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

3 files changed, 104 insertions, 4 deletions

diff --git a/src/include/pg_config.h.in b/src/include/pg_config.h.in
index 42a3fc862e9..b9dfdd41c1a 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 <crtdefs.h> header file. */
 #undef HAVE_CRTDEFS_H
 
diff --git a/src/include/pg_config.h.win32 b/src/include/pg_config.h.win32
index ceb8b7956ee..199668c1876 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 16caf6eee3d..b10d6f019f9 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 <time.h>
+
+/*
+ * 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 <sys/time.h>
 
 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 */