Assert that keys don't point to bitmap pages.

storage/maria/ma_blockrec.h:
  assertion
storage/maria/ma_locking.c:
  With concurrent INSERTs, it is possible that two threads enter _ma_mark_file_changed()
  at the same time, so they should serialize their access to the "changed" state/share
  members; another reason is that this function may call _ma_update_state_lsns_sub()
  which may call translog_deassign_id_from_share() (I saw it during testing of
  online backup) which requires the intern_lock mutex.
  As INSERTs only change from "not changed" to "changed", we can first check without mutex:
  if it says "changed", some other thread has set or is setting the variables now,
  we don't need to do it; if it says "not changed", we serialize and re-check.

storage/maria/ha_maria.cc:
  comment
storage/maria/ma_checkpoint.c:
  comment
storage/maria/ma_extra.c:
  use bitmap mutex when changing bitmap.changed, sounds safer
storage/maria/ma_pagecache.c:
  comment
storage/maria/ma_recovery.c:
  comments

don't treat deadlock or timeout as a sign of corruption

storage/maria/ma_close.c:
  Reset state_history when closing file.
  This fixes a race condition when the checkpoint is using the file while the file is beeing closed.
  The bug was that checkpoint was accessing the freed state_history.

(this is to ease ./mtr --mysqld=--default-storage-engine=maria)

about problems when the R-tree index is not the first index.

added a missing DBUG_VOID_RETURN

mysql-test/suite/maria/r/maria-recovery3.result:
  result update
mysql-test/suite/maria/t/maria-recovery3.test:
  Test for BUG#42112; before the bugfix, recovery would assert like this:
  ma_blockrec.c:6051: _ma_apply_redo_insert_row_head_or_tail: Assertion `rownr == 0 && new_page' failed.
storage/maria/ma_create.c:
  Fix for BUG#42112; plus some intentional crashes to test the fix. The bug was that if crash happened during
  TRUNCATE TABLE, in maria_create(), after the index file's state has been written but before its LSNs
  have been updated (so, if crash happened between _ma_state_info_write_sub() and _ma_update_state__lsns_sub()),
  then that would leave a table with create_rename_lsn==0. Recovery would then try old pre-TRUNCATE REDOs
  on this table, and fail as this table is already partly shortened. Fix is to write create_rename_lsn==LSN_MAX
  as soon as TRUNCATE touches the index file, so that Recovery ignores this table. This allows Maria to start;
  the table is still corrupted but the user can successfully repeat TRUNCATE TABLE (which required Maria to start).
storage/maria/ma_delete_all.c:
  A comment.

trnman_get_state/trnman_set_state renamed to trnman_get/set_flags
TRN::state - to TRN::flags accordingly

sql/sql_insert.cc:
  Removed DBUG_ASSERT() that is triggered by deadlock-innodb test
storage/maria/ma_loghandler.c:
  Removed compiler warnings
storage/maria/trnman_public.h:
  Fixed wrong code from last push

Added DBUG_ASSERT() to unlikely error senario
Don't use errno == 0 in maria_create() / myisam_create()

sql/sql_insert.cc:
  Added DBUG_ASSERT() for case that should never happen in real life
  Added my_error() to avoid assert if mysql_lock() or postlock() doesn't call my_error()
storage/maria/ha_maria.cc:
  Log queries to maria_log if compiled with EXTRA_DEBUG
storage/maria/ma_create.c:
  Don't use errno == 0
storage/maria/ma_loghandler.c:
  Added logging of debug info to maria_log
storage/maria/ma_loghandler.h:
  Added logging of debug info to maria_log
storage/maria/ma_recovery.c:
  Added printing of debug info from maria_log
storage/maria/trnman.c:
  Added functions to read/store state in TRN
storage/maria/trnman.h:
  Added functions to read/store state in TRN
storage/maria/trnman_public.h:
  Added state in TRN to remmeber if we have already logged the query
storage/myisam/mi_create.c:
  Don't use errno == 0

include/atomic/generic-msvc.h:
  prevent possible compiler warnings
include/lf.h:
  comments, better definition for LF_HASH_OVERHEAD
include/maria.h:
  define MARIA_CANNOT_ROLLBACK here
include/my_pthread.h:
  avoid possible name clash
include/waiting_threads.h:
  comments, const, move WT_RESOURCE to waiting_threads.c
mysql-test/suite/maria/r/maria_notembedded.result:
  new test
mysql-test/suite/maria/t/maria_notembedded.test:
  new test - 5-way deadlock
mysys/lf_hash.c:
  better definition for LF_HASH_OVERHEAD
mysys/my_static.c:
  comment
mysys/my_thr_init.c:
  casts
mysys/waiting_threads.c:
  comments, asserts, etc
server-tools/instance-manager/parse.cc:
  fix my_init_dynamic_array() to follow new calling conventions
sql/mysqld.cc:
  call wt_init after set_proper_floating_point_mode
sql/sql_class.h:
  comment
storage/maria/ha_maria.cc:
  move MARIA_CANNOT_ROLLBACK to a common header
storage/maria/ma_commit.c:
  comment
storage/maria/ma_write.c:
  comments, check for HA_ERR_FOUND_DUPP_KEY
storage/maria/trnman.c:
  comments, assert
storage/maria/trnman.h:
  comments
storage/maria/unittest/trnman-t.c:
  be paranoid
unittest/mysys/lf-t.c:
  comments
unittest/mysys/waiting_threads-t.c:
  comments, safety, memory leak

Fix for BUG#41493 "Maria: two recovery failures (wrong logging of BLOB pages)" and some more debugging
output related to this.

mysql-test/suite/maria/r/maria-recovery3.result:
  result update
mysql-test/suite/maria/t/maria-recovery3.test:
  Test for bug; before the fix, the "CHECK TABLE EXTENDED" would mention a bad bitmap, because the
  REDO_INSERT_ROW_BLOBS was containing a page number which was actually the one of a tail, so execution of this
  record would mark the tail page as full in bitmap (like if it were a blob page), though it wasn't full.
  Also, the assertion added around ma_blockrec.c:6580 in the present revision fired.
storage/maria/ma_blockrec.c:
  - fix for BUG#41493: if we found out that logging was not needed at this point (blob_length==0 i.e. tail page),
  then we forgot to increment tmp_block, so in the second iteration (assuming two BLOB columns), we would log the
  page range of the first iteration (i.e. the tail page's number) for this second BLOB, which would cause
  Recovery to overwrite the tail page with the second BLOB.
  - assert when marking the table corrupted during REDO phase; this catches some problems earlier
  otherwise they get caught only when a later record wants to use the table.
  - _ma_apply_redo_insert_row_blobs() now fills some synthetic info about the blobs and pages involved
  in a REDO_INSERT_ROW_BLOBS record, for inclusion into maria_recovery.trace: number of blobs, of ranges,
  first and last page (does not tell about any gaps in the middle, but good enough for now). It also asserts
  that it's not overwriting a tail/head page (which happened in the bug).
storage/maria/ma_blockrec.h:
  new prototype for _ma_apply_redo_insert_row_blobs
storage/maria/ma_recovery.c:
  Print info got from _ma_apply_redo_insert_row_blobs() to maria_recovery.trace (so far this file had mentioned
  what head and tail pages a record touched, but not blob pages).

of Maria tests with --big (back-ports from 6.0-maria)

mysql-test/suite/maria/r/maria-big.result:
  result update
mysql-test/suite/maria/r/maria-recovery-big.result:
  result update
mysql-test/suite/maria/t/maria-big.test:
  this test didn't work anymore due to changes in 5.1 around max_allowed_packet allowed settings
  (fix back-ported from 6.0-maria)
mysql-test/suite/maria/t/maria-recovery-big-master.opt:
  this test didn't work anymore due to changes in 5.1 around max_allowed_packet allowed settings
  (fix back-ported from 6.0-maria)
mysql-test/suite/maria/t/maria-recovery-big.test:
  this test didn't work anymore due to changes in 5.1 around max_allowed_packet allowed settings
  (fix back-ported from 6.0-maria)
mysql-test/suite/maria/t/maria-recovery2-master.opt:
  Fix for BUG#42012 "Maria: test maria-recovery2 fails with --embedded": file is updated like
  maria-recovery-master.opt had already been in guilhem@mysql.com-20080701204709-pa4megwfvllq3s7g

(in debug builds pagecache does many small mallocs, and safemalloc sanity checks take ages)

Bug #41962 Maria: view-related test failures (insert, view, maria, trigger tests)
Added error handling for wrong update of view.
See Bug #41760 Inserting into multiple-table views is not working

mysql-test/r/delayed.result:
  Fixed test as we are now testing values before fields.
  Added new tests to test all error combinations
mysql-test/suite/maria/r/maria.result:
  Added error handling for not supported update of view.
mysql-test/suite/maria/t/maria.test:
  Added error handling for not supported update of view.
mysql-test/t/delayed.test:
  Fixed test as we are now testing values before fields.
  Added new tests to test all error combinations
sql/sql_base.cc:
  Fixed warning from valgrind
sql/sql_insert.cc:
  Don't test from which table values are in case of INSERT ... SELECT
  Run fix_fields() in values before we do it on fields.
  This is needed becasue check_view_single_update() are accessing values.
storage/maria/ma_blockrec.c:
  Don't call pagecache_delete_pages() if no pages to delete.
  This fixes a DBUG_ASSERT() error in maria_test_recovery

storage/maria/maria_def.h:
  Constants for _ma_state_info_write added.

detect a case when a blocker has removed itself and signalled after the condition timed out
but before it (cond_wait) acquired the mutex back

storage/maria/ma_key_recover.c:
  Takes intern_lock before writing CLR_END of bulk insert.
  Added debug info about writing trigger.
  Check of mutex taken in _ma_state_info_write() where we need it.

Added code to detect and give error when doing an insert into a view where we accessed fields in a not yet read table
Disabled test in subselect.test as the CHECK_OPTION for views doesn't work for insert.
This needs to be fixed properly later.

The problem with views are described in Bug #41760 Inserting into multiple-table views is not working

mysql-test/r/insert.result:
  Fixed wrong usage of insert into view.
mysql-test/r/subselect.result:
  Disabled wrong test (temporary)
mysql-test/suite/maria/r/maria.result:
  Added test of size of table
mysql-test/suite/maria/t/maria.test:
  Added test of size of table
mysql-test/t/insert.test:
  Fixed wrong usage of insert into view
  The bug is that during insert/update we currently don't read any of the referenced tables of the view.
  This means that we can't get a value from another table to use as part of the update.
mysql-test/t/subselect.test:
  Disabled not working test until someone has time to fix insert into view properly
  Here we where refering to last used value in t2, which is wrong.
sql/sql_insert.cc:
  Detect if we are trying to update one table in a view based on value in another, not yet read, table.
  This fixes the problem discovered in insert.test
storage/maria/ma_blockrec.c:
  Don't ignore not critical changes to the last page in the table.
  We need to write the last page as otherwise we can during aborting of a row with a duplicate key get
  state.data_file_length and the real length of file out of sync
storage/maria/ma_check.c:
  Flush the page cache even if we got an error during zerofill.
  (This fixes a call to assert() in case of a too short data file)
storage/maria/ma_pagecache.c:
  Mark page as read when we do a write of a full page.
  This fixes a bug when we got an error during read and then used direct write to page to update it
storage/maria/ma_state.c:
  Restore info->lock.type after call to maria_versioning.
  Fixed crash in maria_recover.test
storage/maria/maria_read_log.c:
  Don't write thread id in debug log. (Not needed as maria_read_log is a single treaded program)

Fixed bug when removing a newly inserted record (in case of duplicate key).
The bug caused a crash for rows with several blobs and the first blob was small enough to fit into the head page.
Don't change state_history if nothing changed (speed optimization that also simplifies logic).
Reset state_history if we added/deleted or updated rows without versioning.
Fixed wrong test in trnman_exists_active_transactions() if state is visible or not.

Other bugs fixed:
Fixed wrong argument to (lock->get_status) when we had to wait for TL_WRITE_CONCURRENT_INSERT.
Item_equal::update_used_tables() didn't calculate const_item_cache properly.
Added assert's to detect if join_read_const_table() was called under wrong assumptions..
Fixed that _ma_setup_live_state() is called from thr_lock() instead of handler::external_lock().
This was needed to get versioning information to be setup correctly.
Fixed error in debug binaries during a call to _ma_check_table_is_closed() when another thread was opening/closing a table.
Fixed wrong test when finding right history_state to use.

mysql-test/suite/maria/r/maria.result:
  Added test for Bug#40311 Assert in MARIA_RECORD_POS during pushbuild 2 test
mysql-test/suite/maria/t/maria.test:
  Added test for Bug#40311 Assert in MARIA_RECORD_POS during pushbuild 2 test
mysys/thr_lock.c:
  Fixed wrong argument to (lock->get_status) when we had to wait for TL_WRITE_CONCURRENT_INSERT
sql/item_cmpfunc.cc:
  Item_equal::update_used_tables() didn't calculate const_item_cache properly, which later caused a wrong result for item->const_item()
sql/sql_base.cc:
  In debug mode, Initilize record buffer with unexpected data to catch usage of uninitialized memory
sql/sql_select.cc:
  Fixed indentation
  Added assert's to detect if join_read_const_table() was called under wrong assumptions.
  One assert() is disabled for now as Item_equal() doesn't behave as expected.
storage/maria/ha_maria.cc:
  Move calling to _ma_setup_live_state() to ma_state.c::_ma_block_get_status()
  This was needed as _ma_setup_live_state() needed to know if the table will be used concurrently or not
storage/maria/ma_blockrec.c:
  Fixed bug when removing a newly inserted record (in case of duplicate key).
  The bug caused a crash for rows with several blobs and the first blob was small enough to fit into the head page.
storage/maria/ma_dbug.c:
  Added mutex to protect the open table list during _ma_check_table_is_closed().
  Without the protection we could get a error in debug binaries during a call to _ma_check_table_is_closed()
storage/maria/ma_delete_table.c:
  Removed not used code
storage/maria/ma_rename.c:
  Removed not used code
storage/maria/ma_state.c:
  Fixed wrong test when finding right history_state to use
  Mark in tables->state_current.no_transid if we are using transid's or not.
  Don't change state_history if nothing changed (speed optimization that also simplifies logic)
  Reset state_history if we added/deleted or updated rows without versioning.
  More DBUG_ASSERT's and more DBUG
  Updated maria_versioning() to initialize environment before calling _ma_blok_get_status(). This was needed because of the new logic in _ma_block_get_status()
storage/maria/ma_state.h:
  Added flags to detect if table changed and/or if we changed table without versioning
storage/maria/ma_write.c:
  Simple cleanups (No logic changes)
storage/maria/trnman.c:
  Fixed wrong test in trnman_exists_active_transactions() if state is visible or not.