FAILED IN DEACTIVATE_DDL_LOG_ENTRY
deallocate_ddl_log_entry() can be called without having
locked LOCK_gdl. It uses a global buffer for reading and
writing entries in the ddl_log, and since it is not protected
by any mutex, two concurrent threads can overwrite the
content in the global buffer, so it can be different from
what was read.
Thread a reads from entry 1 into global
buffer, thread b reads from entry 2 into global buffer,
thread a writes from global buffer into entry 1
-> entry 1 is not the content of entry 2.
This is especially bad for replace entries, which uses
two phases, and does not deactivate the whole entry
after the first phase, but increases the phase instead.
Fixed by using thread local storage (stack) instead of global
storage (global buffer).
Also added buffer and size arguments to
read/write_ddl_log_file_entry.
Also only read/write first bytes in entries in
deactivate_ddl_log_entry.
Also fixed the scenario where it will try to recover from a server
compiled with a different value of IO_SIZE (very uncommon!)
updated patch with set_ddl_log_entry_from_buf
and removed read_ddl_log_entry.
Manually tested, no test case included.
SMALL KEY CACHE
The server crashed on division by zero because the key cache was not
initialized and the block length was 0 which was used in a division.
The fix was to not allow CACHE INDEX if the key cache was not initiallized.
Thus never try LOAD INDEX INTO CACHE for an uninitialized key cache.
Also added some windows files/directories to .bzrignore.
TO POSITION FIRST CAN CAUSE DATA TO BE CORRUPTED".
ALTER TABLE MODIFY/CHANGE ... FIRST did nothing except renaming
columns if new version of the table had exactly the same
structure as the old one (i.e. as result of such statement, names
of columns changed their order as specified but data in columns
didn't). The same thing happened for ALTER TABLE DROP COLUMN/ADD
COLUMN statements which were supposed to produce new version of
table with exactly the same structure as the old version of table.
I.e. in the latter case the result was the same as if old column
was renamed instead of being dropped and new column with default
as value being created.
Both these problems were caused by the fact that ALTER TABLE
implementation incorrectly interpreted both these situations as
simple renaming of columns and assumed that in-place ALTER TABLE
algorithm could have been used for them.
This patch fixes this problem by ensuring that in cases when some
column is moved to the first position or some column is dropped
the default ALTER TABLE algorithm involving table copying is
always used. This is achieved by detecting such situations in
mysql_prepare_alter_table() and setting Alter_info::change_level
to ALTER_TABLE_DATA_CHANGED for them.
In sql_class.cc, 'row_count', of type 'ha_rows', was used as last argument for
ER_TRUNCATED_WRONG_VALUE_FOR_FIELD which is
"Incorrect %-.32s value: '%-.128s' for column '%.192s' at row %ld".
So 'ha_rows' was used as 'long'.
On SPARC32 Solaris builds, 'long' is 4 bytes and 'ha_rows' is 'longlong' i.e. 8 bytes.
So the printf-like code was reading only the first 4 bytes.
Because the CPU is big-endian, 1LL is 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x01
so the first four bytes yield 0. So the warning message had "row 0" instead of
"row 1" in test outfile_loaddata.test:
-Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 1
+Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 0
All error-messaging functions which internally invoke some printf-life function
are potential candidate for such mistakes.
One apparently easy way to catch such mistakes is to use
ATTRIBUTE_FORMAT (from my_attribute.h).
But this works only when call site has both:
a) the format as a string literal
b) the types of arguments.
So:
func(ER(ER_BLAH), 10);
will silently not be checked, because ER(ER_BLAH) is not known at
compile time (it is known at run-time, and depends on the chosen
language).
And
func("%s", a va_list argument);
has the same problem, as the *real* type of arguments is not
known at this site at compile time (it's known in some caller).
Moreover,
func(ER(ER_BLAH));
though possibly correct (if ER(ER_BLAH) has no '%' markers), will not
compile (gcc says "error: format not a string literal and no format
arguments").
Consequences:
1) ATTRIBUTE_FORMAT is here added only to functions which in practice
take "string literal" formats: "my_error_reporter" and "print_admin_msg".
2) it cannot be added to the other functions: my_error(),
push_warning_printf(), Table_check_intact::report_error(),
general_log_print().
To do a one-time check of functions listed in (2), the following
"static code analysis" has been done:
1) replace
my_error(ER_xxx, arguments for substitution in format)
with the equivalent
my_printf_error(ER_xxx,ER(ER_xxx), arguments for substitution in
format),
so that we have ER(ER_xxx) and the arguments *in the same call site*
2) add ATTRIBUTE_FORMAT to push_warning_printf(),
Table_check_intact::report_error(), general_log_print()
3) replace ER(xxx) with the hard-coded English text found in
errmsg.txt (like: ER(ER_UNKNOWN_ERROR) is replaced with
"Unknown error"), so that a call site has the format as string literal
4) this way, ATTRIBUTE_FORMAT can effectively do its job
5) compile, fix errors detected by ATTRIBUTE_FORMAT
6) revert steps 1-2-3.
The present patch has no compiler error when submitted again to the
static code analysis above.
It cannot catch all problems though: see Field::set_warning(), in
which a call to push_warning_printf() has a variable error
(thus, not replacable by a string literal); I checked set_warning() calls
by hand though.
See also WL 5883 for one proposal to avoid such bugs from appearing
again in the future.
The issues fixed in the patch are:
a) mismatch in types (like 'int' passed to '%ld')
b) more arguments passed than specified in the format.
This patch resolves mismatches by changing the type/number of arguments,
not by changing error messages of sql/share/errmsg.txt. The latter would be wrong,
per the following old rule: errmsg.txt must be as stable as possible; no insertions
or deletions of messages, no changes of type or number of printf-like format specifiers,
are allowed, as long as the change impacts a message already released in a GA version.
If this rule is not followed:
- Connectors, which use error message numbers, will be confused (by insertions/deletions
of messages)
- using errmsg.sys of MySQL 5.1.n with mysqld of MySQL 5.1.(n+1)
could produce wrong messages or crash; such usage can easily happen if
installing 5.1.(n+1) while /etc/my.cnf still has --language=/path/to/5.1.n/xxx;
or if copying mysqld from 5.1.(n+1) into a 5.1.n installation.
When fixing b), I have verified that the superfluous arguments were not used in the format
in the first 5.1 GA (5.1.30 'bteam@astra04-20081114162938-z8mctjp6st27uobm').
Had they been used, then passing them today, even if the message doesn't use them
anymore, would have been necessary, as explained above.
This is a backport of the patch for MySQL Bug#50574.
Adding a SPATIAL INDEX on non-geometrical columns caused a
segmentation fault when the table was subsequently
inserted into.
A test was added in mysql_prepare_create_table to explicitly
check whether non-geometrical columns are used in a
spatial index, and throw an error if so.
For MySQL 5.5 and later, a new and more meaningful error
message was introduced. For 5.1, we (re-)use an existing
error code.
ALTER TABLE RENAME, DISABLE KEYS.
The code of ALTER TABLE RENAME, DISABLE KEYS could
issue a commit while holding LOCK_open mutex.
This is a regression introduced by the fix for
Bug 54453.
This failed an assert guarding us against a potential
deadlock with connections trying to execute
FLUSH TABLES WITH READ LOCK.
The fix is to move acquisition of LOCK_open outside
the section that issues ha_autocommit_or_rollback().
LOCK_open is taken to protect against concurrent
operations with .frms and the table definition
cache, and doesn't need to cover the call to commit.
A test case added to innodb_mysql.test.
The patch is to be null-merged to 5.5, which
already has 54453 null-merged to it.
data dictionary confusion
On file systems with case insensitive file names, and
lower_case_table_names set to '2', the server could crash
due to a table definition cache inconsistency. This is
the default setting on MacOSX, but may also be set and
used on MS Windows.
The bug is caused by using two different strategies for
creating the hash key for the table definition cache, resulting
in failure to look up an entry which is present in the cache,
or failure to delete an existing entry. One strategy was to
use the real table name (with case preserved), and the other
to use a normalized table name (i.e a lower case version).
This is manifested in two cases. One is during 'DROP DATABASE',
where all known files are removed. The removal from
the table definition cache is done via a generated list of
TABLE_LIST with keys (wrongly) created using the case preserved
name. The other is during CREATE TABLE, where the cache lookup
is also (wrongly) based on the case preserved name.
The fix was to use only the normalized table name when
creating hash keys.
adding new indexes
A fast alter table requires that the existing (old) table
and indices are unchanged (i.e only new indices can be
added). To verify this, the layout and flags of the old
table/indices are compared for equality with the new.
The PACK_KEYS option is a no-op in InnoDB, but the flag
exists, and is used in the table compare. We need to
check this (table) option flag before deciding whether an
index should be packed or not. If the table has
explicitly set PACK_KEYS to 0, the created indices should
not be marked as packed/packable.
Fix warnings flagged by the new warning option -Wunused-but-set-variable
that was added to GCC 4.6 and that is enabled by -Wunused and -Wall. The
option causes a warning whenever a local variable is assigned to but is
later unused. It also warns about meaningless pointer dereferences.
table with active trx
Essentially, the problem is that InnoDB does a implicit commit
when a cursor (table handler) is unlocked/closed, creating
a dissonance between the transaction state within the server
layer and the storage engine layer. Theoretically, a statement
transaction can encompass several table instances in a similar
manner to a multiple statement transaction, hence it does not
make sense to limit a statement transaction to the lifetime of
the table instances (cursors) used within it.
Since this particular instance of the problem is only triggerable
on 5.1 and is masked on 5.5 due 2PC being skipped (assertion is in
the prepare phase of a 2PC), the solution (which is less risky) is
to explicitly end the transaction before the cached table is unlock
on rename table.
The patch is to be null merged into trunk.
This crash occured after ALTER TABLE was used on a temporary
transactional table locked by LOCK TABLES. Any later attempts to
execute LOCK/UNLOCK TABLES, caused the server to crash.
The reason for the crash was the list of locked tables would
end up having a pointer to a free'd table instance. This happened
because ALTER TABLE deleted the table without also removing the
table reference from the locked tables list.
This patch fixes the problem by making sure ALTER TABLE also
removes the table from the locked tables list.
Test case added to innodb_mysql.test.
strict aliasing violations.
One somewhat major source of strict-aliasing violations and
related warnings is the SQL_LIST structure. For example,
consider its member function `link_in_list` which takes
a pointer to pointer of type T (any type) as a pointer to
pointer to unsigned char. Dereferencing this pointer, which
is done to reset the next field, violates strict-aliasing
rules and might cause problems for surrounding code that
uses the next field of the object being added to the list.
The solution is to use templates to parametrize the SQL_LIST
structure in order to deference the pointers with compatible
types. As a side bonus, it becomes possible to remove quite
a few casts related to acessing data members of SQL_LIST.
data directory name command
The check_db_name function has been modified to validate tails of
#mysql50#-prefixed database names for compliance with MySQL 5.0
database name encoding rules (the check_table_name function call
has been reused).
This is the 5.1 merge and extension of the fix.
The server was happily accepting paths in table name in all places a table
name is accepted (e.g. a SELECT). This allowed all users that have some
privilege over some database to read all tables in all databases in all
mysql server instances that the server file system has access to.
Fixed by :
1. making sure no path elements are allowed in quoted table name when
constructing the path (note that the path symbols are still valid in table names
when they're properly escaped by the server).
2. checking the #mysql50# prefixed names the same way they're checked for
path elements in mysql-5.0.
Problem: ALTER TABLE ADD INDEX may lead to table copying if there's
numeric field(s) with non-default display width modificator specified.
Fix: compare numeric field's storage lenghts when we decide whether
they can be considered 'equal' for table alteration purposes.
definition at engine
If a single ALTER TABLE contains both DROP INDEX and ADD INDEX using
the same index name (a.k.a. index modification) we need to disable
in-place alter table because we can't ask the storage engine to have
two copies of the index with the same name even temporarily (if we
first do the ADD INDEX and then DROP INDEX) and we can't modify
indexes that are needed by e.g. foreign keys if we first do
DROP INDEX and then ADD INDEX.
Fixed the problem by disabling in-place ALTER TABLE for these cases.
for same data when using bit fields
Problem: checksum for BIT fields may be computed incorrectly
in some cases due to its storage peculiarity.
Fix: convert a BIT field to a string then calculate its checksum.
There was two problems:
The first was the symptom, caused by bad error handling in
ha_partition. It did not handle print_error etc. when
having no partitions (when used by dummy handler).
The second was the real problem that when dropping tables
it reused the table type (storage engine) from when the lock
was asked for, not the table type that it had when gaining
the exclusive name lock. So that it tried to delete tables
from wrong storage engines.
Solutions for the first problem was to accept some handler
calls to the partitioning handler even if it was not setup
with any partitions, and also if possible fallback
to use the base handler's default functions.
Solution for the second problem was to remove the optimization
to reuse the definition from the cache, instead always check
the frm-file when holding the LOCK_open mutex
(updated with a fix for a debug print crash and better
comments as required by reviewer, and removed optimization
to avoid reading the frm-file).
Rename method as to not hide a base.
Reorder attributes initialization.
Remove unused variable.
Rework code to silence a warning due to assignment used as truth value.
REORGANIZE PARTITION
There were several problems which lead to this this,
all related to bad error handling.
1) There was several bugs preventing the ddl-log to be used for
cleaning up created files on error.
2) The error handling after the copy partition rows did not close
and unlock the tables, resulting in deletion of partitions
which were in use, which lead InnoDB to put the partition to
drop in a background queue.
The problem is a somewhat common misusage of the strmake function.
The strmake(dst, src, len) function writes at most /len/ bytes to
the string pointed to by src, not including the trailing null byte.
Hence, if /len/ is the exact length of the destination buffer, a
one byte buffer overflow can occur if the length of the source
string is equal to or greater than /len/.
with temporary table and partitions
It was possible to create temporary partitioned tables
via create table ... like ... (which is not allowed with
create temporary table). This lead to a new HA_EXTRA flag
(HA_EXTRA_MMAP) was sent to the partitioning handler,
which was caught on an assert in debug builds.
Solution was to check for partitioned tables when
doing create table ... like ... and disallow it.
<tmp_tbl> with RBL
When binlogging the statement, the server always handle the existing
object as a table, even though it is a view. However a view is
handled differently in other parts of the code thus leading the
statement to crash in RBL if the view exists.
This happens because the underlying tables for the view are not opened
when we try to call store_create_info() on the view in order to build
a CREATE TABLE statement.
This patch will only address the crash problem, other binlogging
problems related to CREATE TABLE IF NOT EXISTS LIKE when the existing
object is a view will be solved by BUG 47442.
Implemented the server infrastructure for the fix:
1. Added a function LEX_STRING *thd_query_string(THD) to return
a LEX_STRING structure instead of char *.
This is the function that must be called in innodb instead of
thd_query()
2. Did some encapsulation in THD : aggregated thd_query and
thd_query_length into a LEX_STRING and made accessor and mutator
methods for easy code updating.
3. Updated the server code to use the new methods where applicable.
In RBR, 'DROP TEMPORARY TABLE IF EXISTS...' statement is binlogged when the table
does not exist.
In fact, 'DROP TEMPORARY TABLE ...' statement should never be binlogged in RBR
no matter if the table exists or not.
This patch addresses this by checking whether we are dropping a
temporary table or not, when building the custom drop statement.
Invalid (old?) table or database name in logs
Problem was still not completely fixed, due to
qouting.
This is the server side only fix (in explain_filename),
the change from filename_to_tablename to use explain_filename
in the InnoDB code must be done before the bug is
fixed.
checksum)"
The problem was that checksum of GEOMETRY type used memory addresses
in the computation, making it un-repeatable thus useless.
(This patch is a backport from 6.0 branch)
Despite copying the value of the old table's row type
we don't always have to mark row type as being specified.
Innodb uses this to check if it can do fast ALTER TABLE
or not.
Fixed by correctly flagging the presence of row_type
only when it's actually changed.
Added a test case for 39200.
can crash under load
Backport from 5.1.
Does also include key cache fixes from:
Bug 44068 (RESTORE can disable the MyISAM Key Cache)
Bug 40944 (Backup: crash after myisampack)
CREATE TABLE...LIKE...
The mysql server option 'sync_frm' is ignored when table is created with
syntax CREATE TABLE .. LIKE..
Fixed by adding the MY_SYNC flag and calling my_sync() from my_copy() when
the flag is set.
In mysql_create_table(), when the 'sync_frm' is set, MY_SYNC flag is passed
to my_copy().
Note: TestCase is not attached and can be tested manually using debugger.
Invalid (old?) table or database name in logs
Post push patch.
Bug was that a non partitioned table file was not
converted to system_charset, (due to table_name_len was not set).
Also missing DBUG_RETURN.
And Innodb adds quotes after calling the function,
so I added one more mode where explain_filename does not
add quotes. But it still appends the [sub]partition name
as a comment.
Also caught a minor quoting bug, the character '`' was
not quoted in the identifier. (so 'a`b' was quoted as `a`b`
and not `a``b`, this is mulitbyte characters aware.)
