row_sel_field_store_in_mysql_format(): Do not pad the unused part of
the buffer reserved for a True VARCHAR column (introduced in 5.0.3).
Add Valgrind instrumentation ensuring that the unused part will be
flagged uninitialized.

row_sel_copy_cached_field_for_mysql(): New function: Copy a field
that is in the MySQL row format, not copying the unused tail of
VARCHAR columns.

row_sel_pop_cached_row_for_mysql(): Invoke
row_sel_copy_cached_field_for_mysql() for copying fields.
When the row is long, copy it field-by-field.

rb:715 approved by Inaam Rana

for compressed InnoDB tables

ha_innodb::info_low(): For calculating data_length or index_length,
use the compressed page size for compressed tables instead of UNIV_PAGE_SIZE.

rb:714 approved by Sunny Bains

BUG #11754979 - 46675: ON DUPLICATE KEY UPDATE AND UPDATECOUNT() POSSIBLY WRONG 

The mysql_affected_rows() client call returns 3 instead of 2 on
INSERT ... ON DUPLICATE KEY UPDATE query with a duplicated key value.

The fix for the old bug #29692 was incomplete: unnecessary double
increment of "touched" rows still happened.

This bugfix removes:
1) unneeded increment of "touched" rows and
2) useless double resetting of auto-increment value.


sql/sql_insert.cc:
  write_record() function:
  Unneeded increment of "touched" rows and useless double resetting
  of auto-increment value has been removed.
tests/mysql_client_test.c:
  New test case.

There is an optimization of DISTINCT in JOIN::optimize()
which depends on THD::used_tables value. Each SELECT statement
inside SP resets used_tables value(see mysql_select()) and it
leads to wrong result. The fix is to replace THD::used_tables
with LEX::used_tables.


mysql-test/r/sp.result:
  test case
mysql-test/t/sp.test:
  test case
sql/sql_base.cc:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_class.cc:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_class.h:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_insert.cc:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_lex.cc:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_lex.h:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_prepare.cc:
  THD::used_tables is replaced with LEX::used_tables
sql/sql_select.cc:
  THD::used_tables is replaced with LEX::used_tables

The problem is that TIME_FUZZY_DATE is explicitly used for get_arg0_date()
function in Item_date_typecast::get_date method. The fix is to use real
fuzzy_date value.


mysql-test/r/func_time.result:
  test case
mysql-test/t/func_time.test:
  test case
sql/item_timefunc.cc:
  use real fuzzy_date value

In 5.5, REFRESH SLAVE is used as an alias for RESET SLAVE and
was removed in 5.6. Reseting a slave through REFRESH SLAVE was
causing errors in the valgrind platform since reset_slave_info
was undefined.

To fix the problem, we have set reset_slave_info while calling
REFRESH SLAVE.

SHOW ALL PROBLEMS FOR MERGE TABLE COMPLIANCE IN 5.1".

The problem was that CHECK/REPAIR TABLE for a MERGE table which
had several children missing or in wrong engine reported only
issue with the first such table in its result-set. While in 5.0
this statement returned the whole list of problematic tables.

Ability to report problems for all children was lost during
significant refactorings of MERGE code which were done as part
of work on 5.1 and 5.5 releases.

This patch restores status quo ante refactorings by changing
code in such a way that:
1) Failure to open child table due to its absence during CHECK/
   REPAIR TABLE for a MERGE table is not reported immediately
   when its absence is discovered in open_tables(). Instead
   handling/error reporting in such a situation is postponed
   until the moment when children are attached.
2) Code performing attaching of children no longer stops when
   it encounters first problem with one of the children during
   CHECK/REPAIR TABLE. Instead it continues iteration through
   the child list until all problems caused by child absence/
   wrong engine are reported.

Note that even after this change problem with mismatch of
child/parent definition won't be reported if there is also
another child missing, but this is how it was in 5.0 as well.

mysql-test/r/merge.result:
  Added test case for bug #11754210 - "45777: CHECK TABLE DOESN'T
  SHOW ALL PROBLEMS FOR MERGE TABLE COMPLIANCE IN 5.1".
  Adjusted results of existing tests to the fact that CHECK/REPAIR
  TABLE statements now try to report problems about missing table/
  wrong engine for all underlying tables, and to the fact that
  mismatch of parent/child definitions is always reported as an
  error and not a warning.
mysql-test/t/merge.test:
  Added test case for bug #11754210 - "45777: CHECK TABLE DOESN'T
  SHOW ALL PROBLEMS FOR MERGE TABLE COMPLIANCE IN 5.1".
sql/sql_base.cc:
  Changed code responsible for opening tables to ignore the fact
  that underlying tables of a MERGE table are missing, if this
  table is opened for CHECK/REPAIR TABLE.
  The absence of underlying tables in this case is now detected and
  appropriate error is reported at the point when child tables are
  attached. At this point we can produce full list of problematic
  child tables/errors to be returned as part of CHECK/REPAIR TABLE
  result-set.
storage/myisammrg/ha_myisammrg.cc:
  Changed myisammrg_attach_children_callback() to handle new
  situation, when during CHECK/REPAIR TABLE we do not report 
  error about missing child immediately when this fact is 
  discovered during open_tables() but postpone error-reporting
  till the time when children are attached. 
  Also this callback is now responsible for pushing an error
  mentioning problematic child table to the list of errors to 
  be reported by CHECK/REPAIR TABLE statements.
  Finally, since now myrg_attach_children() no longer relies on
  return value from callback to determine the end of the children
  list, callback no longer needs to set my_errno value and can
  be simplified.
  
  Changed myrg_print_wrong_table() to always report a problem
  with child table as an error and not as a warning. This makes
  reporting for different types of issues with child tables
  more consistent and compatible with 5.0 behavior.
storage/myisammrg/myrg_open.c:
  Changed code in myrg_attach_children() not to abort on the
  first problem with a child table when attaching children to
  parent MERGE table during CHECK/REPAIR TABLE statement 
  execution. This allows CHECK/REPAIR TABLE to report problems 
  about absence/wrong engine for all underlying tables as
  part of their result-set.

TOOLS

Backport a fix for Bug 57094 from 5.5.
The following revision was backported:

# revision-id: alexander.nozdrin@oracle.com-20101006150613-ls60rb2tq5dpyb5c
# parent: bar@mysql.com-20101006121559-am1e05ykeicwnx48
# committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
# branch nick: mysql-5.5-bugteam-bug57094
# timestamp: Wed 2010-10-06 19:06:13 +0400
# message:
#   Fix for Bug 57094 (Copyright notice incorrect?).
#   
#   The fix is to:
#     - introduce ORACLE_WELCOME_COPYRIGHT_NOTICE define to have a single place
#       to specify copyright notice;
#     - replace custom copyright notices with ORACLE_WELCOME_COPYRIGHT_NOTICE
#       in programs.

             FUNCTION 'PTHREAD_INIT' 

The problem was that compilation would fail with a warning:
Implicit declaration of function 'pthread_init' if MySQL was
compiled on OS X 10.7 (Lion). The reason was that pthread_init()
is now part of an internal OS X pthread library so it was found
by CMake.

This patch fixes the problem by removing HAVE_PTHREAD_INIT and
related code. pthread_init() was specific to MIT-pthreads which
has not been supported since 4.1 and was therefore no longer
relevant.

No test case added.

mysql-test/t/implicit_commit.test:
  Test fails if server is compiled with -DENABLED_PROFILING=0
sql/sql_class.cc:
  Let class PROFILING do its own handling of the input file name.
sql/sql_profile.cc:
  Store only basename of file argument.

The title of the bug is a little confusing. The actual fix is to
reintroduce random readahead inside InnoDB with a dynamic, global
switch innodb_random_read_ahead [default = off].

Approved by: Sunny Bains
rb://696

Before BUG#28796, an empty host was used to identify that an instance was no
longer a slave. However, BUG#28796 changed this behavior and one cannot set
an empty host. Besides, a RESET SLAVE only cleans up information on the next
event to retrieve from the master, disables ssl and resets heartbeat period.
So a call to SHOW SLAVE STATUS after issuing a RESET SLAVE still returns some
valid information, such as host, port, user and password.

To fix this problem, we have introduced the command RESET SLAVE ALL that does
what a regular RESET SLAVE does and also clears host, port, user and password
information thus allowing users to identify when an instance is no longer a
slave.

Truncate result of decimal division before converting to integer.


mysql-test/r/func_math.result:
  New test case.
mysql-test/t/func_math.test:
  New test case.
sql/item_func.cc:
  Item_func_int_div::val_int():
  Truncate result of decimal division before converting to integer.

Turns out the DBUG_ASSERT added by fix for Bug#11792200 was overly pessimistic:
'stop0' is used in the main loop of do_div_mod, but we only dereference 'buf0'
for div operations, not for mod.


mysql-test/r/func_math.result:
  New test case.
mysql-test/t/func_math.test:
  New test case.
strings/decimal.c:
  Move DBUG_ASSERT down to where we actually dereference the loop pointer.

Original changeset:
revision-id: alexander.nozdrin@oracle.com-20101221122349-6h8ammcro70a4pac
parent: sven.sandberg@oracle.com-20101221121948-hnivuulyohzch1v4
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: mysql-trunk-bugfixing
timestamp: Tue 2010-12-21 15:23:49 +0300
message:
  A patch for Bug#59060 (Valgrind warning in Protocol_text::store()).
  
  We should not assume to have zero-terminated strings.

HA_ERR was returning 0 (null string) when no error happened 
(error=0). Since HA_ERR is used in DBUG_PRINT, regardless there 
was an error or not, the server could crash in solaris debug
builds.

We fix this by:

  - deploying an assertion that ensures that the function 
    is not called when no error has happened;
  - making sure that HA_ERR is only called when an error 
    happened;
  - making HA_ERR return "No Error", instead of 0, for 
    non-debug builds if it is called when no error happened.

This will make HA_ERR return values to work with DBUG_PRINT on
solaris debug builds.