfo.arg[1] = values2[i2];
locfcinfo.argnull[0] = false;
locfcinfo.argnull[1] = false;
locfcinfo.isnull = false;
cmpresult = DatumGetInt32(FunctionCallInvoke(&locfcinfo));
if (cmpresult < 0)
{
/* arg1 is less than arg2 */
result = -1;
break;
}
else if (cmpresult > 0)
{
/* arg1 is greater than arg2 */
result = 1;
break;
}
}
/* equal, so continue to next column */
i1++, i2++, j++;
}
/*
* If we didn't break out of the loop early, check for column count
* mismatch. (We do not report such mismatch if we found unequal column
* values; is that a feature or a bug?)
*/
if (result == 0)
{
if (i1 != ncolumns1 || i2 != ncolumns2)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("cannot compare record types with different numbers of columns")));
}
pfree(values1);
pfree(nulls1);
pfree(values2);
pfree(nulls2);
ReleaseTupleDesc(tupdesc1);
ReleaseTupleDesc(tupdesc2);
/* Avoid leaking memory when handed toasted input. */
PG_FREE_IF_COPY(record1, 0);
PG_FREE_IF_COPY(record2, 1);
return result;
}
/*----------
* _bt_compare() -- Compare scankey to a particular tuple on the page.
*
* keysz: number of key conditions to be checked (might be less than the
* total length of the scan key!)
* page/offnum: location of btree item to be compared to.
*
* This routine returns:
* <0 if scankey < tuple at offnum;
* 0 if scankey == tuple at offnum;
* >0 if scankey > tuple at offnum.
* NULLs in the keys are treated as sortable values. Therefore
* "equality" does not necessarily mean that the item should be
* returned to the caller as a matching key!
*
* CRUCIAL NOTE: on a non-leaf page, the first data key is assumed to be
* "minus infinity": this routine will always claim it is less than the
* scankey. The actual key value stored (if any, which there probably isn't)
* does not matter. This convention allows us to implement the Lehman and
* Yao convention that the first down-link pointer is before the first key.
* See backend/access/nbtree/README for details.
*----------
*/
int32
_bt_compare(Relation rel,
int keysz,
ScanKey scankey,
Page page,
OffsetNumber offnum)
{
TupleDesc itupdesc = RelationGetDescr(rel);
BTPageOpaque opaque = (BTPageOpaque) PageGetSpecialPointer(page);
BTItem btitem;
IndexTuple itup;
int i;
/*
* Force result ">" if target item is first data item on an internal
* page --- see NOTE above.
*/
if (!P_ISLEAF(opaque) && offnum == P_FIRSTDATAKEY(opaque))
return 1;
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, offnum));
itup = &(btitem->bti_itup);
/*
* The scan key is set up with the attribute number associated with
* each term in the key. It is important that, if the index is
* multi-key, the scan contain the first k key attributes, and that
* they be in order. If you think about how multi-key ordering works,
* you'll understand why this is.
*
* We don't test for violation of this condition here, however. The
* initial setup for the index scan had better have gotten it right
* (see _bt_first).
*/
for (i = 0; i < keysz; i++)
{
ScanKey entry = &scankey[i];
Datum datum;
bool isNull;
int32 result;
datum = index_getattr(itup, entry->sk_attno, itupdesc, &isNull);
/* see comments about NULLs handling in btbuild */
if (entry->sk_flags & SK_ISNULL) /* key is NULL */
{
if (isNull)
result = 0; /* NULL "=" NULL */
else
result = 1; /* NULL ">" NOT_NULL */
}
else if (isNull) /* key is NOT_NULL and item is NULL */
{
result = -1; /* NOT_NULL "<" NULL */
}
else
{
result = DatumGetInt32(FunctionCall2(&entry->sk_func,
entry->sk_argument,
datum));
}
/* if the keys are unequal, return the difference */
if (result != 0)
return result;
}
/* if we get here, the keys are equal */
return 0;
}
Datum /* have to return HeapTuple to Executor */
timetravel(PG_FUNCTION_ARGS)
{
TriggerData *trigdata = (TriggerData *) fcinfo->context;
Trigger *trigger; /* to get trigger name */
int argc;
char **args; /* arguments */
int attnum[MaxAttrNum]; /* fnumbers of start/stop columns */
Datum oldtimeon,
oldtimeoff;
Datum newtimeon,
newtimeoff,
newuser,
nulltext;
Datum *cvals; /* column values */
char *cnulls; /* column nulls */
char *relname; /* triggered relation name */
Relation rel; /* triggered relation */
HeapTuple trigtuple;
HeapTuple newtuple = NULL;
HeapTuple rettuple;
TupleDesc tupdesc; /* tuple description */
int natts; /* # of attributes */
EPlan *plan; /* prepared plan */
char ident[2 * NAMEDATALEN];
bool isnull; /* to know is some column NULL or not */
bool isinsert = false;
int ret;
int i;
/*
* Some checks first...
*/
/* Called by trigger manager ? */
if (!CALLED_AS_TRIGGER(fcinfo))
elog(ERROR, "timetravel: not fired by trigger manager");
/* Should be called for ROW trigger */
if (TRIGGER_FIRED_FOR_STATEMENT(trigdata->tg_event))
elog(ERROR, "timetravel: can't process STATEMENT events");
/* Should be called BEFORE */
if (TRIGGER_FIRED_AFTER(trigdata->tg_event))
elog(ERROR, "timetravel: must be fired before event");
/* INSERT ? */
if (TRIGGER_FIRED_BY_INSERT(trigdata->tg_event))
isinsert = true;
if (TRIGGER_FIRED_BY_UPDATE(trigdata->tg_event))
newtuple = trigdata->tg_newtuple;
trigtuple = trigdata->tg_trigtuple;
rel = trigdata->tg_relation;
relname = SPI_getrelname(rel);
/* check if TT is OFF for this relation */
if (0 == findTTStatus(relname))
{
/* OFF - nothing to do */
pfree(relname);
return PointerGetDatum((newtuple != NULL) ? newtuple : trigtuple);
}
trigger = trigdata->tg_trigger;
argc = trigger->tgnargs;
if (argc != MinAttrNum && argc != MaxAttrNum)
elog(ERROR, "timetravel (%s): invalid (!= %d or %d) number of arguments %d",
relname, MinAttrNum, MaxAttrNum, trigger->tgnargs);
args = trigger->tgargs;
tupdesc = rel->rd_att;
natts = tupdesc->natts;
for (i = 0; i < MinAttrNum; i++)
{
attnum[i] = SPI_fnumber(tupdesc, args[i]);
if (attnum[i] < 0)
elog(ERROR, "timetravel (%s): there is no attribute %s", relname, args[i]);
if (SPI_gettypeid(tupdesc, attnum[i]) != ABSTIMEOID)
elog(ERROR, "timetravel (%s): attribute %s must be of abstime type",
relname, args[i]);
}
for (; i < argc; i++)
{
attnum[i] = SPI_fnumber(tupdesc, args[i]);
if (attnum[i] < 0)
elog(ERROR, "timetravel (%s): there is no attribute %s", relname, args[i]);
if (SPI_gettypeid(tupdesc, attnum[i]) != TEXTOID)
elog(ERROR, "timetravel (%s): attribute %s must be of text type",
relname, args[i]);
}
/* create fields containing name */
newuser = DirectFunctionCall1(textin, CStringGetDatum(GetUserNameFromId(GetUserId())));
nulltext = (Datum) NULL;
if (isinsert)
{ /* INSERT */
int chnattrs = 0;
int chattrs[MaxAttrNum];
Datum newvals[MaxAttrNum];
char newnulls[MaxAttrNum];
oldtimeon = SPI_getbinval(trigtuple, tupdesc, attnum[a_time_on], &isnull);
if (isnull)
{
newvals[chnattrs] = GetCurrentAbsoluteTime();
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_time_on];
chnattrs++;
}
oldtimeoff = SPI_getbinval(trigtuple, tupdesc, attnum[a_time_off], &isnull);
if (isnull)
{
if ((chnattrs == 0 && DatumGetInt32(oldtimeon) >= NOEND_ABSTIME) ||
(chnattrs > 0 && DatumGetInt32(newvals[a_time_on]) >= NOEND_ABSTIME))
elog(ERROR, "timetravel (%s): %s is infinity", relname, args[a_time_on]);
newvals[chnattrs] = NOEND_ABSTIME;
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_time_off];
chnattrs++;
}
else
{
if ((chnattrs == 0 && DatumGetInt32(oldtimeon) > DatumGetInt32(oldtimeoff)) ||
(chnattrs > 0 && DatumGetInt32(newvals[a_time_on]) > DatumGetInt32(oldtimeoff)))
elog(ERROR, "timetravel (%s): %s gt %s", relname, args[a_time_on], args[a_time_off]);
}
pfree(relname);
if (chnattrs <= 0)
return PointerGetDatum(trigtuple);
if (argc == MaxAttrNum)
{
/* clear update_user value */
newvals[chnattrs] = nulltext;
newnulls[chnattrs] = 'n';
chattrs[chnattrs] = attnum[a_upd_user];
chnattrs++;
/* clear delete_user value */
newvals[chnattrs] = nulltext;
newnulls[chnattrs] = 'n';
chattrs[chnattrs] = attnum[a_del_user];
chnattrs++;
/* set insert_user value */
newvals[chnattrs] = newuser;
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_ins_user];
chnattrs++;
}
rettuple = SPI_modifytuple(rel, trigtuple, chnattrs, chattrs, newvals, newnulls);
return PointerGetDatum(rettuple);
/* end of INSERT */
}
/* UPDATE/DELETE: */
oldtimeon = SPI_getbinval(trigtuple, tupdesc, attnum[a_time_on], &isnull);
if (isnull)
elog(ERROR, "timetravel (%s): %s must be NOT NULL", relname, args[a_time_on]);
oldtimeoff = SPI_getbinval(trigtuple, tupdesc, attnum[a_time_off], &isnull);
if (isnull)
elog(ERROR, "timetravel (%s): %s must be NOT NULL", relname, args[a_time_off]);
/*
* If DELETE/UPDATE of tuple with stop_date neq INFINITY then say
* upper Executor to skip operation for this tuple
*/
if (newtuple != NULL)
{ /* UPDATE */
newtimeon = SPI_getbinval(newtuple, tupdesc, attnum[a_time_on], &isnull);
if (isnull)
elog(ERROR, "timetravel (%s): %s must be NOT NULL", relname, args[a_time_on]);
newtimeoff = SPI_getbinval(newtuple, tupdesc, attnum[a_time_off], &isnull);
if (isnull)
elog(ERROR, "timetravel (%s): %s must be NOT NULL", relname, args[a_time_off]);
if (oldtimeon != newtimeon || oldtimeoff != newtimeoff)
elog(ERROR, "timetravel (%s): you can't change %s and/or %s columns (use set_timetravel)",
relname, args[a_time_on], args[a_time_off]);
}
if (oldtimeoff != NOEND_ABSTIME)
{ /* current record is a deleted/updated
* record */
pfree(relname);
return PointerGetDatum(NULL);
}
newtimeoff = GetCurrentAbsoluteTime();
/* Connect to SPI manager */
if ((ret = SPI_connect()) < 0)
elog(ERROR, "timetravel (%s): SPI_connect returned %d", relname, ret);
/* Fetch tuple values and nulls */
cvals = (Datum *) palloc(natts * sizeof(Datum));
cnulls = (char *) palloc(natts * sizeof(char));
for (i = 0; i < natts; i++)
{
cvals[i] = SPI_getbinval(trigtuple, tupdesc, i + 1, &isnull);
cnulls[i] = (isnull) ? 'n' : ' ';
}
/* change date column(s) */
cvals[attnum[a_time_off] - 1] = newtimeoff; /* stop_date eq current
* date */
cnulls[attnum[a_time_off] - 1] = ' ';
if (!newtuple)
{ /* DELETE */
if (argc == MaxAttrNum)
{
cvals[attnum[a_del_user] - 1] = newuser; /* set delete user */
cnulls[attnum[a_del_user] - 1] = ' ';
}
}
/*
* Construct ident string as TriggerName $ TriggeredRelationId and try
* to find prepared execution plan.
*/
snprintf(ident, sizeof(ident), "%s$%u", trigger->tgname, rel->rd_id);
plan = find_plan(ident, &Plans, &nPlans);
/* if there is no plan ... */
if (plan->splan == NULL)
{
void *pplan;
Oid *ctypes;
char sql[8192];
char separ = ' ';
/* allocate ctypes for preparation */
ctypes = (Oid *) palloc(natts * sizeof(Oid));
/*
* Construct query: INSERT INTO _relation_ VALUES ($1, ...)
*/
snprintf(sql, sizeof(sql), "INSERT INTO %s VALUES (", relname);
for (i = 1; i <= natts; i++)
{
ctypes[i - 1] = SPI_gettypeid(tupdesc, i);
if (!(tupdesc->attrs[i - 1]->attisdropped)) /* skip dropped columns */
{
snprintf(sql + strlen(sql), sizeof(sql) - strlen(sql), "%c$%d", separ, i);
separ = ',';
}
}
snprintf(sql + strlen(sql), sizeof(sql) - strlen(sql), ")");
elog(DEBUG4, "timetravel (%s) update: sql: %s", relname, sql);
/* Prepare plan for query */
pplan = SPI_prepare(sql, natts, ctypes);
if (pplan == NULL)
elog(ERROR, "timetravel (%s): SPI_prepare returned %d", relname, SPI_result);
/*
* Remember that SPI_prepare places plan in current memory context
* - so, we have to save plan in Top memory context for latter
* use.
*/
pplan = SPI_saveplan(pplan);
if (pplan == NULL)
elog(ERROR, "timetravel (%s): SPI_saveplan returned %d", relname, SPI_result);
plan->splan = pplan;
}
/*
* Ok, execute prepared plan.
*/
ret = SPI_execp(plan->splan, cvals, cnulls, 0);
if (ret < 0)
elog(ERROR, "timetravel (%s): SPI_execp returned %d", relname, ret);
/* Tuple to return to upper Executor ... */
if (newtuple)
{ /* UPDATE */
int chnattrs = 0;
int chattrs[MaxAttrNum];
Datum newvals[MaxAttrNum];
char newnulls[MaxAttrNum];
newvals[chnattrs] = newtimeoff;
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_time_on];
chnattrs++;
newvals[chnattrs] = NOEND_ABSTIME;
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_time_off];
chnattrs++;
if (argc == MaxAttrNum)
{
/* set update_user value */
newvals[chnattrs] = newuser;
newnulls[chnattrs] = ' ';
chattrs[chnattrs] = attnum[a_upd_user];
chnattrs++;
/* clear delete_user value */
newvals[chnattrs] = nulltext;
newnulls[chnattrs] = 'n';
chattrs[chnattrs] = attnum[a_del_user];
chnattrs++;
/* set insert_user value */
newvals[chnattrs] = nulltext;
newnulls[chnattrs] = 'n';
chattrs[chnattrs] = attnum[a_ins_user];
chnattrs++;
}
rettuple = SPI_modifytuple(rel, newtuple, chnattrs, chattrs, newvals, newnulls);
/*
* SPI_copytuple allocates tmptuple in upper executor context -
* have to free allocation using SPI_pfree
*/
/* SPI_pfree(tmptuple); */
}
else
/* DELETE case */
rettuple = trigtuple;
SPI_finish(); /* don't forget say Bye to SPI mgr */
pfree(relname);
return PointerGetDatum(rettuple);
}
/*
* ntile
* compute an exact numeric value with scale 0 (zero),
* ranging from 1 (one) to n, per spec.
*/
Datum
window_ntile(PG_FUNCTION_ARGS)
{
WindowObject winobj = PG_WINDOW_OBJECT();
ntile_context *context;
context = (ntile_context *)
WinGetPartitionLocalMemory(winobj, sizeof(ntile_context));
if (context->ntile == 0)
{
/* first call */
int64 total;
int32 nbuckets;
bool isnull;
total = WinGetPartitionRowCount(winobj);
nbuckets = DatumGetInt32(WinGetFuncArgCurrent(winobj, 0, &isnull));
/*
* per spec: If NT is the null value, then the result is the null
* value.
*/
if (isnull)
PG_RETURN_NULL();
/*
* per spec: If NT is less than or equal to 0 (zero), then an
* exception condition is raised.
*/
if (nbuckets <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_ARGUMENT_FOR_NTILE),
errmsg("argument of ntile must be greater than zero")));
context->ntile = 1;
context->rows_per_bucket = 0;
context->boundary = total / nbuckets;
if (context->boundary <= 0)
context->boundary = 1;
else
{
/*
* If the total number is not divisible, add 1 row to leading
* buckets.
*/
context->remainder = total % nbuckets;
if (context->remainder != 0)
context->boundary++;
}
}
context->rows_per_bucket++;
if (context->boundary < context->rows_per_bucket)
{
/* ntile up */
if (context->remainder != 0 && context->ntile == context->remainder)
{
context->remainder = 0;
context->boundary -= 1;
}
context->ntile += 1;
context->rows_per_bucket = 1;
}
PG_RETURN_INT32(context->ntile);
}
