agora inbox for [email protected]  

help / color / mirror / Atom feed

[PATCH v21 2/8] Row pattern recognition patch (parse/analysis).
268+ messages / 2 participants
[nested] [flat]

* [PATCH v21 2/8] Row pattern recognition patch (parse/analysis).
@ 2024-08-26 04:32 Tatsuo Ishii <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Tatsuo Ishii @ 2024-08-26 04:32 UTC (permalink / raw)

---
 src/backend/parser/parse_agg.c    |   7 +
 src/backend/parser/parse_clause.c | 296 +++++++++++++++++++++++++++++-
 src/backend/parser/parse_expr.c   |   6 +
 src/backend/parser/parse_func.c   |   3 +
 4 files changed, 311 insertions(+), 1 deletion(-)

diff --git a/src/backend/parser/parse_agg.c b/src/backend/parser/parse_agg.c
index bee7d8346a..9bc22a836a 100644
--- a/src/backend/parser/parse_agg.c
+++ b/src/backend/parser/parse_agg.c
@@ -577,6 +577,10 @@ check_agglevels_and_constraints(ParseState *pstate, Node *expr)
 			errkind = true;
 			break;
 
+		case EXPR_KIND_RPR_DEFINE:
+			errkind = true;
+			break;
+
 			/*
 			 * There is intentionally no default: case here, so that the
 			 * compiler will warn if we add a new ParseExprKind without
@@ -967,6 +971,9 @@ transformWindowFuncCall(ParseState *pstate, WindowFunc *wfunc,
 		case EXPR_KIND_CYCLE_MARK:
 			errkind = true;
 			break;
+		case EXPR_KIND_RPR_DEFINE:
+			errkind = true;
+			break;
 
 			/*
 			 * There is intentionally no default: case here, so that the
diff --git a/src/backend/parser/parse_clause.c b/src/backend/parser/parse_clause.c
index 8118036495..9762dce81f 100644
--- a/src/backend/parser/parse_clause.c
+++ b/src/backend/parser/parse_clause.c
@@ -98,7 +98,14 @@ static WindowClause *findWindowClause(List *wclist, const char *name);
 static Node *transformFrameOffset(ParseState *pstate, int frameOptions,
 								  Oid rangeopfamily, Oid rangeopcintype, Oid *inRangeFunc,
 								  Node *clause);
-
+static void transformRPR(ParseState *pstate, WindowClause *wc, WindowDef *windef,
+						 List **targetlist);
+static List *transformDefineClause(ParseState *pstate, WindowClause *wc, WindowDef *windef,
+								   List **targetlist);
+static void transformPatternClause(ParseState *pstate, WindowClause *wc,
+								   WindowDef *windef);
+static List *transformMeasureClause(ParseState *pstate, WindowClause *wc,
+									WindowDef *windef);
 
 /*
  * transformFromClause -
@@ -2956,6 +2963,10 @@ transformWindowDefinitions(ParseState *pstate,
 											 rangeopfamily, rangeopcintype,
 											 &wc->endInRangeFunc,
 											 windef->endOffset);
+
+		/* Process Row Pattern Recognition related clauses */
+		transformRPR(pstate, wc, windef, targetlist);
+
 		wc->winref = winref;
 
 		result = lappend(result, wc);
@@ -3820,3 +3831,286 @@ transformFrameOffset(ParseState *pstate, int frameOptions,
 
 	return node;
 }
+
+/*
+ * transformRPR
+ *		Process Row Pattern Recognition related clauses
+ */
+static void
+transformRPR(ParseState *pstate, WindowClause *wc, WindowDef *windef,
+			 List **targetlist)
+{
+	/*
+	 * Window definition exists?
+	 */
+	if (windef == NULL)
+		return;
+
+	/*
+	 * Row Pattern Common Syntax clause exists?
+	 */
+	if (windef->rpCommonSyntax == NULL)
+		return;
+
+	/* Check Frame option. Frame must start at current row */
+	if ((wc->frameOptions & FRAMEOPTION_START_CURRENT_ROW) == 0)
+		ereport(ERROR,
+				(errcode(ERRCODE_SYNTAX_ERROR),
+				 errmsg("FRAME must start at current row when row patttern recognition is used")));
+
+	/* Transform AFTER MACH SKIP TO clause */
+	wc->rpSkipTo = windef->rpCommonSyntax->rpSkipTo;
+
+	/* Transform AFTER MACH SKIP TO variable */
+	wc->rpSkipVariable = windef->rpCommonSyntax->rpSkipVariable;
+
+	/* Transform SEEK or INITIAL clause */
+	wc->initial = windef->rpCommonSyntax->initial;
+
+	/* Transform DEFINE clause into list of TargetEntry's */
+	wc->defineClause = transformDefineClause(pstate, wc, windef, targetlist);
+
+	/* Check PATTERN clause and copy to patternClause */
+	transformPatternClause(pstate, wc, windef);
+
+	/* Transform MEASURE clause */
+	transformMeasureClause(pstate, wc, windef);
+}
+
+/*
+ * transformDefineClause Process DEFINE clause and transform ResTarget into
+ *		list of TargetEntry.
+ *
+ * XXX we only support column reference in row pattern definition search
+ * condition, e.g. "price". <row pattern definition variable name>.<column
+ * reference> is not supported, e.g. "A.price".
+ */
+static List *
+transformDefineClause(ParseState *pstate, WindowClause *wc, WindowDef *windef,
+					  List **targetlist)
+{
+	/* DEFINE variable name initials */
+	static char *defineVariableInitials = "abcdefghijklmnopqrstuvwxyz";
+
+	ListCell   *lc,
+			   *l;
+	ResTarget  *restarget,
+			   *r;
+	List	   *restargets;
+	List	   *defineClause;
+	char	   *name;
+	int			initialLen;
+	int			i;
+
+	/*
+	 * If Row Definition Common Syntax exists, DEFINE clause must exist. (the
+	 * raw parser should have already checked it.)
+	 */
+	Assert(windef->rpCommonSyntax->rpDefs != NULL);
+
+	/*
+	 * Check and add "A AS A IS TRUE" if pattern variable is missing in DEFINE
+	 * per the SQL standard.
+	 */
+	restargets = NIL;
+	foreach(lc, windef->rpCommonSyntax->rpPatterns)
+	{
+		A_Expr	   *a;
+		bool		found = false;
+
+		if (!IsA(lfirst(lc), A_Expr))
+			ereport(ERROR,
+					errmsg("node type is not A_Expr"));
+
+		a = (A_Expr *) lfirst(lc);
+		name = strVal(a->lexpr);
+
+		foreach(l, windef->rpCommonSyntax->rpDefs)
+		{
+			restarget = (ResTarget *) lfirst(l);
+
+			if (!strcmp(restarget->name, name))
+			{
+				found = true;
+				break;
+			}
+		}
+
+		if (!found)
+		{
+			/*
+			 * "name" is missing. So create "name AS name IS TRUE" ResTarget
+			 * node and add it to the temporary list.
+			 */
+			A_Const    *n;
+
+			restarget = makeNode(ResTarget);
+			n = makeNode(A_Const);
+			n->val.boolval.type = T_Boolean;
+			n->val.boolval.boolval = true;
+			n->location = -1;
+			restarget->name = pstrdup(name);
+			restarget->indirection = NIL;
+			restarget->val = (Node *) n;
+			restarget->location = -1;
+			restargets = lappend((List *) restargets, restarget);
+		}
+	}
+
+	if (list_length(restargets) >= 1)
+	{
+		/* add missing DEFINEs */
+		windef->rpCommonSyntax->rpDefs =
+			list_concat(windef->rpCommonSyntax->rpDefs, restargets);
+		list_free(restargets);
+	}
+
+	/*
+	 * Check for duplicate row pattern definition variables.  The standard
+	 * requires that no two row pattern definition variable names shall be
+	 * equivalent.
+	 */
+	restargets = NIL;
+	foreach(lc, windef->rpCommonSyntax->rpDefs)
+	{
+		restarget = (ResTarget *) lfirst(lc);
+		name = restarget->name;
+
+		/*
+		 * Add DEFINE expression (Restarget->val) to the targetlist as a
+		 * TargetEntry if it does not exist yet. Planner will add the column
+		 * ref var node to the outer plan's target list later on. This makes
+		 * DEFINE expression could access the outer tuple while evaluating
+		 * PATTERN.
+		 *
+		 * XXX: adding whole expressions of DEFINE to the plan.targetlist is
+		 * not so good, because it's not necessary to evalute the expression
+		 * in the target list while running the plan. We should extract the
+		 * var nodes only then add them to the plan.targetlist.
+		 */
+		findTargetlistEntrySQL99(pstate, (Node *) restarget->val,
+								 targetlist, EXPR_KIND_RPR_DEFINE);
+
+		/*
+		 * Make sure that the row pattern definition search condition is a
+		 * boolean expression.
+		 */
+		transformWhereClause(pstate, restarget->val,
+							 EXPR_KIND_RPR_DEFINE, "DEFINE");
+
+		foreach(l, restargets)
+		{
+			char	   *n;
+
+			r = (ResTarget *) lfirst(l);
+			n = r->name;
+
+			if (!strcmp(n, name))
+				ereport(ERROR,
+						(errcode(ERRCODE_SYNTAX_ERROR),
+						 errmsg("row pattern definition variable name \"%s\" appears more than once in DEFINE clause",
+								name),
+						 parser_errposition(pstate, exprLocation((Node *) r))));
+		}
+		restargets = lappend(restargets, restarget);
+	}
+	list_free(restargets);
+
+	/*
+	 * Create list of row pattern DEFINE variable name's initial. We assign
+	 * [a-z] to them (up to 26 variable names are allowed).
+	 */
+	restargets = NIL;
+	i = 0;
+	initialLen = strlen(defineVariableInitials);
+
+	foreach(lc, windef->rpCommonSyntax->rpDefs)
+	{
+		char		initial[2];
+
+		restarget = (ResTarget *) lfirst(lc);
+		name = restarget->name;
+
+		if (i >= initialLen)
+		{
+			ereport(ERROR,
+					(errcode(ERRCODE_SYNTAX_ERROR),
+					 errmsg("number of row pattern definition variable names exceeds %d",
+							initialLen),
+					 parser_errposition(pstate,
+										exprLocation((Node *) restarget))));
+		}
+		initial[0] = defineVariableInitials[i++];
+		initial[1] = '\0';
+		wc->defineInitial = lappend(wc->defineInitial,
+									makeString(pstrdup(initial)));
+	}
+
+	defineClause = transformTargetList(pstate, windef->rpCommonSyntax->rpDefs,
+									   EXPR_KIND_RPR_DEFINE);
+
+	/* mark column origins */
+	markTargetListOrigins(pstate, defineClause);
+
+	/* mark all nodes in the DEFINE clause tree with collation information */
+	assign_expr_collations(pstate, (Node *) defineClause);
+
+	return defineClause;
+}
+
+/*
+ * transformPatternClause
+ *		Process PATTERN clause and return PATTERN clause in the raw parse tree
+ */
+static void
+transformPatternClause(ParseState *pstate, WindowClause *wc,
+					   WindowDef *windef)
+{
+	ListCell   *lc;
+
+	/*
+	 * Row Pattern Common Syntax clause exists?
+	 */
+	if (windef->rpCommonSyntax == NULL)
+		return;
+
+	wc->patternVariable = NIL;
+	wc->patternRegexp = NIL;
+	foreach(lc, windef->rpCommonSyntax->rpPatterns)
+	{
+		A_Expr	   *a;
+		char	   *name;
+		char	   *regexp;
+
+		if (!IsA(lfirst(lc), A_Expr))
+			ereport(ERROR,
+					errmsg("node type is not A_Expr"));
+
+		a = (A_Expr *) lfirst(lc);
+		name = strVal(a->lexpr);
+
+		wc->patternVariable = lappend(wc->patternVariable, makeString(pstrdup(name)));
+		regexp = strVal(lfirst(list_head(a->name)));
+
+		wc->patternRegexp = lappend(wc->patternRegexp, makeString(pstrdup(regexp)));
+	}
+}
+
+/*
+ * transformMeasureClause
+ *		Process MEASURE clause
+ *	XXX MEASURE clause is not supported yet
+ */
+static List *
+transformMeasureClause(ParseState *pstate, WindowClause *wc,
+					   WindowDef *windef)
+{
+	if (windef->rowPatternMeasures == NIL)
+		return NIL;
+
+	ereport(ERROR,
+			(errcode(ERRCODE_SYNTAX_ERROR),
+			 errmsg("%s", "MEASURE clause is not supported yet"),
+			 parser_errposition(pstate, exprLocation((Node *) windef->rowPatternMeasures))));
+	return NIL;
+}
diff --git a/src/backend/parser/parse_expr.c b/src/backend/parser/parse_expr.c
index 56e413da9f..c187b3278d 100644
--- a/src/backend/parser/parse_expr.c
+++ b/src/backend/parser/parse_expr.c
@@ -577,6 +577,7 @@ transformColumnRef(ParseState *pstate, ColumnRef *cref)
 		case EXPR_KIND_COPY_WHERE:
 		case EXPR_KIND_GENERATED_COLUMN:
 		case EXPR_KIND_CYCLE_MARK:
+		case EXPR_KIND_RPR_DEFINE:
 			/* okay */
 			break;
 
@@ -1860,6 +1861,9 @@ transformSubLink(ParseState *pstate, SubLink *sublink)
 		case EXPR_KIND_GENERATED_COLUMN:
 			err = _("cannot use subquery in column generation expression");
 			break;
+		case EXPR_KIND_RPR_DEFINE:
+			err = _("cannot use subquery in DEFINE expression");
+			break;
 
 			/*
 			 * There is intentionally no default: case here, so that the
@@ -3199,6 +3203,8 @@ ParseExprKindName(ParseExprKind exprKind)
 			return "GENERATED AS";
 		case EXPR_KIND_CYCLE_MARK:
 			return "CYCLE";
+		case EXPR_KIND_RPR_DEFINE:
+			return "DEFINE";
 
 			/*
 			 * There is intentionally no default: case here, so that the
diff --git a/src/backend/parser/parse_func.c b/src/backend/parser/parse_func.c
index 9b23344a3b..4c482abb30 100644
--- a/src/backend/parser/parse_func.c
+++ b/src/backend/parser/parse_func.c
@@ -2658,6 +2658,9 @@ check_srf_call_placement(ParseState *pstate, Node *last_srf, int location)
 		case EXPR_KIND_CYCLE_MARK:
 			errkind = true;
 			break;
+		case EXPR_KIND_RPR_DEFINE:
+			errkind = true;
+			break;
 
 			/*
 			 * There is intentionally no default: case here, so that the
-- 
2.25.1


----Next_Part(Mon_Aug_26_13_39_47_2024_878)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v21-0003-Row-pattern-recognition-patch-rewriter.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread

* [PATCH v12 2/3] Use time stamp counter to measure time on Linux/x86
@ 2025-07-26 00:57 Lukas Fittl <[email protected]>
  0 siblings, 0 replies; 268+ messages in thread

From: Lukas Fittl @ 2025-07-26 00:57 UTC (permalink / raw)

We switch to using the time stamp counter (TSC) instead of clock_gettime()
to reduce overhead of EXPLAIN (ANALYZE, TIME ON). Tests showed that runtime
is reduced by around 10% for queries moving lots of rows through the plan.

For now this is only enabled on Linux/x86, in case the system clocksource is
reported as TSC. Relying on the Linux kernel simplifies the logic to detect
if the present TSC is usable (frequency invariant, synchronized between
sockets, etc.). In all other cases we fallback to clock_gettime().

Note, that we intentionally use RDTSC in the fast paths, rather than RDTSCP.
RDTSCP waits for outstanding instructions to retire on out-of-order CPUs.
This adds noticably for little benefit in the typical InstrStartNode() /
InstrStopNode() use case. The macro to be used in such cases is called
INSTR_TIME_SET_CURRENT_FAST(). The original macro INSTR_TIME_SET_CURRENT()
uses RDTSCP and is supposed to be used when precision is more important
than performance.

Author: David Geier <[email protected]>
Author: Andres Freund <[email protected]>
Author: Lukas Fittl <[email protected]>
Reviewed-by:
Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de
---
 src/backend/access/heap/vacuumlazy.c |   4 +-
 src/backend/executor/instrument.c    |  12 +-
 src/backend/utils/init/postinit.c    |   3 +
 src/bin/pgbench/pgbench.c            |   3 +
 src/bin/psql/startup.c               |   4 +
 src/common/Makefile                  |   1 +
 src/common/instr_time.c              | 206 +++++++++++++++++++++++++++
 src/common/meson.build               |   1 +
 src/include/portability/instr_time.h | 136 +++++++++++++++---
 9 files changed, 348 insertions(+), 22 deletions(-)
 create mode 100644 src/common/instr_time.c

diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c
index d2b031fdd06..5027048cac4 100644
--- a/src/backend/access/heap/vacuumlazy.c
+++ b/src/backend/access/heap/vacuumlazy.c
@@ -3409,8 +3409,8 @@ count_nondeletable_pages(LVRelState *vacrel, bool *lock_waiter_detected)
 			INSTR_TIME_SET_CURRENT(currenttime);
 			elapsed = currenttime;
 			INSTR_TIME_SUBTRACT(elapsed, starttime);
-			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
-				>= VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
+			if (INSTR_TIME_GET_MILLISEC(elapsed) >=
+				VACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
 			{
 				if (LockHasWaitersRelation(vacrel->rel, AccessExclusiveLock))
 				{
diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c
index 56e635f4700..01f67c5d972 100644
--- a/src/backend/executor/instrument.c
+++ b/src/backend/executor/instrument.c
@@ -67,9 +67,13 @@ InstrInit(Instrumentation *instr, int instrument_options)
 void
 InstrStartNode(Instrumentation *instr)
 {
-	if (instr->need_timer &&
-		!INSTR_TIME_SET_CURRENT_LAZY(instr->starttime))
-		elog(ERROR, "InstrStartNode called twice in a row");
+	if (instr->need_timer)
+	{
+		if (!INSTR_TIME_IS_ZERO(instr->starttime))
+			elog(ERROR, "InstrStartNode called twice in a row");
+		else
+			INSTR_TIME_SET_CURRENT_FAST(instr->starttime);
+	}
 
 	/* save buffer usage totals at node entry, if needed */
 	if (instr->need_bufusage)
@@ -95,7 +99,7 @@ InstrStopNode(Instrumentation *instr, double nTuples)
 		if (INSTR_TIME_IS_ZERO(instr->starttime))
 			elog(ERROR, "InstrStopNode called without start");
 
-		INSTR_TIME_SET_CURRENT(endtime);
+		INSTR_TIME_SET_CURRENT_FAST(endtime);
 		INSTR_TIME_ACCUM_DIFF(instr->counter, endtime, instr->starttime);
 
 		INSTR_TIME_SET_ZERO(instr->starttime);
diff --git a/src/backend/utils/init/postinit.c b/src/backend/utils/init/postinit.c
index 641e535a73c..d573409903b 100644
--- a/src/backend/utils/init/postinit.c
+++ b/src/backend/utils/init/postinit.c
@@ -810,6 +810,9 @@ InitPostgres(const char *in_dbname, Oid dboid,
 	/* Initialize portal manager */
 	EnablePortalManager();
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/*
 	 * Load relcache entries for the shared system catalogs.  This must create
 	 * at least entries for pg_database and catalogs used for authentication.
diff --git a/src/bin/pgbench/pgbench.c b/src/bin/pgbench/pgbench.c
index 1515ed405ba..79bef2d2aec 100644
--- a/src/bin/pgbench/pgbench.c
+++ b/src/bin/pgbench/pgbench.c
@@ -7290,6 +7290,9 @@ main(int argc, char **argv)
 		initRandomState(&state[i].cs_func_rs);
 	}
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	/* opening connection... */
 	con = doConnect();
 	if (con == NULL)
diff --git a/src/bin/psql/startup.c b/src/bin/psql/startup.c
index 249b6aa5169..d615df593c7 100644
--- a/src/bin/psql/startup.c
+++ b/src/bin/psql/startup.c
@@ -24,6 +24,7 @@
 #include "help.h"
 #include "input.h"
 #include "mainloop.h"
+#include "portability/instr_time.h"
 #include "settings.h"
 
 /*
@@ -327,6 +328,9 @@ main(int argc, char *argv[])
 
 	PQsetNoticeProcessor(pset.db, NoticeProcessor, NULL);
 
+	/* initialize high-precision interval timing */
+	INSTR_TIME_INITIALIZE();
+
 	SyncVariables();
 
 	if (options.list_dbs)
diff --git a/src/common/Makefile b/src/common/Makefile
index 2c720caa509..1a2fbbe887f 100644
--- a/src/common/Makefile
+++ b/src/common/Makefile
@@ -59,6 +59,7 @@ OBJS_COMMON = \
 	file_perm.o \
 	file_utils.o \
 	hashfn.o \
+	instr_time.o \
 	ip.o \
 	jsonapi.o \
 	keywords.o \
diff --git a/src/common/instr_time.c b/src/common/instr_time.c
new file mode 100644
index 00000000000..fdf47699f20
--- /dev/null
+++ b/src/common/instr_time.c
@@ -0,0 +1,206 @@
+/*-------------------------------------------------------------------------
+ *
+ * instr_time.c
+ *	   Non-inline parts of the portable high-precision interval timing
+ *	 implementation
+ *
+ * Portions Copyright (c) 2025, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/port/instr_time.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#if defined(HAVE__GET_CPUID) || (defined(HAVE__CPUIDEX) && !defined(_MSC_VER))
+#include <cpuid.h>
+#endif
+
+#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER))
+#include <intrin.h>
+#endif
+
+#include "portability/instr_time.h"
+
+#ifndef WIN32
+/*
+ * Stores what the number of cycles needs to be multiplied with to end up
+ * with nanoseconds using integer math. See comment in pg_initialize_rdtsc()
+ * for more details.
+ *
+ * By default assume we are using clock_gettime() as a fallback which uses
+ * nanoseconds as ticks. Hence, we set the multiplier to the precision scalar
+ * so that the division in INSTR_TIME_GET_NANOSEC() won't change the nanoseconds.
+ *
+ * When using the RDTSC instruction directly this is filled in during initialization
+ * based on the relevant CPUID fields.
+ */
+int64		ticks_per_ns_scaled = TICKS_TO_NS_PRECISION;
+int64		ticks_per_sec = NS_PER_S;
+int64		max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION;
+
+#if defined(__x86_64__) && defined(__linux__)
+/*
+ * Indicates if RDTSC can be used (Linux/x86 only, when OS uses TSC clocksource)
+ */
+bool		has_rdtsc = false;
+
+/*
+ * Indicates if RDTSCP can be used. True if RDTSC can be used and RDTSCP is available.
+ */
+bool		has_rdtscp = false;
+
+#define CPUID_HYPERVISOR_VMWARE(words) (words[1] == 0x61774d56 && words[2] == 0x4d566572 && words[3] == 0x65726177) /* VMwareVMware */
+#define CPUID_HYPERVISOR_KVM(words) (words[1] == 0x4b4d564b && words[2] == 0x564b4d56 && words[3] == 0x0000004d)	/* KVMKVMKVM */
+
+static bool
+get_tsc_frequency_khz(uint32 *tsc_freq)
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x15, &r[0] /* denominator */ , &r[1] /* numerator */ , &r[2] /* hz */ , &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x15);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[2] > 0)
+	{
+		if (r[0] == 0 || r[1] == 0)
+			return false;
+
+		*tsc_freq = r[2] / 1000 * r[1] / r[0];
+		return true;
+	}
+
+	/* Some CPUs only report frequency in 16H */
+
+#if defined(HAVE__GET_CPUID)
+	__get_cpuid(0x16, &r[0] /* base_mhz */ , &r[1], &r[2], &r[3]);
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x16);
+#else
+#error cpuid instruction not available
+#endif
+
+	if (r[0] > 0)
+	{
+		*tsc_freq = r[0] * 1000;
+		return true;
+	}
+
+	/*
+	 * Check if we have a KVM or VMware Hypervisor passing down TSC frequency
+	 * to us in a guest VM
+	 *
+	 * Note that accessing the 0x40000000 leaf for Hypervisor info requires
+	 * use of __cpuidex to set ECX to 0. The similar __get_cpuid_count
+	 * function does not work as expected since it contains a check for
+	 * __get_cpuid_max, which has been observed to be lower than the special
+	 * Hypervisor leaf.
+	 */
+#if defined(HAVE__CPUIDEX)
+	__cpuidex((int32 *) r, 0x40000000, 0);
+	if (r[0] >= 0x40000010 && (CPUID_HYPERVISOR_VMWARE(r) || CPUID_HYPERVISOR_KVM(r)))
+	{
+		__cpuidex((int32 *) r, 0x40000010, 0);
+		if (r[0] > 0)
+		{
+			*tsc_freq = r[0];
+			return true;
+		}
+	}
+#endif
+
+	return false;
+}
+
+static bool
+is_rdtscp_available()
+{
+	uint32		r[4] = {0, 0, 0, 0};
+
+#if defined(HAVE__GET_CPUID)
+	if (!__get_cpuid(0x80000001, &r[0], &r[1], &r[2], &r[3]))
+		return false;
+#elif defined(HAVE__CPUID)
+	__cpuid(r, 0x80000001);
+#else
+#error cpuid instruction not available
+#endif
+
+	return (r[3] & (1 << 27)) != 0;
+}
+
+/*
+ * Decide whether we use the RDTSC instruction at runtime, for Linux/x86,
+ * instead of incurring the overhead of a full clock_gettime() call.
+ *
+ * This can't be reliably determined at compile time, since the
+ * availability of an "invariant" TSC (that is not affected by CPU
+ * frequency changes) is dependent on the CPU architecture. Additionally,
+ * there are cases where TSC availability is impacted by virtualization,
+ * where a simple cpuid feature check would not be enough.
+ *
+ * Since Linux already does a significant amount of work to determine
+ * whether TSC is a viable clock source, decide based on that.
+ */
+void
+pg_initialize_rdtsc(void)
+{
+	FILE	   *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r");
+
+	if (fp)
+	{
+		char		buf[128];
+
+		if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0)
+		{
+			/*
+			 * Compute baseline CPU peformance, determines speed at which
+			 * RDTSC advances.
+			 */
+			uint32		tsc_freq;
+
+			if (get_tsc_frequency_khz(&tsc_freq))
+			{
+				/*
+				 * Ticks to nanoseconds conversion requires floating point
+				 * math because because:
+				 *
+				 * sec = ticks / frequency_hz ns  = ticks / frequency_hz *
+				 * 1,000,000,000 ns  = ticks * (1,000,000,000 / frequency_hz)
+				 * ns  = ticks * (1,000,000 / frequency_khz) <-- now in
+				 * kilohertz
+				 *
+				 * Here, 'ns' is usually a floating number. For example for a
+				 * 2.5 GHz CPU the scaling factor becomes 1,000,000 /
+				 * 2,500,000 = 1.2.
+				 *
+				 * To be able to use integer math we work around the lack of
+				 * precision. We first scale the integer up and after the
+				 * multiplication by the number of ticks in
+				 * INSTR_TIME_GET_NANOSEC() we divide again by the same value.
+				 * We picked the scaler such that it provides enough precision
+				 * and is a power-of-two which allows for shifting instead of
+				 * doing an integer division.
+				 */
+				ticks_per_ns_scaled = INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq;
+				ticks_per_sec = tsc_freq * 1000;	/* KHz->Hz */
+				max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled;
+
+				has_rdtsc = true;
+				has_rdtscp = is_rdtscp_available();
+			}
+		}
+
+		fclose(fp);
+	}
+}
+#endif							/* defined(__x86_64__) && defined(__linux__) */
+
+#endif							/* WIN32 */
diff --git a/src/common/meson.build b/src/common/meson.build
index 1540ba67cca..62b90b3e609 100644
--- a/src/common/meson.build
+++ b/src/common/meson.build
@@ -13,6 +13,7 @@ common_sources = files(
   'file_perm.c',
   'file_utils.c',
   'hashfn.c',
+  'instr_time.c',
   'ip.c',
   'jsonapi.c',
   'keywords.c',
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index f71a851b18d..e2e339a0c4f 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -4,9 +4,11 @@
  *	  portable high-precision interval timing
  *
  * This file provides an abstraction layer to hide portability issues in
- * interval timing.  On Unix we use clock_gettime(), and on Windows we use
- * QueryPerformanceCounter().  These macros also give some breathing room to
- * use other high-precision-timing APIs.
+ * interval timing. On Linux/x86 we use the rdtsc instruction when a TSC
+ * clocksource is also used on the host OS.  Otherwise, and on other
+ * Unix-like systems we use clock_gettime() and on Windows we use
+ * QueryPerformanceCounter(). These macros also give some breathing
+ * room to use other high-precision-timing APIs.
  *
  * The basic data type is instr_time, which all callers should treat as an
  * opaque typedef.  instr_time can store either an absolute time (of
@@ -17,10 +19,11 @@
  *
  * INSTR_TIME_SET_ZERO(t)			set t to zero (memset is acceptable too)
  *
- * INSTR_TIME_SET_CURRENT(t)		set t to current time
+ * INSTR_TIME_SET_CURRENT_FAST(t)	set t to current time without waiting
+ * 									for instructions in out-of-order window
  *
- * INSTR_TIME_SET_CURRENT_LAZY(t)	set t to current time if t is zero,
- *									evaluates to whether t changed
+ * INSTR_TIME_SET_CURRENT(t)		set t to current time while waiting for
+ * 									instructions in OOO to retire
  *
  * INSTR_TIME_ADD(x, y)				x += y
  *
@@ -81,6 +84,15 @@ typedef struct instr_time
 
 #ifndef WIN32
 
+/*
+ * Make sure this is a power-of-two, so that the compiler can turn the
+ * multiplications and divisions into shifts.
+ */
+#define TICKS_TO_NS_PRECISION (1<<14)
+
+extern int64 ticks_per_ns_scaled;
+extern int64 ticks_per_sec;
+extern int64 max_ticks_no_overflow;
 
 /* Use clock_gettime() */
 
@@ -106,9 +118,18 @@ typedef struct instr_time
 #define PG_INSTR_CLOCK	CLOCK_REALTIME
 #endif
 
-/* helper for INSTR_TIME_SET_CURRENT */
+#if defined(__x86_64__) && defined(__linux__)
+#include <x86intrin.h>
+#include <cpuid.h>
+
+extern bool has_rdtsc;
+extern bool has_rdtscp;
+
+extern void pg_initialize_rdtsc(void);
+#endif
+
 static inline instr_time
-pg_clock_gettime_ns(void)
+pg_clock_gettime(void)
 {
 	instr_time	now;
 	struct timespec tmp;
@@ -119,11 +140,94 @@ pg_clock_gettime_ns(void)
 	return now;
 }
 
+static inline instr_time
+pg_get_ticks_fast(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtsc)
+	{
+		instr_time	now;
+
+		now.ticks = __rdtsc();
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline instr_time
+pg_get_ticks(void)
+{
+#if defined(__x86_64__) && defined(__linux__)
+	if (has_rdtscp)
+	{
+		instr_time	now;
+		uint32		unused;
+
+		now.ticks = __rdtscp(&unused);
+		return now;
+	}
+#endif
+
+	return pg_clock_gettime();
+}
+
+static inline int64_t
+pg_ticks_to_ns(instr_time t)
+{
+	/*
+	 * Would multiplication overflow? If so perform computation in two parts.
+	 * Check overflow without actually overflowing via: a * b > max <=> a >
+	 * max / b
+	 */
+	int64		ns = 0;
+
+	if (unlikely(t.ticks > max_ticks_no_overflow))
+	{
+		/*
+		 * Compute how often the maximum number of ticks fits completely into
+		 * the number of elapsed ticks and convert that number into
+		 * nanoseconds. Then multiply by the count to arrive at the final
+		 * value. In a 2nd step we adjust the number of elapsed ticks and
+		 * convert the remaining ticks.
+		 */
+		int64		count = t.ticks / max_ticks_no_overflow;
+		int64		max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+
+		ns = max_ns * count;
+
+		/*
+		 * Subtract the ticks that we now already accounted for, so that they
+		 * don't get counted twice.
+		 */
+		t.ticks -= count * max_ticks_no_overflow;
+		Assert(t.ticks >= 0);
+	}
+
+	ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+	return ns;
+}
+
+static inline void
+pg_initialize_get_ticks()
+{
+#if defined(__x86_64__) && defined(__linux__)
+	pg_initialize_rdtsc();
+#endif
+}
+
+#define INSTR_TIME_INITIALIZE() \
+	pg_initialize_get_ticks()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_get_ticks_fast())
+
 #define INSTR_TIME_SET_CURRENT(t) \
-	((t) = pg_clock_gettime_ns())
+	((t) = pg_get_ticks())
 
 #define INSTR_TIME_GET_NANOSEC(t) \
-	((int64) (t).ticks)
+	pg_ticks_to_ns(t)
 
 
 #else							/* WIN32 */
@@ -131,7 +235,7 @@ pg_clock_gettime_ns(void)
 
 /* Use QueryPerformanceCounter() */
 
-/* helper for INSTR_TIME_SET_CURRENT */
+/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
 static inline instr_time
 pg_query_performance_counter(void)
 {
@@ -153,6 +257,11 @@ GetTimerFrequency(void)
 	return (double) f.QuadPart;
 }
 
+#define INSTR_TIME_INITIALIZE()
+
+#define INSTR_TIME_SET_CURRENT_FAST(t) \
+	((t) = pg_query_performance_counter())
+
 #define INSTR_TIME_SET_CURRENT(t) \
 	((t) = pg_query_performance_counter())
 
@@ -168,13 +277,8 @@ GetTimerFrequency(void)
 
 #define INSTR_TIME_IS_ZERO(t)	((t).ticks == 0)
 
-
 #define INSTR_TIME_SET_ZERO(t)	((t).ticks = 0)
 
-#define INSTR_TIME_SET_CURRENT_LAZY(t) \
-	(INSTR_TIME_IS_ZERO(t) ? INSTR_TIME_SET_CURRENT(t), true : false)
-
-
 #define INSTR_TIME_ADD(x,y) \
 	((x).ticks += (y).ticks)
 
-- 
2.47.3


--vtqqtrpooseurzip
Content-Type: text/x-diff; charset=utf-8
Content-Disposition: attachment;
	filename="v12-0003-pg_test_timing-Add-fast-flag-to-test-fast-timing.patch"



^ permalink  raw  reply  [nested|flat] 268+ messages in thread