Valgrind warning happpens because of uninitialized null bytes.
In row_sel_push_cache_row_for_mysql() function we fill fetch cache
with necessary field values, row_sel_store_mysql_rec() is called
for this and leaves null bytes untouched.
Later row_sel_pop_cached_row_for_mysql() rewrites table record
buffer with uninited null bytes. We can see the problem from the
test case:
At 'SELECT...' we call row_sel_push...->row_sel_store...->row_sel_pop_cached...
chain which rewrites table->record[0] buffer with uninitialized null bytes.
When we call 'UPDATE...' statement, compare_record uses this buffer and
valgrind warning occurs.
The fix is to init null bytes with default values.
for InnoDB plugin
dict_load_table(): Pass the correct tablespace flags to
fil_open_single_table_tablespace(). For ROW_FORMAT=COMPACT and REDUNDANT,
the tablespace flags are 0. The table flags would be 0 or DICT_TF_COMPACT.
In semi-consistent read, only unlock freshly locked non-matching records.
lock_rec_lock_fast(): Return LOCK_REC_SUCCESS,
LOCK_REC_SUCCESS_CREATED, or LOCK_REC_FAIL instead of TRUE/FALSE.
enum db_err: Add DB_SUCCESS_LOCKED_REC for indicating a successful
operation where a record lock was created.
lock_sec_rec_read_check_and_lock(),
lock_clust_rec_read_check_and_lock(), lock_rec_enqueue_waiting(),
lock_rec_lock_slow(), lock_rec_lock(), row_ins_set_shared_rec_lock(),
row_ins_set_exclusive_rec_lock(), sel_set_rec_lock(),
row_sel_get_clust_rec_for_mysql(): Return DB_SUCCESS_LOCKED_REC if a
new record lock was created. Adjust callers.
row_unlock_for_mysql(): Correct the function documentation.
row_prebuilt_t::new_rec_locks: Correct the documentation.
In semi-consistent read, only unlock freshly locked non-matching records.
Define DB_SUCCESS_LOCKED_REC for indicating a successful operation
where a record lock was created.
lock_rec_lock_fast(): Return LOCK_REC_SUCCESS,
LOCK_REC_SUCCESS_CREATED, or LOCK_REC_FAIL instead of TRUE/FALSE.
lock_sec_rec_read_check_and_lock(),
lock_clust_rec_read_check_and_lock(), lock_rec_enqueue_waiting(),
lock_rec_lock_slow(), lock_rec_lock(), row_ins_set_shared_rec_lock(),
row_ins_set_exclusive_rec_lock(), sel_set_rec_lock(),
row_sel_get_clust_rec_for_mysql(): Return DB_SUCCESS_LOCKED_REC if a
new record lock was created. Adjust callers.
row_unlock_for_mysql(): Correct the function documentation.
row_prebuilt_t::new_rec_locks: Correct the documentation.
lock_rec_unlock(): Cache first_lock and rewrite while() loops as for().
btr_cur_optimistic_update(): Use common error handling return.
row_create_prebuilt(): Add Valgrind instrumentation.
------------------------------------------------------------
revno: 3127
revision-id: vasil.dimov@oracle.com-20100531152341-x2d4hma644icamh1
parent: vasil.dimov@oracle.com-20100531105923-kpjwl4rbgfpfj13c
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk-innodb
timestamp: Mon 2010-05-31 18:23:41 +0300
message:
Fix Bug #53947 InnoDB: Assertion failure in thread 4224 in file .\sync\sync0sync.c line 324
Destroy the rw-lock object before freeing the memory it is occupying.
If we do not do this, then the mutex that is contained in the rw-lock
object btr_search_latch_temp->mutex gets "freed" and subsequently
mutex_free() from sync_close() hits a mutex whose memory has been
freed and crashes.
Approved by: Heikki (via IRC)
Discussed with: Calvin
err_index could be not a member of the share structure or prebuilt
structure passed from MySQL. For now, we resort to the traditional
way of scanning index->table for the index number.
Add code to waiting for a set of errors.
Add code to waiting for an error instead of waiting for io thread to stop, as
after 'START SLAVE', the status of io thread is still not running.
But it doesn't mean slave io thread encounters an error.
without FOR UPDATE is causing a lock".
SELECT statements with subqueries referencing InnoDB tables
were acquiring shared locks on rows in these tables when they
were executed in REPEATABLE-READ mode and with statement or
mixed mode binary logging turned on.
This was a regression which were introduced when fixing
bug 39843.
The problem was that for tables belonging to subqueries
parser set TL_READ_DEFAULT as a lock type. In cases when
statement/mixed binary logging at open_tables() time this
type of lock was converted to TL_READ_NO_INSERT lock at
open_tables() time and caused InnoDB engine to acquire
shared locks on reads from these tables. Although in some
cases such behavior was correct (e.g. for subqueries in
DELETE) in case of SELECT it has caused unnecessary locking.
This patch implements minimal version of the fix for the
specific problem described in the bug-report which supposed
to be not too risky for pushing into 5.1 tree.
The 5.5 tree already contains a more appropriate solution
which also addresses other related issues like bug 53921
"Wrong locks for SELECTs used stored functions may lead
to broken SBR".
This patch tries to solve the problem by ensuring that
TL_READ_DEFAULT lock which is set in the parser for
tables participating in subqueries at open_tables()
time is interpreted as TL_READ_NO_INSERT or TL_READ.
TL_READ is used only if we know that this is a SELECT
and that this particular table is not used by a stored
function.
Test coverage is added for both InnoDB and MyISAM.
This patch introduces an "incompatible" change in locking
scheme for subqueries used in SELECT ... FOR UPDATE and
SELECT .. IN SHARE MODE.
In 4.1 (as well as in 5.0 and 5.1 before fix for bug 39843)
the server would use a snapshot InnoDB read for subqueries
in SELECT FOR UPDATE and SELECT .. IN SHARE MODE statements,
regardless of whether the binary log is on or off.
If the user required a different type of read (i.e. locking
read), he/she could request so explicitly by providing FOR
UPDATE/IN SHARE MODE clause for each individual subquery.
The patch for bug 39843 broke this behaviour (which was not
documented or tested), and started to use locking reads for
all subqueries in SELECT ... FOR UPDATE/IN SHARE MODE.
This patch restores 4.1 behaviour.
This patch should be mostly null-merged into 5.5 tree.
There are two problems:
1. In simplify_joins function we calculate table dependencies. If STRAIGHT_JOIN hint
is used for whole SELECT we do not count it and as result some dependendecies
might be lost. It leads to incorrect table order which is returned by
join_tab_cmp_straight() function.
2. make_join_statistics() calculate the transitive closure for relations a particular
JOIN_TAB is 'dependent on'.
We aggregate the dependent table_map of a JOIN_TAB by adding dependencies from other
tables which we depend on. However, this may also cause new dependencies to be
available after we have completed processing a certain JOIN_TAB.
Both these problems affect condition pushdown and as result condition might be pushed
into wrong table which leads to crash or even omitted which leads to wrong result.
The fix:
1. Use modified 'transitive closure' algorithm provided by Ole John Aske
2. Update table dependences in simplify_joins according to
global STRAIGHT_JOIN hint.
Note: the patch also fixes bugs 46091 & 51492
