Commit graph

180051 commits

Author SHA1 Message Date
Marko Mäkelä
a778a62822 Re-record a result; repl.proto_max is 8, not 7 2018-02-13 16:39:51 +02:00
Alexander Barkov
991649e3ed A cleanup for MDEV-15287: removung unused code
The patch for MDEV-15287 made the code in Item_func_min_max::get_date_native()
dead, as the TIME data type is now handled in get_time_native().
Removing the TIME related code from get_date_native().
2018-02-13 09:33:06 +04:00
Monty
54db0be3be Added Max_index_length and Temporary to SHOW TABLE STATUS
- Max_index_length is supported by MyISAM and Aria tables.
- Temporary is a placeholder to signal that a table is a
  temporary table. For the moment this is always "N", except
  "Y" for generated information_schema tables and NULL for
  views. Full temporary table support will be done in another task.
  (No reason to have to update a lot of result files twice in a row)
2018-02-12 17:17:26 +02:00
Monty
19dd14e648 Updated galera .result files 2018-02-12 17:17:26 +02:00
Alexander Barkov
4087683e1a MDEV-15287 Bad result for LEAST/GREATEST(datetime_alike_string, time) 2018-02-12 17:55:48 +04:00
Alexander Barkov
da99e086f9 Merge remote-tracking branch 'origin/10.2' into bb-10.2-ext 2018-02-12 10:03:28 +04:00
Elena Stepanova
7a106d1961 Updated list of unstable tests for 10.2.13 2018-02-12 03:20:09 +02:00
Sergei Golubchik
49bcc82686 Merge branch '10.1' into 10.2 2018-02-11 13:47:16 +01:00
Alexey Botchkov
b88542681b MDEV-14611 ALTER TABLE EXCHANGE PARTITION does not work properly when used with DATA DIRECTORY.
When table is renamed, the InnoDB's dictionary cache didn't
        change the ib_table->data_dir_path accordingly.
        Now it's set to NULL.
2018-02-10 22:17:49 +04:00
Alexander Barkov
a2feeb3d6f After-merge fixes for "sys_vars.sysvars_innodb '32bit,xtradb'" test results. 2018-02-10 21:51:24 +04:00
Monty
7ed0156384 Removed compiler warnings 2018-02-10 14:32:24 +02:00
Monty
76dc6af24c TokuDB didn't compile with valgrind
TokuDB uses USE_VALGRIND while MariaDB uses HAVE_valgrind

Fixed by defining USE_VALGRIND in TokuDB if HAVE_valgrind
is used
2018-02-10 14:32:24 +02:00
Monty
12d5307e95 MDEV-13508 ALTER TABLE that renames columns and CHECK constraints
Fixed by adding Item::rename_fields_processor

Signed-off-by: Monty <monty@mariadb.org>
2018-02-10 14:32:24 +02:00
Alexander Barkov
e07451f71f Merge remote-tracking branch 'origin/10.0' into 10.1 2018-02-10 00:06:10 +04:00
Alexander Barkov
b0a92333c0 MDEV-15262 Wrong results for SELECT..WHERE non_indexed_datetime_column=indexed_time_column 2018-02-09 19:47:00 +04:00
Jan Lindström
564891c532 MDEV-14508: encryption.innodb-compressed-blob failed in buildbot, assertion in btr0cur.cc line 1398
Before that line there is call to buf_page_get_gen that could
return block = NULL when decrypting a page fails. However,
we should set error to be != DB_SUCCESS also. In error log
there was error about decompression but in that code there
is one case where error is not set correctly.
2018-02-09 17:17:32 +02:00
Marko Mäkelä
5559905d41 Merge 10.2 into bb-10.2-ext 2018-02-09 12:48:23 +02:00
Sergei Petrunia
7beaa5e34e Merge branch 'bb-10.2-mariarocks' into 10.2 2018-02-09 13:44:38 +03:00
Marko Mäkelä
7c1bd8fa04 Add back (and clean up) the test innodb_zip.prefix_index_liftedlimit
This MySQL 5.5 test innodb_zip.innodb_prefix_index_lifted
was renamed in MySQL 5.7. In
commit 2e814d4702
the test was inadvertently removed, instead of being renamed.

The absence of this test caused a regression in MariaDB 10.2:
MDEV-15257 Invalid CREATE INDEX fails to report error correctly
2018-02-09 10:03:53 +02:00
Marko Mäkelä
e4da20d438 MDEV-14238 Bogus assertion on row_get_rec_trx_id()
page_zip_write_rec(): Do not attempt to access a non-existing
DB_TRX_ID column when writing a record to a non-leaf page.
2018-02-09 08:26:47 +02:00
Sergei Golubchik
b75d8453d4 MDEV-14868 MariaDB server crashes after using ROLLBACK TO when encrypt_tmp_files=ON
add a test that restores the data from a binlog
2018-02-08 22:54:37 +01:00
Sergei Golubchik
60dfe12be3 MDEV-14868 MariaDB server crashes after using ROLLBACK TO when encrypt_tmp_files=ON
Fix reinit_io_cache(WRITE_CACHE) with non-zero seek_offset.
Run encryption.tempfiles with and without binlog checksums.
2018-02-08 22:54:37 +01:00
Sergei Golubchik
47d1679ac6 fix encryption.tempfiles to check that encrypt_tmp_files is ON 2018-02-08 22:54:37 +01:00
Vicențiu Ciorbaru
6f0b316fbe Update wrong xtradb version
The update was lost during merge.
2018-02-08 21:12:11 +02:00
Igor Babaev
8411a8ff60 Corrected the patch for mdev-15119 that caused a failure for
cte_nonrecursive.test with --embedded.

This also fixed some problems for embedded CTEs.
Adjusted test results accordingly.
2018-02-08 09:48:36 -08:00
Jan Lindström
06d77eb43a MDEV-14427: encryption.innodb-bad-key-change failed in buildbot
Timing problem as sometimes table is marked as encrypted but
sometimes we are not sure and table is just marked missing.
2018-02-08 18:53:01 +02:00
Marko Mäkelä
c567369ad7 Merge 10.2 into bb-10.2-ext 2018-02-08 17:48:19 +02:00
Alexander Barkov
3cad31f2a7 Merge remote-tracking branch 'origin/10.2' into bb-10.2-ext 2018-02-08 19:06:25 +04:00
Vladislav Vaintroub
f01ce62c8f add mdev-504 to unstable tests, to fix appveyor 2018-02-08 13:46:43 +00:00
Vladislav Vaintroub
76600199cb rocksdb : Disable the constantly failing bloomfilter test, until MDEV-14562 is fixed 2018-02-08 13:05:13 +00:00
Otto Kekäläinen
cca9b3a933
Merge pull request #595 from MariaDB/add_myrocks_gotbackup
Add myrocks hotbackup
2018-02-08 13:57:46 +02:00
Marko Mäkelä
db25305780 MDEV-14407 Assertion failure during rollback
Rollback attempted to dereference DB_ROLL_PTR=0, which cannot possibly
be a valid undo log pointer. A safe canonical value would be
roll_ptr_t(1) << ROLL_PTR_INSERT_FLAG_POS
which is what was chosen in MDEV-12288.

This bug was reproduced in 10.3 only. Potentially, the problem could
have been introduced by MDEV-11415, which suppresses undo logging for
ALGORITHM=COPY operations. In those operations, we should actually
have written the safe value of DB_ROLL_PTR instead of writing 0.
However, the test in commit 5421e3aee7
demonstrates that access to the rebuilt table by earlier-started
transactions should actually have been refused with ER_TABLE_DEF_CHANGED.

btr_cur_ins_lock_and_undo(): When undo logging is disabled, use the
safe value of DB_ROLL_PTR.

btr_cur_optimistic_insert(): Validate the DB_TRX_ID,DB_ROLL_PTR before
inserting into a clustered index leaf page.

ins_node_t::sys_buf[]: Replaces row_id_buf and trx_id_buf and some
heap usage.

row_ins_alloc_sys_fields(): Initialize ins_node_t::sys_buf[].

trx_undo_page_report_modify(): Assert that the DB_ROLL_PTR is not 0.

trx_undo_get_undo_rec_low(): Assert that the roll_ptr is valid before
trying to dereference it.

dict_index_t::is_primary(): Check if the index is the primary key.
2018-02-08 13:55:57 +02:00
Marko Mäkelä
be6307c0fa Merge 10.1 into 10.2 2018-02-08 13:53:21 +02:00
Marko Mäkelä
871f2a6ee2 Merge 10.0 into 10.1 2018-02-08 13:29:08 +02:00
Marko Mäkelä
9216a4f69f Make the test innodb.recovery_shutdown more robust
Before killing the server, we have to FLUSH TABLES in order to
prevent the corruption of any MyISAM system tables.
2018-02-08 13:26:44 +02:00
Marko Mäkelä
5421e3aee7 MDEV-15249 Crash in MVCC read after IMPORT TABLESPACE
PageConverter::adjust_cluster_record(): Instead of writing
the invalid value DB_ROLL_PTR=0, write a value that indicates
a fresh insert, that is, prevents the DB_ROLL_PTR from being
dereferenced in any circumstances.

It can be argued that IMPORT TABLESPACE should actually
update the dict_index_t::trx_id to prevent older transactions
from accessing the table, similar to what I did on table
rebuild in MySQL 5.6.6 in
03f81a55f2
2018-02-08 12:59:32 +02:00
Oleksandr Byelkin
b6455479e5 MDEV-15230: column_json breaks cyrillic in 10.1.31
Use unsigned comparison.
2018-02-07 18:14:45 +01:00
Monty
5c057b3fef Disabled galera_ist_progress at it always fails in 10.2 2018-02-07 17:25:46 +02:00
Vicențiu Ciorbaru
5a7b6db671 Merge remote-tracking branch 'connect/10.2' into 10.2 2018-02-07 15:19:53 +02:00
Vicențiu Ciorbaru
1c2a2d3c9d Update Connector/C 2018-02-07 15:19:45 +02:00
Vicentiu Ciorbaru
adbf6caad6 debian/control Include myrocks_hotbackup into mariadb-plugin-rocksdb 2018-02-07 13:08:19 +00:00
Vicentiu Ciorbaru
cbe566dc42 Add myrocks_hotbackup part of rocksdb plugin in cmake 2018-02-07 13:08:19 +00:00
Vicentiu Ciorbaru
699123fdc3 Fix rocksdb compiler version identification with GCC
GCC 7 with only dumpversion flag returns the major version only.
dumpfullversion flag makes it work with both new and old gcc.
2018-02-07 13:07:34 +00:00
Daniel Black
d743b61f89 MDEV-14567: CRYPTO_set_mem_functions fails in FIPS mode
If the CRYPTO_set_mem_functions fails just return success
from check_openssl_compatibility. The only case where
CRYPTO_set_mem_functions fails is the allow_customize==0 (aka FIPS
mode). The check_openssl_compatibility isn't able to complete unless
this function returns success.

ref: https://github.com/openssl/openssl/blob/OpenSSL_1_1_0g/crypto/mem.c#L34
2018-02-07 11:45:01 +01:00
Marko Mäkelä
10590dd39c MDEV-15199 Referential integrity broken in ON DELETE CASCADE
MDEV-14222 Unnecessary 'cascade' memory allocation for every updated row
when there is no FOREIGN KEY

This reverts the MySQL 5.7.2 change
377774689b
which introduced these problems. MariaDB 10.2.2 inherited these problems
in commit 2e814d4702.

The FOREIGN KEY CASCADE and SET NULL operations implemented as
procedural recursion are consuming more than 8 kilobytes of stack
(9 stack frames) per iteration in a non-debug GNU/Linux AMD64 build.
This is why we need to limit the maximum recursion depth to 15 steps
instead of the 255 that it used to be in MySQL 5.7 and MariaDB 10.2.

A corresponding change was made in MySQL 5.7.21 in
7b26dc98a6
2018-02-07 10:39:12 +02:00
Marko Mäkelä
5d7e9fd46c InnoDB UPDATE cleanup
row_upd_store_v_row(): Declare static

row_upd_clust_rec_by_insert(), row_upd_del_mark_clust_rec():
Pass the parameter 'foreign' only #ifdef WITH_WSREP
2018-02-07 10:39:12 +02:00
Marko Mäkelä
b68dac88b3 MDEV-15219 FOREIGN KEY CASCADE or SET NULL operations will not resume after lock wait
This corruption was introduced in MDEV-13331. It would have been caught
by the MySQL 5.7 test innodb.update-cascade which MariaDB was missing
until now.

row_ins_check_foreign_constraint(): Never replace err == DB_LOCK_WAIT
with other values than DB_LOCK_WAIT_TIMEOUT.
2018-02-07 10:39:12 +02:00
Marko Mäkelä
1789e0ff03 Remove useless debug instrumentation row_print_geometry_data 2018-02-07 10:39:12 +02:00
Marko Mäkelä
5815c8ee85 Remove useless code
srv_mbr_print(): Remove. Geometry fields are already being output
in raw form, and we do not need bogus MBR output for non-spatial
indexes.
2018-02-07 10:39:12 +02:00
Marko Mäkelä
4cec63af44 Foreign key code cleanup
row_ins_cascade_calc_update_vec(): Remove the output parameter
fts_col_affected, and instead return whether any fulltext index
is affected by the cascade operation.

row_ins_foreign_check_on_constraint(): Narrow the scope of some
variables.

ib_dec_in_dtor: Remove.
2018-02-07 10:39:12 +02:00