Commit graph

85693 commits

Author SHA1 Message Date
Balasubramanian Kandasamy
ce3c0ade5f Bug#18327194 - Resolve confict issues during upgrade 2014-03-06 11:23:15 +01:00
Satya Bodapati
a559f6abf7 Bug#18124788 - COMPRESSED TABLES REGRESSION FROM 5.6.13
After incrementing the number of pending uncompression operations,
buffer pool zip mutex can be released immediately.

Approved by Vasil. rb#4514
2014-02-28 12:45:53 +05:30
Rich Prohaska
4e6d8249db support HTON_SUPPORTS_CLUSTERED_KEYS 2014-02-26 14:39:06 -05:00
Rich Prohaska
7c506b017b support -DTOKUDB_VERSION=SOME_VERSION_STRING 2014-02-26 10:58:44 -05:00
Sergei Golubchik
9e135ce19f Fix code in make_sortkey() that only worked by chance
(assert added by MySQL verified that strnxfrm can only *increase*
the string length if from == to, and the latter is a random
decision made by individual items and String::realloc).

Fix it by avoiding the memcpy in the first place.
2014-02-26 16:15:52 +01:00
Rich Prohaska
5243883431 fix a race condition in the i_s_tokudb_lock_waits_released test 2014-02-25 17:03:10 -05:00
Rich Prohaska
eb500b8cb1 destroy ule status partitioned counters 2014-02-25 11:21:13 -05:00
Rich Prohaska
21421e1f36 fix make problem related to tokudb only tests 2014-02-25 09:57:36 -05:00
Tor Didriksen
3f2b13f979 Bug#18235669 MYSQL_CONFIG TO PROVIDE R FLAG ON SOLARIS
'mysql_config --libs' outputs -L/path/to/library
on SunOS we also want it to output '-R/path/to/library'
in order to find libraries at runtime.


cmake/libutils.cmake:
  Add an informational message, to show dependencies on OS libraries.
2014-02-25 09:07:44 +01:00
sayantan dutta
6f027d9dfc Bug - MTR SHOULD NOT WAIT FOR CHILDREN WHEN ABORTING TEST RUN ON WINDOWS 2014-02-25 18:42:14 +05:30
Rich Prohaska
e9b9fd9077 keep track of and cleanup last_lock_timeout memory 2014-02-24 09:46:24 -05:00
Leif Walsh
c6a6cd8cf8 Merge remote-tracking branch 'origin/garbage_collection_extended' 2014-02-23 16:09:07 -05:00
John Esmet
dfcffcc714 Allow txn manager state to be created cheaply, with the option to later
initialize it for full gc if we think it is worth it. This allows for
full gc to run when bringing a basement node up to date.
2014-02-21 19:57:19 -05:00
John Esmet
9ee82c3cce Do not use the newest of our ancestor's oldest_referenced_xid values -
there may be an abort message waiting to be applied.
2014-02-21 16:10:51 -05:00
John Esmet
36a525f978 Fix test 2014-02-21 16:09:34 -05:00
John Esmet
be301362d8 Add test to detect a case of bad implicition promotion 2014-02-21 15:57:58 -05:00
Rich Prohaska
b77c0addba zero out new null bytes to force valgrind to quit complaining about uninitialized bytes 2014-02-21 15:53:05 -05:00
Rich Prohaska
054633c65d allow lock tree waits to be killable 2014-02-21 12:27:23 -05:00
Rich Prohaska
3fbd54a96d add tokudb_kill_time session variable to set lock tree kill callback frequency 2014-02-21 08:21:31 -05:00
Igor Babaev
3e03c9eae9 After constant row substitution the optimizer should call the method
update_used_tables for the the where condition to update cached
indicators of constant subexpressions. It should be done before further
possible simplification of the where condition.

This change caused simplification of the executed where conditions 
in many test cases.
2014-02-20 21:27:33 -08:00
John Esmet
0bd9ff9d3f Fix up a race where two threads could update the oldest referenced xid
in a different order than they began, and blow up a bad assert.
2014-02-20 16:15:14 -05:00
John Esmet
fb76250b7a Consolidate garbage collection information into txn_gc_info. Explicitly
separate the difference between the oldest referenced xid we can use for
implicit promotion vs simple garbage collection.
2014-02-20 13:32:25 -05:00
John Esmet
f12bdc758c Pass down txn manager state to message application, which it can use to
run full garbage collection when a leafentry has > 1 committed entry.
2014-02-20 13:30:30 -05:00
Rich Prohaska
f07613f140 update cardinality on first table open 2014-02-20 07:41:11 -05:00
Sergey Petrunya
6ebaa4939e Backport the following from 5.5 to 5.3:
MDEV-4556 Server crashes in SEL_ARG::rb_insert with index_merge+index_merge_sort_union, FORCE INDEX
- merge_same_index_scans() may put the same SEL_ARG tree in multiple result plans.
  make it call incr_refs() on the SEL_ARG trees that it does key_or() on, because 
  key_or(sel_arg_tree_1, sel_arg_tree_2) call may invalidate SEL_ARG trees pointed 
  by sel_arg_tree_1 and sel_arg_tree_2.
2014-02-19 18:34:12 +04:00
Sergey Petrunya
ddc21f7977 Fix compile failure:
In function ‘void* memset(void*, int, size_t)’,
    inlined from ‘void Lifo_buffer::set_buffer_space(uchar*, uchar*)’ at sql_lifo_buffer.h:70:5,
    inlined from ‘int DsMrr_impl::dsmrr_init(handler*, RANGE_SEQ_IF*, void*, uint, uint, HANDLER_BUFFER*)’ at multi_range_read.cc:895:62:
/usr/include/i386-linux-gnu/bits/string3.h:82:32: error: call to ‘__warn_memset_zero_len’ declared with attribute warning: memset used with constant zero length parameter; this could be due to transposed parameters [-Werror]

It was intentional that the buffer is set to zero length there.
2014-02-19 17:52:47 +04:00
Sergey Petrunya
260c802e7d Add a debugger helper function that does this:
(gdb) p dbug_print_table_row(table)
  $33 = "SUBQUERY2_t1(col_int_key,col_varchar_nokey)=(7,c)"
2014-02-19 17:47:02 +04:00
Sergey Petrunya
097b644006 MDEV-5600: Wrong result on 2nd execution of PS depending on the length of the query
- Item_direct_view_ref didn't clear its pointer to item_equal in ::cleanup.
- Some Item_direct_view_ref objects have statement lifetime (i.e. they
  survive across multiple EXECUTE commands). Item_equal objects live only for
  the duration of one EXECUTE. This caused Item_direct_view_ref to have a stale pointer,
  which could cause all sorts of effects. (In this bug's testcase it was pointing to
  the wrong Item_equal, causing wrong query result)
- Fixed by doing what Item_field::cleanup() does - don't keep item_equal pointer value.
- There is no testcase because the only testcase I've got is highly fragile (e.g. the
  bug will not show up if @@datadir is of the wrong length).
2014-02-19 17:45:33 +04:00
Alexey Botchkov
84580f950c MDEV-5481 mysqldump fails to dump geometry types properly.
Fixed so the MYSQL_TYPE_GEOMETRY is treated as BLOB.
2014-02-18 17:45:08 +04:00
Alexey Botchkov
820b1a6687 MDEV-5615 crash in Gcalc_function::add_operation.
The result is EMPTY for a buffer(line, -1), but we still
need one FALSE operation to be stored in the condition.
And we actually add it but forgot to alloc memory to store it.
2014-02-18 17:15:25 +04:00
Murthy Narkedimilli
075a2fb716 Updated/added copyright header. Added line "use is subject to license terms"
to copyright header.
2014-02-17 18:19:04 +05:30
Thirunarayanan B
9ea02a1c09 Bug UNIQUE PREFIX INDEX ON BINARY COLUMN:
FAILING ASSERTION: FLEN == LEN

Problem:
       Broken invariant triggered when building a unique index on a
binary column and the input data contains duplicate keys. This was broken
in debug builds only.

Fix:
       Fixed length of the binary datatype can be greater than length of
the shorter prefix on which index is being created.
2014-02-17 13:45:34 +05:30
John Esmet
d93fe66d5e refs split the includes that are used for OSX vs linux builds 2014-02-15 10:38:20 -05:00
John Esmet
ca7651e522 fixes maintain a global map of registered tl_stack objects that we
can iterate over and clean up, which prevents a memory leak when the
fractal tree shared library unloads before all thread-local destructors run
2014-02-15 10:13:23 -05:00
Bradley C. Kuszmaul
0a06abb42e Revert "Make the hotbackup library link up"
This reverts commit 843f0050cb.

Should have been on a branch.
2014-02-14 17:07:26 -05:00
Bradley C. Kuszmaul
d0223116ac Revert "Cleanup"
This reverts commit 7fa771e0f3.

Should have been on a branch.
2014-02-14 17:07:10 -05:00
Bradley C. Kuszmaul
6154c3d31b Revert "Cleanup"
This reverts commit fbd9685991.

Should have been on a branch.
2014-02-14 17:06:22 -05:00
Bradley C. Kuszmaul
d76887a9c2 Revert "backup stress test runs"
This reverts commit d467d3cf42.

Should have been on a branch.
2014-02-14 17:06:01 -05:00
Sergey Petrunya
ccb54beb6d Merge 2014-02-15 01:26:53 +04:00
Sergey Petrunya
8c9b2f3429 MDEV-5581: Server crashes in in JOIN::prepare on 2nd execution of PS with materialization+semijoin
- The problem was that JOIN::prepare() tried to set TABLE::maybe_null
  for a table in join. Non-merged semi-join tables 1) are present as 
  join's base tables on second EXECUTE, but 2) do not yet have a TABLE 
  object.
  Worked around the problem by putting mixed_implicit_grouping into JOIN
  object, and then passing it to JTBM tables in setup_jtbm_semi_joins().
2014-02-15 01:21:46 +04:00
Bradley C. Kuszmaul
7c6bd00edf backup stress test runs 2014-02-14 16:02:58 -05:00
Bradley C. Kuszmaul
00ef6d8498 Cleanup 2014-02-14 15:27:23 -05:00
Bradley C. Kuszmaul
91e2af2b42 Cleanup 2014-02-14 15:16:54 -05:00
Bradley C. Kuszmaul
d71b0003a5 Make the hotbackup library link up 2014-02-14 14:31:14 -05:00
Rich Prohaska
bfd4c05c40 new FT compression ratio status vars added, so fix the result file 2014-02-14 12:08:18 -05:00
Rich Prohaska
2e7e13af2d fix out of range read from uint3korr 2014-02-14 10:11:51 -05:00
Sergei Golubchik
5a21dc7df1 fix suite/sphinx/suite.pm to not start searchd twice 2014-02-14 14:08:16 +01:00
Rich Prohaska
567c09ad91 run atc and tpch table loads with a configurable memory size 2014-02-14 06:48:28 -05:00
Sergei Golubchik
9d5731ef1b fix the test 2014-02-14 11:14:10 +01:00
Sergei Golubchik
8a26d19088 record incorrect result for MDEV-5539 Empty results in UNION with Sphinx engine
(just to have the test in the tree when we merge the upstream fix)
2014-02-14 10:39:25 +01:00