When innobase_allocate_row_for_vcol() returns true (for failure),
it may already have invoked mem_heap_create(). However, some callers
would fail to invoke mem_heap_free().
PROBLEM
-------
1. We are inserting a base column entry which causes an invalid value
by the function provided to generate virtual column,but we go ahead
and insert this due to ignore keyword.
2. We then delete this record, making this record delete marked in innodb.
If we try to insert another record with the same pk as the deleted
record and if the rec is not purged ,then we try to undelete mark this
record and try to build a update vector with previous and updated value
and while calculating the value of virtual column we get error from
server that we cannot calculate this from base column.
Innodb assumes that innobase_get_computed_value() Should always return
a valid value for the base column present in the row. The failure of
this call was not handled ,so we were crashing.
FIX
dict_sys_get_size(): Replace the time-consuming loop with
a crude estimate that can be computed without holding any mutex.
Even before dict_sys->size was removed in MDEV-13325,
not all memory allocations by the InnoDB data dictionary cache
were being accounted for. One example is foreign key constraints.
Another example is virtual column metadata, starting with 10.2.
on startup innodb is checking whether files "ib_logfileN"
(for N from 1 to 100) exist, and whether they're readable.
A non-existent file aborted the scan.
A directory instead of a file made InnoDB to fail.
Now it treats "directory exists" as "file doesn't exist".
When InnoDB is invoking posix_fallocate() to extend data files, it
was missing a call to fsync() to update the file system metadata.
If file system recovery is needed, the file size could be incorrect.
When the setting innodb_flush_method=O_DIRECT_NO_FSYNC
that was introduced in MariaDB 10.0.11 (and MySQL 5.6) is enabled,
InnoDB would wrongly skip fsync() after extending files.
Furthermore, the merge commit d8b45b0c00
inadvertently removed XtraDB error checking for posix_fallocate()
which this fix is restoring.
fil_flush(): Add the parameter bool metadata=false to request that
fil_buffering_disabled() be ignored.
fil_extend_space_to_desired_size(): Invoke fil_flush() with the
extra parameter. After successful posix_fallocate(), invoke
os_file_flush(). Note: The bookkeeping for fil_flush() would not be
updated the posix_fallocate() code path, so the "redundant"
fil_flush() should be a no-op.
thd_destructor_proxy() may miss abort signal if innobase_end() is running
concurrently, which causes server hang in pthread_join() on shutdown.
The problem was that aborting wasn't protected by mutex:
proxy thr: while (!myvar->abort)
end thr: running->abort = 1;
end thr: mysql_cond_broadcast(...);
proxy thr: mysql_cond_wait(...); // nobody to awake it
end thr: pthread_join(...); // waits for proxy thr
Also made main.mysqld_option_err reentrant.
row_drop_table_for_mysql(): Fix a regression introduced in MDEV-16515.
Similar to the follow-up fixes MDEV-16647 and MDEV-17470, we must make
the internal tables of FULLTEXT INDEX immune to kills, to avoid noise
and resource leakage on DROP TABLE or ALTER TABLE. (Orphan internal tables
would be dropped at the next InnoDB startup only.)
server shutdown.
Prior to adding current thd to the global thd list, initialize
thd->mysys_var->current_cond / thd->mysys_var->current_mutex.
Otherwise thd_destructor_proxy can miss the abort condition, which is set
by the shutdown thread.
Problem:
========
MLOG_FILE_WRITE_CRYPT_DATA redo log fails to apply type for
the crypt_data present in the space. While processing the double-write
buffer pages, page fails to decrypt. It leads to warning message.
Fix:
====
Set the type while parsing MLOG_FILE_WRITE_CRYPT_DATA redo log.
If type and length is of invalid type then mark it as corrupted.
When performing a hash search via HASH_SEARCH we first look at a key of a node
and then at its pointer to the next node in chain. If we have those in one cache
line instead of a two we reduce memory reads.
I found dict_table_t, fil_space_t and buf_page_t suitable for such improvement.
This was introduced in 1a7a018939
MDEV-16557 Remove INNOBASE_SHARE::idx_trans_tbl
ha_innobase::innobase_get_index: remove incorrect assertion.
Index nullability is checked in subsequent ifs.
Closes#1079
During database recovery, a transaction with wsrep XID is
recovered from InnoDB in prepared state. However, when the
transaction is looked up with trx_get_trx_by_xid() in
innobase_commit_by_xid(), trx->xid gets cleared in
trx_get_trx_by_xid_low() and commit time serialization history
write does not update the wsrep XID in trx sys header for
that recovered trx. As a result the transaction gets
committed during recovery but the wsrep position does not
get updated appropriately.
As a fix, we preserve trx->xid for Galera over transaction
commit in recovery phase.
Fix authored by: Teemu Ollakka (GaleraCluster) and Marko Mäkelä.
modified: mysql-test/suite/galera/disabled.def
modified: mysql-test/suite/galera/r/galera_gcache_recover_full_gcache.result
modified: mysql-test/suite/galera/r/galera_gcache_recover_manytrx.result
modified: mysql-test/suite/galera/t/galera_gcache_recover_full_gcache.test
modified: mysql-test/suite/galera/t/galera_gcache_recover_manytrx.test
modified: storage/innobase/trx/trx0trx.cc
modified: storage/xtradb/trx/trx0trx.cc
This is a regression after MDEV-13671.
The bug is related to key part prefix lengths wich are stored in SYS_FIELDS.
Storage format is not obvious and was handled incorrectly which led to data
dictionary corruption.
SYS_FIELDS.POS actually contains prefix length too in case if any key part
has prefix length.
innobase_rename_column_try(): fixed prefixes handling
Tests for prefixed indexes added too.
Closes#1063
Orphan #sql* tables may remain after ALTER TABLE
was interrupted by timeout or KILL or client disconnect.
This is a regression caused by MDEV-16515.
Similar to temporary tables (MDEV-16647), we had better ignore the
KILL when dropping the original table in the final part of ALTER TABLE.
Closes#1020
This fixes a regression that was introduced in MySQL 5.6.6
in an error handling code path, in the following change:
commit 024f363d6b5f09b20d1bba411af55be95c7398d3
Author: kevin.lewis@oracle.com <>
Date: Fri Jun 15 09:01:42 2012 -0500
Bug #14169459 INNODB; DROP TABLE DOES NOT DELETE THE IBD FILE
FOR A TEMPORARY TABLE.
In MDEV-13103, I made a mistake in the error handling of
page_compressed=1 decryption when the default
innodb_compression_algorithm=zlib is used.
Due to this mistake, with certain versions of zlib,
MariaDB would fail to detect a corrupted page.
The problem was uncovered by the following tests:
mariabackup.unencrypted_page_compressed
mariabackup.encrypted_page_compressed
Before MDEV-5800 in MariaDB 10.2.2, InnoDB knew nothing about
virtual columns. But, they would be present in the TABLE_SHARE.
It appears that the MDEV-12936 fix only worked in the special case
when the virtual columns are the last columns in a table definition.
mysql_fields(): Remove.
build_template_needs_field(): Omit virtual columns if the
InnoDB table definition is known to not contain them, based
on the .frm file version.
build_template_field(): Clean up some code.
ha_innobase::build_template(): Remove redundant computations.
Loop over all MariaDB columns. Skip virtual columns if
InnoDB does not know about them.
calc_row_difference(): Skip not-known-to-InnoDB virtual columns.
wsrep_calc_row_hash(): Skip all virtual columns. This would
fix an out-of-bounds access to dict_table_t::cols[]. Also,
remove some redundant computations and variables.
create_table_check_doc_id_col(): Assert that the .frm file
for a newly created table will be in a format where InnoDB
knows about virtual columns.
Problem:
=======
Mariabackup seems to fail to verify the pages of compressed tables.
The reason is that both fil_space_verify_crypt_checksum() and
buf_page_is_corrupted() will skip the validation for compressed pages.
Fix:
====
Mariabackup should call fil_page_decompress() for compressed and encrypted
compressed page. After that, call buf_page_is_corrupted() to
check the page corruption.
Starting with commit 6b6987154a
the encrypted page checksum is only computed with crc32.
Encrypted data pages that were written earlier can contain
other checksums, but new ones will only contain crc32.
Because of this, it does not make sense to implement strict
checks of innodb_checksum_algorithm for other than strict_crc32.
fil_space_verify_crypt_checksum(): Treat strict_innodb as innodb
and strict_none as none. That is, allow a match from any of the
algorithms none, innodb, crc32. (This is how it worked before the
second MDEV-12112 fix.)
Thanks to Thirunarayanan Balathandayuthapani for pointing this out.
