This was triggered by innodb.innodb_multi_update, where we had a static length row without nulls and xtradb didn't fill in the delete-marker byte


include/my_bitmap.h:
  Added prototype for bitmap_union_is_set_all()
mysys/my_bitmap.c:
  Added function to check if union of two bit maps covers all bits.
sql/mysql_priv.h:
  Updated protype for compare_record()
sql/sql_insert.cc:
  Send to compare_record() flag if all fields are used.
sql/sql_select.cc:
  Set share->null_bytes_for_compare.
sql/sql_update.cc:
  In compare_record() don't use the fast cmp_record() (which is basically memcmp) if we don't know that all fields exists.
  Don't compare the null_bytes if there is no data there.
sql/table.cc:
  Store in share->null_bytes_for_compare the number of bytes that has null or bit fields (but not delete marker)
  Store in can_cmp_whole_record if we can use memcmp() (assuming all rows are read) to compare rows in compare_record()
sql/table.h:
  Added two elements in table->share to speed up checking how updated rows can be compared.

Speed up some PBXT tests by adding begin...commit around creating of testing tables.

include/my_base.h:
  Fixed wrong constant
mysql-test/mysql-test-run.pl:
  Print MariaDB instead of MySQL
mysql-test/r/range.result:
  Move test that required partitions to parts.optimizer
mysql-test/suite/innodb_plugin/t/disabled.def:
  Disable test that causes valgrind warning about not released memory in xtradb
mysql-test/suite/parts/r/optimizer.result:
  Moved from range.result
mysql-test/suite/parts/t/optimizer.test:
  Moved from range.test
mysql-test/suite/pbxt/r/join_nested.result:
  Updated results after optimizer changes
mysql-test/suite/pbxt/r/renamedb.result:
  Updated test for new error message
mysql-test/suite/pbxt/t/check.test:
  Speed up test
mysql-test/suite/pbxt/t/count_distinct2.test:
  Speed up test
mysql-test/suite/pbxt/t/derived.test:
  Speed up test
mysql-test/suite/pbxt/t/renamedb.test:
  Updated test for new error message
mysql-test/suite/rpl/r/rpl_log_pos.result:
  Updated results
mysql-test/suite/rpl/t/rpl_log_pos.test:
  Update test to read from a position that has 'known wrong' data.
  The orignal test read a timestamp, so the error message could differ between runs.
mysql-test/suite/rpl/t/rpl_temporary_errors.test:
  Sync to slave to make test predictable
mysql-test/t/events_time_zone.test:
  Extend wait to make test predictable
mysql-test/t/range.test:
  Move test that required partitions to parts.optimizer
sql/sql_list.h:
  Fixed compiler warning
sql/sql_load.cc:
  buffer was not freed in some error conditions
tests/mysql_client_test.c:
  Fixed compiler warning

Added 'Aria' as an alias for the Maria storage engine. Real rename is happening in 5.2

sql/handler.cc:
  Added 'Aria' as an alias for the Maria storage engine.

Fixed Bug#52005 'JOIN_TAB->dependent' may be incorrectly propageted for multilevel outer joins' in a better way (patch from Sergey Petrunya)

Fix for LP#602604: RQG: ma_blockrec.c:6187: _ma_apply_redo_insert_row_head_or_tail: Assertion `0' failed on Maria engine recovery
More DBUG_PRINT (to simplify future debugging)
Aria: Added STATE_IN_REPAIR, which is set on start of repair. This allows us to see if 'crashed' flag was set intentionally.
Aria: Some trivial speedup optimization
Aria: Better warning if table was marked crashed by unfinnished repair


mysql-test/lib/v1/mysql-test-run.pl:
  Fix so one can run RQG
mysql-test/suite/maria/r/maria-recovery2.result:
  Update for new error message.
mysys/stacktrace.c:
  Fixed compiler warning
storage/maria/ha_maria.cc:
  More DBUG_PRINT
  Added STATE_IN_REPAIR flag, which is set on start of repair. This allows us to see if 'crashed' flag was set intentionally.
  Don't log query for dropping temporary table.
storage/maria/ha_maria.h:
  Added prototype for drop_table()
storage/maria/ma_blockrec.c:
  More DBUG_PRINT
  Make read_long_data() inline for most cases. (Trivial speedup optimization)
storage/maria/ma_check.c:
  Better warning if table was marked crashed by unfinnished repair
storage/maria/ma_open.c:
  More DBUG_PRINT
storage/maria/ma_recovery.c:
  Give warning if found crashed table.
  Changed warning for tables that can't be opened.
storage/maria/ma_recovery_util.c:
  Write warnings to DBUG file
storage/maria/maria_chk.c:
  Added STATE_IN_REPAIR flag, which is set on start of repair. This allows us to see if 'crashed' flag was set intentionally.
storage/maria/maria_def.h:
  Added maria_mark_in_repair(x)
storage/maria/maria_read_log.c:
  Added option: --character-sets-dir
storage/maria/trnman.c:
  By default set min_read_from to max value.
  This allows us to remove TRN:s from rows during recovery to get more space.
  This fixes bug LP#602604: RQG: ma_blockrec.c:6187: _ma_apply_redo_insert_row_head_or_tail: Assertion `0' failed on Maria engine recovery

mysql-test/suite/federated/federated_server.result:
  Update to new error message
mysql-test/suite/federated/federated_server.test:
  Update to new error message
storage/federatedx/ha_federatedx.cc:
  Return real error code, not 0, as in some cases before.

mysql client: Ignore --comments at start of command line. This allows one to more easily run mysqltest tests trough the command line.
Fixed bug: LP#603026 RQG: pagecache_read: Assertion `pageno < ((1ULL) << 40)' on OPTIMIZE TABLE of a Maria table  

client/mysql.cc:
  Removed Oracle copyright from stdout, as Oracle doesn't have copyright to all code in this file.
  Ignore --comments at start of command line. This allows one to more easily run mysqltest tests trough the command line.
mysql-test/suite/maria/r/optimize.result:
  Added test for LP#603026
mysql-test/suite/maria/t/optimize.test:
  Added test for LP#603026
sql/net_serv.cc:
  Removed DBUG_ASSERT(), as this code can happen during testing.
storage/maria/ma_check.c:
  Fixed bug: LP#603026 RQG: pagecache_read: Assertion `pageno < ((1ULL) << 40)' on OPTIMIZE TABLE of a Maria table  
  The problem was duplicated memory usage with long packed keys.

was not complete. Bootstrap failed to pick up necessary files needed by test and hence
all tests failed.

Merge up to sunny.bains@oracle.com-20100625081841-ppulnkjk1qlazh82 .
There are 8 more changesets in mysql-5.1-innodb, but PB2 shows a
failure for a test added in one of them. If that is resolved quickly
then those 8 more changesets will be merged too.

redone locking in TC_LOG_MMAP::log_xid

POSIX requires that a signal handler defined with sigaction()
is not reset on delivering a signal unless SA_NODEFER or
SA_RESETHAND is set. It is therefore unnecessary to redefine
the handler on signal delivery on platforms where sigaction()
is used without those flags.

include/my_alarm.h:
  Renamed DONT_REMEMBER_SIGNAL ->
  SIGNAL_HANDLER_RESET_ON_DELIVERY.
include/my_global.h:
  Renamed DONT_REMEMBER_SIGNAL ->
  SIGNAL_HANDLER_RESET_ON_DELIVERY. The latter is now defined
  only on non-BSD platforms missing the POSIX sigaction()
  function.
libmysql/libmysql.c:
  Renamed DONT_REMEMBER_SIGNAL ->
  SIGNAL_HANDLER_RESET_ON_DELIVERY
mysys/thr_alarm.c:
  Renamed DONT_REMEMBER_SIGNAL ->
  SIGNAL_HANDLER_RESET_ON_DELIVERY
sql/mysqld.cc:
  Renamed DONT_REMEMBER_SIGNAL ->
  SIGNAL_HANDLER_RESET_ON_DELIVERY

The problem is that QUICK_SELECT_DESC behaviour depends
on used_key_parts value which can be bigger than selected
best_key_parts value if an engine supports clustered key.
But used_key_parts is overwritten with best_key_parts
value that prevents from correct selection of index
access method. The fix is to preserve used_key_parts
value for further use in QUICK_SELECT_DESC.


mysql-test/r/innodb_mysql.result:
  test case
mysql-test/t/innodb_mysql.test:
  test case
sql/sql_select.cc:
  preserve used_key_parts value for further use in QUICK_SELECT_DESC

automatic reconnect

A client with automatic reconnect enabled will see the error
message "Lost connection to MySQL server during query" if the
connection is lost between mysql_stmt_prepare() and
mysql_stmt_execute(). The mysql_stmt_errno() number, however,
is 0 -- not the corresponding value 2013.

This patch checks for the case where the prepared statement
has been pruned due to a connection loss (i.e., stmt->mysql
has been set to NULL) during a call to cli_advanced_command(),
and avoids changing the last_errno to the result of the last
reconnect attempt.

platform dependent diffs.