The RDTSC instruction, which was introduced in the Intel Pentium,
has been used in MariaDB for a long time. But, the __rdtsc()
wrapper is not available by default in some x86 build environments.
The simplest solution seems to replace the inlined instruction
with a call to the wrapper function my_timer_cycles(). The overhead
for the call should not affect the measurement threshold.
On Windows and on AMD64, we will keep using __rdtsc() directly.
Starting with the Intel Skylake microarchitecture, the PAUSE
instruction latency is about 140 clock cycles instead of earlier 10.
On AMD processors, the latency could be 10 or 50 clock cycles,
depending on microarchitecture.
Because of this big range of latency, let us scale the loops around
the PAUSE instruction based on timing results at server startup.
my_cpu_relax_multiplier: New variable: How many times to invoke PAUSE
in a loop. Only defined for IA-32 and AMD64.
my_cpu_init(): Determine with RDTSC the time to run 16 PAUSE instructions
in two unrolled loops according, and based on the quicker of the two
runs, initialize my_cpu_relax_multiplier. This form of calibration was
suggested by Mikhail Sinyavin from Intel.
LF_BACKOFF(), ut_delay(): Use my_cpu_relax_multiplier when available.
ut_delay(): Define inline in my_cpu.h.
UT_COMPILER_BARRIER(): Remove. This does not seem to have any effect,
because in our ut_delay() implementation, no computations are being
performed inside the loop. The purpose of UT_COMPILER_BARRIER() was to
prohibit the compiler from reordering computations. It was not
emitting any code.
MDEV-19585 Assertion with S3 table and flush_tables
The limit has to be increased so that MariaDB can create system tables.
It should not have any notable impact on performance.
There should not be any notable performance differences between 1K and 4K,
especially for temporary tables. In most cases using bigger blocks is also
faster (with the possible exception of doing key reads of not fixed length
keys).
The problem was two fault:
- flush_tables() wrongly gave errors when failing to open read only tables
- backup_block_ddl() didn't properly ignores errors from flush_tables()
The test case for this will be pushed in 10.5 as the test involves
S3 tables.
dict_table_t::stats_latch_created: remove along with related stuff
dict_table_t::stats_latch: make value member, not pointer. And always lock this
for simplicity, even to stats cloned table.
based on the work of Sergey Vojtovich
The bug occured when the optimizer decided to use a rowid filter built
by a range index scan to access an InnoDB table with generated clustered
index.
When a table is accessed by a secondary index Idx employing a rowid filter the
the value of pk contained in the found index tuple is checked against the
filter. A call of the handler function position is supposed to put the
pk value into the handler::ref buffer. However for generated clustered
primary keys it did not happened. The patch fixes this problem.
The problem was that the code in maria_extra assumed that there could be
only one table open when doing maria_extra(MA_FORCE_REOPEN)
However in the case of triggers, there can be multiple copies of
the table open.
Fixed by removing assert.
Also, move part of the test back to innodb.innodb_mysql
and another part to a new test innodb.purge.
Last but not least, merge the tests innodb_zip.4k and innodb_zip.8k
to innodb_zip.page_size.
Make Field::is_equal() const and return bool as it's a naturally fitting
type for it. Also it's agrument was narrowed to Column_definition.
InnoDB can change type of some columns by itself. InnoDB-specific code used to
reside in Field_xxx:is_equal() methods. Now engine-specific stuff was
moved to a virtual methods of handler::can_convert{string,varstring,blob,geom}.
These methods are called by Field::can_be_converted_by_engine() which is a
double dispatch pattern.
Some InnoDB-specific code still resides in compare_keys_but_name(). It should
be moved from here someday to handler::compare_key_parts(...) or similar.
IS_EQUAL_WITH_REINTERPRET_COMPATIBLE_CHARSET
IS_EQUAL_WITH_REINTERPRET_COMPATIBLE_CHARSET_BUT_COLLATE: both was removed
IS_EQUAL_NO, IS_EQUAL_YES are not needed now and should be removed
along with deprecated handler::check_if_incompatible_data().
HA_EXTENDED_TYPES_CONVERSION: was removed as such logic is not needed now by
server code.
ALTER_COLUMN_EQUAL_PACK_LENGTH: was renamed to a more generic
ALTER_COLUMN_TYPE_CHANGE_BY_ENGINE
Patch is about two cases:
1) On some collate changes it's possible to rebuild only secondary indexes
2) For non-indexed columns collate can be changed INSTANTly
Implemented mostly in Field_{string,varstring,blob}::is_equal().
Make this method return how exactly collationa differs.
This information is later used by fill_alter_inplace_info() to pass
correct info to engine.
MariaDB doesn't support Read-Free replication, so showing them only causes
confusion.
Removed variables:
- @@rocksdb_read_free_rpl
- @@rocksdb_read_free_rpl_tables
The test cases for the MDEV found several independent bugs
in MariaDB server and Aria:
- If a temporary table was marked as crashed, it could never
be deleted.
- Opening of a crashed temporary table gave an error message
but the error was never forwarded to the caller which caused
an assert() in my_ok()
- init_read_record() did mmap of all temporary tables, which is
probably not a good idea as this area can potentially be
very big. Changed code to only mmap internal temporary tables.
- mmap-ed tables where not unmapped in case of repair/optimize
which caused bad data in table and crashes if the original
table files where replaced with new ones (as the old mmap
was still in place). Fixed by removing the mmap in case
of repair.
- Cleaned up usage of code that disabled mmap in Aria
Problem was that in case of implicit rollback for alter table
Aria did try to run commit twice.
The test case for this is tricky to do in 10.2, so it will
be added to 10.4 as part of BACKUP STAGE testing.
Empty write sets will not trigger apply callback, and will not
update thread wsrep_trx_meta.gtid.seqno. Because of that assert will
be triggered when commit callback is called.
In collaboration with Sergey Vojtovich <svoj@mariadb.org>
The COMPRESSED clause is now a part of the data type and goes immediately
after the data type and length, but before the CHARACTER SET clause,
and before column attributes such as DEFAULT, COLLATE, ON UPDATE,
SYSTEM VERSIONING, engine specific column attributes.
In the old reduction, the COMPRESSED clause was a column attribute.
New syntax:
<varchar or text data type> <length> <compression> <character set> <column attributes>
<varbinary or blob data type> <length> <compression> <column attributes>
New syntax examples:
VARCHAR(1000) COMPRESSED CHARACTER SET latin1 DEFAULT ''
BLOB COMPRESSED DEFAULT ''
Deprecate syntax examples:
VARCHAR(1000) CHARACTER SET latin1 COMPRESSED DEFAULT ''
TEXT CHARACTER SET latin1 DEFAULT '' COMPRESSED
VARBINARY(1000) DEFAULT '' COMPRESSED
As a side effect:
- COMPRESSED is not valid as an SP label name in SQL/PSM routines any more
(but it's still valid as an SP label name in sql_mode=ORACLE)
- COMPRESSED is now allowed in combination with GENERATED ALWAYS AS:
TEXT COMPRESSED GENERATED ALWAYS AS REPEAT('a',1000)
