From d84b2ad09898e81a6107ac3d64cd13d8526b1f18 Mon Sep 17 00:00:00 2001 From: Lukas Fittl Date: Fri, 25 Jul 2025 17:57:20 -0700 Subject: [PATCH v4 2/3] Use time stamp counter to measure time on Linux/x86 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 Author: Andres Freund Author: Lukas Fittl Reviewed-by: Discussion: https://www.postgresql.org/message-id/flat/20200612232810.f46nbqkdhbutzqdg%40alap3.anarazel.de --- src/backend/executor/instrument.c | 12 +- src/backend/utils/misc/guc_parameters.dat | 10 + src/backend/utils/misc/guc_tables.c | 10 + src/common/Makefile | 1 + src/common/instr_time.c | 283 ++++++++++++++++++++++ src/common/meson.build | 1 + src/include/portability/instr_time.h | 156 ++++++++++-- src/include/utils/guc_hooks.h | 2 + src/include/utils/guc_tables.h | 1 + 9 files changed, 455 insertions(+), 21 deletions(-) create mode 100644 src/common/instr_time.c diff --git a/src/backend/executor/instrument.c b/src/backend/executor/instrument.c index edab92a0ebe..ebdad31ca3b 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/misc/guc_parameters.dat b/src/backend/utils/misc/guc_parameters.dat index f0260e6e412..ca6c0d2d148 100644 --- a/src/backend/utils/misc/guc_parameters.dat +++ b/src/backend/utils/misc/guc_parameters.dat @@ -1035,6 +1035,16 @@ max => '3', }, +{ name => 'fast_clock_source', type => 'enum', context => 'PGC_USERSET', group => 'RESOURCES_TIME', + short_desc => 'Controls the use of fast clock source on architectures that support it.', + long_desc => 'This enables the use of the faster RDTSC instruction on x86 systems (if available) to support timing measurements during EXPLAIN and other instrumentation.', + variable => 'fast_clock_source', + boot_val => 'FAST_CLOCK_SOURCE_AUTO', + options => 'fast_clock_source_options', + check_hook => 'check_fast_clock_source', + assign_hook => 'assign_fast_clock_source', +}, + { name => 'file_copy_method', type => 'enum', context => 'PGC_USERSET', group => 'RESOURCES_DISK', short_desc => 'Selects the file copy method.', variable => 'file_copy_method', diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c index 13c569d8790..f92bbef943d 100644 --- a/src/backend/utils/misc/guc_tables.c +++ b/src/backend/utils/misc/guc_tables.c @@ -90,6 +90,7 @@ #include "tcop/tcopprot.h" #include "tsearch/ts_cache.h" #include "utils/builtins.h" +#include "portability/instr_time.h" #include "utils/bytea.h" #include "utils/float.h" #include "utils/guc_hooks.h" @@ -371,6 +372,15 @@ static const struct config_enum_entry huge_pages_options[] = { {NULL, 0, false} }; +static const struct config_enum_entry fast_clock_source_options[] = { + {"auto", FAST_CLOCK_SOURCE_AUTO, false}, + {"off", FAST_CLOCK_SOURCE_OFF, false}, +#if !defined(WIN32) && defined(__x86_64__) + {"rdtsc", FAST_CLOCK_SOURCE_RDTSC, false}, +#endif + {NULL, 0, false} +}; + static const struct config_enum_entry huge_pages_status_options[] = { {"off", HUGE_PAGES_OFF, false}, {"on", HUGE_PAGES_ON, false}, 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..4cd45bdca97 --- /dev/null +++ b/src/common/instr_time.c @@ -0,0 +1,283 @@ +/*------------------------------------------------------------------------- + * + * 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 +#endif + +#if defined(HAVE__CPUID) || (defined(HAVE__CPUIDEX) && defined(_MSC_VER)) +#include +#endif + +#include "portability/instr_time.h" + +#if !defined(WIN32) && defined(__x86_64__) +static void set_ticks_per_ns(void); +static void pg_initialize_rdtsc(void); +#endif + +#ifdef FRONTEND + +void +pg_initialize_fast_clock_source() +{ +#if !defined(WIN32) && defined(__x86_64__) + pg_initialize_rdtsc(); + use_tsc = has_rdtsc && has_rdtscp; + set_ticks_per_ns(); +#endif +} + +#else + +#include "utils/guc_hooks.h" + +int fast_clock_source = FAST_CLOCK_SOURCE_AUTO; + +bool +check_fast_clock_source(int *newval, void **extra, GucSource source) +{ +#if !defined(WIN32) && defined(__x86_64__) + if (*newval == FAST_CLOCK_SOURCE_AUTO || *newval == FAST_CLOCK_SOURCE_RDTSC) + pg_initialize_rdtsc(); + + if (*newval == FAST_CLOCK_SOURCE_RDTSC && (!has_rdtsc || !has_rdtscp)) + { + GUC_check_errdetail("TSC is not supported as fast clock source"); + return false; + } +#endif + + return true; +} + +void +assign_fast_clock_source(int newval, void *extra) +{ +#if !defined(WIN32) && defined(__x86_64__) + use_tsc = newval == FAST_CLOCK_SOURCE_RDTSC || (newval == FAST_CLOCK_SOURCE_AUTO && pg_fast_clock_source_default()); + + set_ticks_per_ns(); +#endif +} + +#endif /* FRONTEND */ + +/* + * Decides whether to use TSC time source if the user did not specify it + * one way or the other, and it is available (checked separately). + * + * Currently only enabled by default on Linux, since Linux already does a + * significant amount of work to determine whether TSC is a viable clock + * source. + */ +bool +pg_fast_clock_source_default() +{ +#if defined(__x86_64__) && defined(__linux__) + FILE *fp = fopen("/sys/devices/system/clocksource/clocksource0/current_clocksource", "r"); + char buf[128]; + + if (!fp) + return false; + + if (fgets(buf, sizeof(buf), fp) != NULL && strcmp(buf, "tsc\n") == 0) + { + fclose(fp); + return true; + } + + fclose(fp); +#endif + + return false; +} + +#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 max_ticks_no_overflow = PG_INT64_MAX / TICKS_TO_NS_PRECISION; + +#if defined(__x86_64__) + +bool has_rdtsc = false; +bool has_rdtscp = false; +bool use_tsc = false; + +static uint32 tsc_freq; + +static int64 +set_ticks_per_ns_for_system() +{ + return TICKS_TO_NS_PRECISION; +} + +/* + * For TSC, the ticks to nanoseconds conversion requires floating point math + * 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. + */ +static int64 +set_ticks_per_ns_for_tsc() +{ + return INT64CONST(1000000) * TICKS_TO_NS_PRECISION / tsc_freq; +} + +static void +set_ticks_per_ns() +{ + if (use_tsc) + ticks_per_ns_scaled = ticks_per_ns_for_tsc(); + else + ticks_per_ns_scaled = ticks_per_ns_for_system(); + max_ticks_no_overflow = PG_INT64_MAX / ticks_per_ns_scaled; +} + +#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. + */ +static void +pg_initialize_rdtsc(void) +{ + /* + * Compute baseline CPU peformance, determines speed at which RDTSC + * advances. + */ + if (!get_tsc_frequency_khz(&tsc_freq)) + return; + + has_rdtsc = true; + has_rdtscp = is_rdtscp_available(); +} +#endif /* defined(__x86_64__) */ + +#endif /* WIN32 */ diff --git a/src/common/meson.build b/src/common/meson.build index b757618a9c9..042edb7473a 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 490593d1825..635af2477ec 100644 --- a/src/include/portability/instr_time.h +++ b/src/include/portability/instr_time.h @@ -4,9 +4,10 @@ * 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 x86 we use the RDTSC/RDTSCP instruction directly in + * certain cases, or alternatively clock_gettime() on Unix-like systems and + * QueryPerformanceCounter() on Windows. 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 +18,11 @@ * * 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 * @@ -78,11 +80,49 @@ typedef struct instr_time #define NS_PER_MS INT64CONST(1000000) #define NS_PER_US INT64CONST(1000) +/* + * Initialize fast clock source, and enable use if available. + * + * This needs to be called by client programs before using INSTR_* macros, in + * order to benefit from fast clock sources when available. If not called + * INSTR_* macros will utilize the regular source instead. + * + * Not used for backends, where instead the fast_clock_source GUC controls, and + * the GUC check/assign hooks do the necessary initialization. + */ +#ifdef FRONTEND + +extern void + pg_initialize_fast_clock_source(void); + +#else + +typedef enum +{ + FAST_CLOCK_SOURCE_AUTO, + FAST_CLOCK_SOURCE_OFF, + FAST_CLOCK_SOURCE_RDTSC +} FastClockSourceType; + +extern int fast_clock_source; + +#endif /* FRONTEND */ + +extern bool + pg_fast_clock_source_default(void); #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) -/* Use clock_gettime() */ +extern int64 ticks_per_ns_scaled; +extern int64 max_ticks_no_overflow; + +/* Use clock_gettime() or fast clock source */ #include @@ -97,6 +137,9 @@ typedef struct instr_time * than CLOCK_MONOTONIC. In particular, as of macOS 10.12, Apple provides * CLOCK_MONOTONIC_RAW which is both faster to read and higher resolution than * their version of CLOCK_MONOTONIC. + * + * Note this does not get used in case the fast clock source logic is used, + * which directly calls architecture specific timing instructions (e.g. RDTSC). */ #if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW) #define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW @@ -106,9 +149,18 @@ typedef struct instr_time #define PG_INSTR_CLOCK CLOCK_REALTIME #endif -/* helper for INSTR_TIME_SET_CURRENT */ +#if defined(__x86_64__) +#include + +/* Indicates if TSC instructions (RDTSC and RDTSCP) are available. */ +extern bool has_tsc; + +/* Whether to actually use RDTSC/RDTSCP based on availability and GUC settings. */ +extern bool use_tsc; +#endif + static inline instr_time -pg_clock_gettime_ns(void) +pg_clock_gettime(void) { instr_time now; struct timespec tmp; @@ -119,11 +171,83 @@ pg_clock_gettime_ns(void) return now; } +static inline instr_time +pg_get_ticks_fast(void) +{ +#if defined(__x86_64__) + if (likely(use_tsc)) + { + 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__) + if (likely(use_tsc)) + { + 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; +} + +#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 +255,7 @@ 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 +277,9 @@ GetTimerFrequency(void) return (double) f.QuadPart; } +#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 +295,8 @@ 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) diff --git a/src/include/utils/guc_hooks.h b/src/include/utils/guc_hooks.h index b6ecb0e769f..4488d8579fc 100644 --- a/src/include/utils/guc_hooks.h +++ b/src/include/utils/guc_hooks.h @@ -62,6 +62,8 @@ extern void assign_default_text_search_config(const char *newval, void *extra); extern bool check_default_with_oids(bool *newval, void **extra, GucSource source); extern const char *show_effective_wal_level(void); +extern void assign_fast_clock_source(int newval, void *extra); +extern bool check_fast_clock_source(int *newval, void **extra, GucSource source); extern bool check_huge_page_size(int *newval, void **extra, GucSource source); extern void assign_io_method(int newval, void *extra); extern bool check_io_max_concurrency(int *newval, void **extra, GucSource source); diff --git a/src/include/utils/guc_tables.h b/src/include/utils/guc_tables.h index 71a80161961..63440b8e36c 100644 --- a/src/include/utils/guc_tables.h +++ b/src/include/utils/guc_tables.h @@ -60,6 +60,7 @@ enum config_group CONN_AUTH_TCP, CONN_AUTH_AUTH, CONN_AUTH_SSL, + RESOURCES_TIME, RESOURCES_MEM, RESOURCES_DISK, RESOURCES_KERNEL, -- 2.47.1