fix for bug #19690: ORDER BY eliminates rows from the result

Depending on the queries we use different data processing methods
and can lose some data in case of double (and decimal in 4.1) fields.

The fix consists of two parts:
1. double comparison changed, now double a is equal to double b 
if (a-b) is less than 5*0.1^(1 + max(a->decimals, b->decimals)). 
For example, if a->decimals==1, b->decimals==2, a==b if (a-b)<0.005
2. if we use a temporary table, store double values there as is 
to avoid any data conversion (rounding).

1 files changed, 6 insertions, 0 deletions

diff --git a/sql/init.cc b/sql/init.cc
index 4beb8db0c6f..5e1b6532c75 100644
--- a/sql/init.cc
+++ b/sql/init.cc
@@ -45,6 +45,12 @@ void unireg_init(ulong options)
   {					/* It's used by filesort... */
     log_10[i]= nr ; nr*= 10.0;
   }
+  /* Make a tab of powers of 0.1 */
+  for (i= 0, nr= 0.1; i < array_elements(log_01); i++)
+  {
+    log_01[i]= nr;
+    nr*= 0.1;
+  }
   specialflag|=options;			/* Set options from argv */
   DBUG_VOID_RETURN;
 }