From 1d63f7d2d180c8708bc12710254eb7b45823440f Mon Sep 17 00:00:00 2001 From: Tom Lane Date: Mon, 2 Jan 2017 13:41:51 -0500 Subject: [PATCH] Use clock_gettime(), if available, in instr_time measurements. The advantage of clock_gettime() is that the API allows the result to be precise to nanoseconds, not just microseconds as in gettimeofday(). Now that it's routinely possible to do tens of plan node executions in 1us, we really need more precision than gettimeofday() can offer for EXPLAIN ANALYZE to accumulate statistics with. Some research shows that clock_gettime() is available on pretty nearly every modern Unix-ish platform, and as far as I have been able to test, it has about the same execution time as gettimeofday(), so there's no loss in switching over. (By the same token, this doesn't do anything to fix the fact that we really wish clock readings were faster. But there's enough win here to justify changing anyway.) A small side benefit is that on most platforms, we can use CLOCK_MONOTONIC instead of CLOCK_REALTIME and thereby render EXPLAIN impervious to concurrent resets of the system clock. (This means that code must not assume that the contents of struct instr_time have any well-defined interpretation as timestamps, but really that was true before.) Some platforms offer nonstandard clock IDs that might be of interest. This patch knows we should use CLOCK_MONOTONIC_RAW on macOS, because it provides more precision and is faster to read than their CLOCK_MONOTONIC. If there turn out to be many more cases where we need special rules, it might be appropriate to handle the selection of clock ID in configure, but for the moment that doesn't seem worth the trouble. Discussion: https://postgr.es/m/31856.1400021891@sss.pgh.pa.us --- configure | 58 ++++++++++++++- configure.in | 3 +- src/include/pg_config.h.in | 3 + src/include/pg_config.h.win32 | 3 + src/include/portability/instr_time.h | 102 +++++++++++++++++++++++++-- 5 files changed, 163 insertions(+), 6 deletions(-) diff --git a/configure b/configure index 0f143a0fad..e5dd6fb789 100755 --- a/configure +++ b/configure @@ -9055,6 +9055,62 @@ if test "$ac_res" != no; then : fi +{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing clock_gettime" >&5 +$as_echo_n "checking for library containing clock_gettime... " >&6; } +if ${ac_cv_search_clock_gettime+:} false; then : + $as_echo_n "(cached) " >&6 +else + ac_func_search_save_LIBS=$LIBS +cat confdefs.h - <<_ACEOF >conftest.$ac_ext +/* end confdefs.h. */ + +/* Override any GCC internal prototype to avoid an error. + Use char because int might match the return type of a GCC + builtin and then its argument prototype would still apply. */ +#ifdef __cplusplus +extern "C" +#endif +char clock_gettime (); +int +main () +{ +return clock_gettime (); + ; + return 0; +} +_ACEOF +for ac_lib in '' rt posix4; do + if test -z "$ac_lib"; then + ac_res="none required" + else + ac_res=-l$ac_lib + LIBS="-l$ac_lib $ac_func_search_save_LIBS" + fi + if ac_fn_c_try_link "$LINENO"; then : + ac_cv_search_clock_gettime=$ac_res +fi +rm -f core conftest.err conftest.$ac_objext \ + conftest$ac_exeext + if ${ac_cv_search_clock_gettime+:} false; then : + break +fi +done +if ${ac_cv_search_clock_gettime+:} false; then : + +else + ac_cv_search_clock_gettime=no +fi +rm conftest.$ac_ext +LIBS=$ac_func_search_save_LIBS +fi +{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_search_clock_gettime" >&5 +$as_echo "$ac_cv_search_clock_gettime" >&6; } +ac_res=$ac_cv_search_clock_gettime +if test "$ac_res" != no; then : + test "$ac_res" = "none required" || LIBS="$ac_res $LIBS" + +fi + # Solaris: { $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing fdatasync" >&5 $as_echo_n "checking for library containing fdatasync... " >&6; } @@ -12520,7 +12576,7 @@ fi LIBS_including_readline="$LIBS" LIBS=`echo "$LIBS" | sed -e 's/-ledit//g' -e 's/-lreadline//g'` -for ac_func in cbrt dlopen fdatasync getifaddrs getpeerucred getrlimit mbstowcs_l memmove poll pstat pthread_is_threaded_np readlink setproctitle setsid shm_open symlink sync_file_range towlower utime utimes wcstombs wcstombs_l +for ac_func in cbrt clock_gettime dlopen fdatasync getifaddrs getpeerucred getrlimit mbstowcs_l memmove poll pstat pthread_is_threaded_np readlink setproctitle setsid shm_open symlink sync_file_range towlower utime utimes wcstombs wcstombs_l do : as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh` ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var" diff --git a/configure.in b/configure.in index b9831bc340..77aa3b41d7 100644 --- a/configure.in +++ b/configure.in @@ -1016,6 +1016,7 @@ AC_SEARCH_LIBS(getopt_long, [getopt gnugetopt]) AC_SEARCH_LIBS(crypt, crypt) AC_SEARCH_LIBS(shm_open, rt) AC_SEARCH_LIBS(shm_unlink, rt) +AC_SEARCH_LIBS(clock_gettime, [rt posix4]) # Solaris: AC_SEARCH_LIBS(fdatasync, [rt posix4]) # Required for thread_test.c on Solaris @@ -1415,7 +1416,7 @@ PGAC_FUNC_WCSTOMBS_L LIBS_including_readline="$LIBS" LIBS=`echo "$LIBS" | sed -e 's/-ledit//g' -e 's/-lreadline//g'` -AC_CHECK_FUNCS([cbrt dlopen fdatasync getifaddrs getpeerucred getrlimit mbstowcs_l memmove poll pstat pthread_is_threaded_np readlink setproctitle setsid shm_open symlink sync_file_range towlower utime utimes wcstombs wcstombs_l]) +AC_CHECK_FUNCS([cbrt clock_gettime dlopen fdatasync getifaddrs getpeerucred getrlimit mbstowcs_l memmove poll pstat pthread_is_threaded_np readlink setproctitle setsid shm_open symlink sync_file_range towlower utime utimes wcstombs wcstombs_l]) AC_REPLACE_FUNCS(fseeko) case $host_os in diff --git a/src/include/pg_config.h.in b/src/include/pg_config.h.in index 42a3fc862e..b9dfdd41c1 100644 --- a/src/include/pg_config.h.in +++ b/src/include/pg_config.h.in @@ -105,6 +105,9 @@ /* Define to 1 if you have the `class' function. */ #undef HAVE_CLASS +/* Define to 1 if you have the `clock_gettime' function. */ +#undef HAVE_CLOCK_GETTIME + /* Define to 1 if you have the header file. */ #undef HAVE_CRTDEFS_H diff --git a/src/include/pg_config.h.win32 b/src/include/pg_config.h.win32 index ceb8b7956e..199668c187 100644 --- a/src/include/pg_config.h.win32 +++ b/src/include/pg_config.h.win32 @@ -75,6 +75,9 @@ /* Define to 1 if you have the `class' function. */ /* #undef HAVE_CLASS */ +/* Define to 1 if you have the `clock_gettime' function. */ +/* #undef HAVE_CLOCK_GETTIME */ + /* Define to 1 if you have the `crypt' function. */ /* #undef HAVE_CRYPT */ diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h index 16caf6eee3..b10d6f019f 100644 --- a/src/include/portability/instr_time.h +++ b/src/include/portability/instr_time.h @@ -4,10 +4,10 @@ * portable high-precision interval timing * * This file provides an abstraction layer to hide portability issues in - * interval timing. On Unix we use gettimeofday(), but on Windows that - * gives a low-precision result so we must use QueryPerformanceCounter() - * instead. These macros also give some breathing room to use other - * high-precision-timing APIs on yet other platforms. + * interval timing. On Unix we use clock_gettime() if available, else + * gettimeofday(). On Windows, gettimeofday() gives a low-precision result + * so we must use QueryPerformanceCounter() instead. 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 @@ -54,6 +54,94 @@ #ifndef WIN32 +#ifdef HAVE_CLOCK_GETTIME + +/* Use clock_gettime() */ + +#include + +/* + * The best clockid to use according to the POSIX spec is CLOCK_MONOTONIC, + * since that will give reliable interval timing even in the face of changes + * to the system clock. However, POSIX doesn't require implementations to + * provide anything except CLOCK_REALTIME, so fall back to that if we don't + * find CLOCK_MONOTONIC. + * + * Also, some implementations have nonstandard clockids with better properties + * 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. + */ +#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW) +#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW +#elif defined(CLOCK_MONOTONIC) +#define PG_INSTR_CLOCK CLOCK_MONOTONIC +#else +#define PG_INSTR_CLOCK CLOCK_REALTIME +#endif + +typedef struct timespec instr_time; + +#define INSTR_TIME_IS_ZERO(t) ((t).tv_nsec == 0 && (t).tv_sec == 0) + +#define INSTR_TIME_SET_ZERO(t) ((t).tv_sec = 0, (t).tv_nsec = 0) + +#define INSTR_TIME_SET_CURRENT(t) ((void) clock_gettime(PG_INSTR_CLOCK, &(t))) + +#define INSTR_TIME_ADD(x,y) \ + do { \ + (x).tv_sec += (y).tv_sec; \ + (x).tv_nsec += (y).tv_nsec; \ + /* Normalize */ \ + while ((x).tv_nsec >= 1000000000) \ + { \ + (x).tv_nsec -= 1000000000; \ + (x).tv_sec++; \ + } \ + } while (0) + +#define INSTR_TIME_SUBTRACT(x,y) \ + do { \ + (x).tv_sec -= (y).tv_sec; \ + (x).tv_nsec -= (y).tv_nsec; \ + /* Normalize */ \ + while ((x).tv_nsec < 0) \ + { \ + (x).tv_nsec += 1000000000; \ + (x).tv_sec--; \ + } \ + } while (0) + +#define INSTR_TIME_ACCUM_DIFF(x,y,z) \ + do { \ + (x).tv_sec += (y).tv_sec - (z).tv_sec; \ + (x).tv_nsec += (y).tv_nsec - (z).tv_nsec; \ + /* Normalize after each add to avoid overflow/underflow of tv_nsec */ \ + while ((x).tv_nsec < 0) \ + { \ + (x).tv_nsec += 1000000000; \ + (x).tv_sec--; \ + } \ + while ((x).tv_nsec >= 1000000000) \ + { \ + (x).tv_nsec -= 1000000000; \ + (x).tv_sec++; \ + } \ + } while (0) + +#define INSTR_TIME_GET_DOUBLE(t) \ + (((double) (t).tv_sec) + ((double) (t).tv_nsec) / 1000000000.0) + +#define INSTR_TIME_GET_MILLISEC(t) \ + (((double) (t).tv_sec * 1000.0) + ((double) (t).tv_nsec) / 1000000.0) + +#define INSTR_TIME_GET_MICROSEC(t) \ + (((uint64) (t).tv_sec * (uint64) 1000000) + (uint64) ((t).tv_nsec / 1000)) + +#else /* !HAVE_CLOCK_GETTIME */ + +/* Use gettimeofday() */ + #include typedef struct timeval instr_time; @@ -113,8 +201,13 @@ typedef struct timeval instr_time; #define INSTR_TIME_GET_MICROSEC(t) \ (((uint64) (t).tv_sec * (uint64) 1000000) + (uint64) (t).tv_usec) + +#endif /* HAVE_CLOCK_GETTIME */ + #else /* WIN32 */ +/* Use QueryPerformanceCounter() */ + typedef LARGE_INTEGER instr_time; #define INSTR_TIME_IS_ZERO(t) ((t).QuadPart == 0) @@ -149,6 +242,7 @@ GetTimerFrequency(void) QueryPerformanceFrequency(&f); return (double) f.QuadPart; } + #endif /* WIN32 */ #endif /* INSTR_TIME_H */