Bug#35745: SELECT COUNT(*) is not correct for some partitioned tables.

problem was that ha_partition::records was not implemented, thus
using the default handler::records, which is not correct if the engine
does not support HA_STATS_RECORDS_IS_EXACT.
Solution was to implement ha_partition::records as a wrapper around
the underlying partitions records.

The rows column in explain partitions will now include the total
number of records in the partitioned table.

(recommit after removing out-commented code)

1 files changed, 28 insertions, 0 deletions

diff --git a/sql/ha_partition.cc b/sql/ha_partition.cc
index 3a78e6601f0..b06e2572f44 100644
--- a/sql/ha_partition.cc
+++ b/sql/ha_partition.cc
@@ -5364,6 +5364,34 @@ ha_rows ha_partition::estimate_rows_upper_bound()
 }
 
 
+/**
+  Number of rows in table. see handler.h
+
+  SYNOPSIS
+    records()
+
+  RETURN VALUE
+    Number of total rows in a partitioned table.
+*/
+
+ha_rows ha_partition::records()
+{
+  ha_rows rows, tot_rows= 0;
+  handler **file;
+  DBUG_ENTER("ha_partition::records");
+
+  file= m_file;
+  do
+  {
+    rows= (*file)->records();
+    if (rows == HA_POS_ERROR)
+      DBUG_RETURN(HA_POS_ERROR);
+    tot_rows+= rows;
+  } while (*(++file));
+  DBUG_RETURN(tot_rows);
+}
+
+
 /*
   Is it ok to switch to a new engine for this table