Length value is the length of the field,
Max_length is the length of the field value.
So Max_length can not be more than Length.
The fix: fixed calculation of the Item_empty_string item length

(Patch applied and queued on demand of Trudy/Davi.)

When the fractional part in a multiplication of DECIMALs
overflowed, we truncated the first operand rather than the
longest. Now truncating least significant places instead
for more precise multiplications.

(Queuing at demand of Trudy/Davi.)

Due to unknown changes the test failed in some ways.

Fixed by checking the test case in detail, commenting the expected behavior,
and fixing error directives.

In the course of the analyze unneeded get_lock()/release_lock() use,
unneeded send/reap use, and unneeded sleeps were removed. The lock wait
timeout was reduced to 1 second, so that this is no big-test any more.

The test was split into two parts, one running the tests with
--innodb_locks_unsafe_for_binlog, the other part without.
The main part (include/concurrent.inc) conditionally expects
lock wait timeouts based on the value of the system variable
innodb_locks_unsafe_for_binlog.

The major part of the patch comes from Kristofer Pettersson.

(Chad queues this patch on demand by Trudy/Davi.)

	update accross partitions.
      
It's not Innodb-specific bug.
ha_partition::update_row() didn't set
table->timestamp_field_type= TIMESTAMP_NO_AUTO_SET when
orig_timestamp_type == TIMESTAMP_AUTO_SET_ON_INSERT.

So that a partition sets the timestamp field when a record
is moved to a different partition.

Fixed by doing '= TIMESTAMP_NO_AUTO_SET' unconditionally.
Also ha_partition::write_row() is fixed in same way as now
Field_timestamp::set() is called twice in SET_ON_INSERT case.

(Chad queues this patch on demand by Trudy/Davi.)

Fix for a valgrind warning due to a jump on a uninitialized
variable. The problem was that the sql profile preparation
function wasn't being called for all possible code paths
of query execution.

The solution is to ensure that query profiling is always
started before dispatch_command function is called and to
explicitly call the profile preparation function on bootstrap.

The problem:
CSV storage engine open function returns success even
thought it failed to open the data file
The fix:
return error
Additional fixes:
added MY_WME to my_open to avoid mysterious error message
free share struct if open the file was unsuccessful

with blobs containing nulls

Problem: FEDERATED SE improperly stores NULL fields in the record buffer.

Fix: store them properly.

Problem: missed "break" in a switch leads to unexpected assertion failure
of 'myisamchk compressed_table'.

Fix: add the break.

Problem: REGEXP in functions/PSs may return wrong results
due to improper initialization.

Fix: initialize required REGEXP params.

The problem was because the event allocated in mysql_client_binlog_statement
was not freed when an error occured while applying the event.

    Test "funcs_1.processlist_val_ps" fails in various ways
+ corrections of logic in poll routines
+ minor improvements

Details:
- add subtest with drop unrelated view
- rearrange existing tests so that a distinction
  between drop procedure and drop function effects
  is possible

partition is corrupt

Post push fix

an DBUG_ASSERT broke the embedded server, fixed by initializing
it in the embedded version of Protocol_text::prepare_for_resend

used causes server crash.

When the loose index scan access method is used values of aggregated functions
are precomputed by it. Aggregation of such functions shouldn't be performed
in this case and functions should be treated as normal ones.
The create_tmp_table function wasn't taking this into account and this led to
a crash if a query has MIN/MAX aggregate functions and employs temporary table
and loose index scan.
Now the JOIN::exec and the create_tmp_table functions treat MIN/MAX aggregate
functions as normal ones when the loose index scan is used.

Post push fix (compiler warning)

This fix is for 5.1 only : back porting the 6.0 patch manually

The parser code in sql/sql_yacc.yy needs to be more robust to out of
memory conditions, so that when parsing a query fails due to OOM,
the thread gracefully returns an error.

Before this fix, a new/alloc returning NULL could:
- cause a crash, if dereferencing the NULL pointer,
- produce a corrupted parsed tree, containing NULL nodes,
- alter the semantic of a query, by silently dropping token values or nodes

With this fix:
- C++ constructors are *not* executed with a NULL "this" pointer
when operator new fails.
This is achieved by declaring "operator new" with a "throw ()" clause,
so that a failed new gracefully returns NULL on OOM conditions.

- calls to new/alloc are tested for a NULL result,

- The thread diagnostic area is set to an error status when OOM occurs.
This ensures that a request failing in the server properly returns an
ER_OUT_OF_RESOURCES error to the client.

- OOM conditions cause the parser to stop immediately (MYSQL_YYABORT).
This prevents causing further crashes when using a partially built parsed
tree in further rules in the parser.

No test scripts are provided, since automating OOM failures is not
instrumented in the server.
Tested under the debugger, to verify that an error in alloc_root cause the
thread to returns gracefully all the way to the client application, with
an ER_OUT_OF_RESOURCES error.