Hello,
Thanks for the catch and the proposed fix! Indeed, on errors the timing is
not updated appropriately.
ISTM that the best course is to update the elapsed time whenever a result
is obtained, so that a sensible value is always available.
See attached patch which is a variant of Richard's version.
fabien=# SELECT 1 as one \; SELECT 1/0 \; SELECT 2 as two;
┌─────┐
│ one │
├─────┤
│ 1 │
└─────┘
(1 row)
ERROR: division by zero
Time: 0,352 ms
Probably it would be appropriate to add a test case. I'll propose
something later.
--
Fabien.
diff --git a/src/bin/psql/common.c b/src/bin/psql/common.c
index feb1d547d4..773673cc62 100644
--- a/src/bin/psql/common.c
+++ b/src/bin/psql/common.c
@@ -1560,6 +1560,16 @@ ExecQueryAndProcessResults(const char *query, double *elapsed_msec, bool *svpt_g
else
result = PQgetResult(pset.db);
+ /*
+ * Get current timing measure in case an error occurs
+ */
+ if (timing)
+ {
+ INSTR_TIME_SET_CURRENT(after);
+ INSTR_TIME_SUBTRACT(after, before);
+ *elapsed_msec = INSTR_TIME_GET_MILLISEC(after);
+ }
+
continue;
}
else if (svpt_gone_p && !*svpt_gone_p)
@@ -1612,7 +1622,7 @@ ExecQueryAndProcessResults(const char *query, double *elapsed_msec, bool *svpt_g
last = (next_result == NULL);
/*
- * Get timing measure before printing the last result.
+ * Update current timing measure.
*
* It will include the display of previous results, if any.
* This cannot be helped because the server goes on processing
@@ -1623,7 +1633,7 @@ ExecQueryAndProcessResults(const char *query, double *elapsed_msec, bool *svpt_g
* With combined queries, timing must be understood as an upper bound
* of the time spent processing them.
*/
- if (last && timing)
+ if (timing)
{
INSTR_TIME_SET_CURRENT(after);
INSTR_TIME_SUBTRACT(after, before);
