Merge dl145s.mysql.com:/users/mhansson/mysql/push/bug23856/my50-bug23856

into  dl145s.mysql.com:/users/mhansson/mysql/push/bug23856/my51-bug23856


mysql-test/r/func_gconcat.result:
  Auto merged
mysql-test/t/func_gconcat.test:
  Auto merged
sql/field.h:
  Auto merged
sql/field_conv.cc:
  Auto merged
sql/sql_select.cc:
  Auto merged
sql/item_sum.cc:
  Bug#23856: Manual merge 5.0->5.1. Changed comparison to strict inequality between 
  convert_blob_length and UINT_MAX16. Replaced UINT_MAX16 with Field_varstring::MAX_SIZE.

1 files changed, 14 insertions, 0 deletions

diff --git a/sql/field_conv.cc b/sql/field_conv.cc
index 9771cbc12b1..e042c0ad76d 100644
--- a/sql/field_conv.cc
+++ b/sql/field_conv.cc
@@ -529,7 +529,21 @@ void Copy_field::set(char *to,Field *from)
 }
 
 
+/*
+  To do: 
+
+  If 'save\ is set to true and the 'from' is a blob field, do_copy is set to
+  do_save_blob rather than do_conv_blob.  The only differences between them
+  appears to be:
 
+  - do_save_blob allocates and uses an intermediate buffer before calling 
+    Field_blob::store. Is this in order to trigger the call to 
+    well_formed_copy_nchars, by changing the pointer copy->tmp.ptr()?
+    That call will take place anyway in all known cases.
+
+  - The above causes a truncation to MAX_FIELD_WIDTH. Is this the intended 
+    effect? Truncation is handled by well_formed_copy_nchars anyway.
+ */
 void Copy_field::set(Field *to,Field *from,bool save)
 {
   if (to->type() == MYSQL_TYPE_NULL)