If InnoDB crash recovery was needed, the InnoDB function srv_start()
would invoke extra validation, reading something from every InnoDB
data file. This should be unnecessary now that MDEV-14717 made
RENAME operations crash-safe inside InnoDB (which can be
disabled in MariaDB 10.2 by setting innodb_safe_truncate=OFF).
dict_check_sys_tables(): Skip tables that would be dropped by
row_mysql_drop_garbage_tables(). Perform extra validation only
if innodb_safe_truncate=OFF, innodb_force_recovery=0 and
crash recovery was needed.
dict_load_table_one(): Validate the root page of the table.
In this way, we can deny access to corrupted or mismatching tables
not only after crash recovery, but also after a clean shutdown.
In MariaDB, InnoDB tables will always contain DATA_N_SYS_COLS = 3
columns, 2 or 3 of which are present in the clustered index.
We remove the predicate that was added in MySQL 5.7 as part of WL#7682.
`zypper install krb5-devel` installs executables outside of $PATH.
It also installs /etc/profile.d/krb5.sh that is sourced by a new
shell to add the new location to the $PATH. But this doesn't affect
the current shell.
Now decent Linux distros remind the user to run `. /etc/profile`
to reload paths in such a case. SUSE doesn't and for a good reason -
it doesn't work there. Because SUSE sets PROFILEREAD=true in the
environment and /etc/profile does not do anything.
By this point, one should not really expect `unset PROFILEREAD` to help,
and it does not - PROFILEREAD is readonly, and cannot be unset.
Apparently SUSE really *really* wants you to re-login between installing
MariaDB build dependencies and actually running the rpmbuild.
Which we cannot do it buildbot. And it would look very user-un-friendly
in the Build Instructions section of the manual.
So, we work around it - by adding SUSE krb5 path to the search list.
THIS IS SUSEEEEEE!!!
On Microsoft Windows, InnoDB writes the path separator \ to the
redo log file, while on all other platforms, / is being used.
fil_name_parse(): Normalize the parsed path separators to the
native format. This allows backups or data sets to be portable
between Windows and other systems.
MariaDB before MDEV-5800 in version 10.2.2 did not support
indexed virtual columns. Non-persistent virtual columns were
hidden from storage engines. Only starting with MDEV-5800, InnoDB
would create internal metadata on virtual columns.
Similar to what was done in MDEV-18084, MDEV-18090, MDEV-18960, we adjust
one more code path for the old tables.
innobase_build_col_map(): Allocate space for virtual columns in col_map[]
but leave the entries at ULINT_UNDEFINED, noting that the virtual columns
were missing before the table was being rebuilt.
InnoDB includes 3 parsers, which use 3 lexical analyzers that
are generated with flex. Flex versions before 2.6 emitted
the keyword "register", which is deprecated in C++17.
The lexical analyzers were regenerated as follows:
for s in storage/innobase storage/xtradb
do
(cd "$s"/pars; ./make_flex.sh)
touch "$s"/fts/*.l
make -C "$s"/fts -f Makefile.query
done
The MDEV-11738/MDEV-11581 fix was supposed to add the column
ROTATING_OR_FLUSHING to the INFORMATION_SCHEMA table
INNODB_TABLESPACES_ENCRYPTION, but it also added that column to
INNODB_TABLESPACES_SCRUBBING in InnoDB (not XtraDB).
The extra column was never initialized. We will remove it,
because key rotation has nothing to do with the scrubbing of
tablespace data.
When HAVE_YASSL is defined (due to cmake -DWITH_SSL=bundled
or otherwise), mysys_ssl/my_crypt.cc will #include "yassl.cc"
from the same directory.
When MariaDB 10.2 or later is compiled with GCC 8 and optimizations
are enabled, then the check
if (iv)
in EVP_CipherInit_ex() can be wrongly optimized away.
The reason appears to be that __attribute__((nonnull)) is attached
to the variable iv, because there is a (no-op) call
memcpy(oiv, iv, ivlen=0) earlier in the code path.
It is possible that this started failing after the code was
refactored in MDEV-10332 (MariaDB 10.2.6). In MariaDB 10.1,
there is a similar memcpy() call in MyCTX_nopad::init(),
but the code appears to work fine.
The cause for this was fix MDEV-15372, which was trying to speed up
the parallel slave.
Part#1: Do not attempt the "optimization" for transactions that are not
replication slave workers.
Problem:
========
When attempting to delay a Slave attached with GTID, there appears to be an
extra delay applied initially. For example, this output reflects a Slave that is
already delayed by 43200 seconds. When switching to GTID replication,
replication is paused until SQL_Remaining_Delay counts down to 0:
CHANGE MASTER TO master_use_gtid=current_pos; CHANGE MASTER TO
MASTER_DELAY=43200;
Seconds_Behind_Master: 44847
Using_Gtid: Current_Pos
SQL_Delay: 43200
SQL_Remaining_Delay: 43089
Slave_SQL_Running_State: Waiting until MASTER_DELAY seconds after master
executed event
Analysis:
=========
When slave initiates a GTID based connection request to master, the master sends
two GTID_LIST events. The first one is actual GTID_LIST event and the second
one is a fake GTID_LIST event. This is sent by master to provide its current
binlary log file position. The fake GTID_LIST events will have their ev->when=0.
'when' (the timestamp) is set to 0 so that slave could distinguish between real
and fake Rotate events.
On slave side when MASTER_DELAY is configured to "X" the applier will ensure
that there is a time delay of "X" seconds before the event is applied.
General behaviour of MASTER_DELAY example:-
Master
timestamp of event e1=10
timestamp of event e2=11
On slave MASTER_DELAY=5
Event e1 will be applied at = 15
e2 will be applied at =16
In bug scenario:-
On Master: With GTIDs
timestamp of event e1=10
timestamp of event e2=0
On Slave:
e1 will be applied at = 10 + 5 =15
For e2, since "e2->when=0" e2->when is set to current timestamp.
i.e since the e2->when and current timestamp on slave is the same applier waits
for additional master_delay=5 seconds. the ev->when contributes to
"rli->last_master_timestamp".
rli->last_master_timestamp= ev->when + (time_t) ev->exec_time;
Fake events should not update the "ev->when" to "current timestamp" on slave.
Fix:
===
Remove the assignment of current timestamp to "ev->when" when "ev->when=0".
This reverts commit 21b2fada7a
and commit 81d71ee6b2.
The MDEV-18464 change introduces a few data race issues. Contrary to
the documentation, the field trx_t::victim is not always being protected
by lock_sys_t::mutex and trx_t::mutex. Most importantly, it seems
that KILL QUERY could wrongly avoid acquiring both mutexes when
invoking lock_trx_handle_wait_low(), in case another thread had
already set trx->victim=true.
We also revert MDEV-12009, because it should depend on the MDEV-18464
fix being present.
1) Avoid writing of MLOG_INDEX_LOAD redo log record during inplace
alter table when the table is empty and also for spatial index.
2) Avoid creation of temporary merge file for spatial index during
index creation process.
To fix the crash there we need to make sure that the
server while storing the statistical values in statistical tables should do it
in a multi-byte safe way.
Also there is no need to throw warnings if there is truncation while storing
values from statistical fields.
As noted on kill_one_thread SUPER should be able to kill even
system threads i.e. threads/query flagged as high priority or
wsrep applier thread. Normal user, should not able to kill
threads/query flagged as high priority (BF) or wsrep applier
thread.
Pushed the decision for innodb transaction and system
locking down to lock0lock.cc level. With this,
we can avoid releasing these mutexes for executions
where these mutexes were acquired upfront.
This patch will also fix BF aborting of native threads, e.g.
threads which have declared wsrep_on=OFF. Earlier, we have
used, for innodb trx locks, was_chosen_as_deadlock_victim
flag, for marking inodb transactions, which are victims for
wsrep BF abort. With native threads (wsrep_on==OFF), re-using
was_chosen_as_deadlock_victim flag may lead to inteference
with real deadlock, and to deal with this, the patch has added new
flag for marking wsrep BF aborts only: victim=true
Similar way if replication decides to abort one of the threads
we mark victim by: victim=true
innobase_kill_query
Remove lock sys and trx mutex handling.
wsrep_innobase_kill_one_trx
Mark victim trx with victim=true
trx0trx.h
Remove trx_abort_t type and abort type variable from
trx struct. Add victim variable to trx.
wsrep_kill_victim
Remove abort_type
lock_report_waiters_to_mysql
Take also trx mutex and mark trx as a victim for
replication abort.
lock_trx_handle_wait_low
New low level function to check whether the transaction
has already been rolled back because it was selected as
a deadlock victim, or if it has to wait then cancel
the wait lock.
lock_trx_handle_wait
If transaction is not marked as victim take lock sys
and trx mutex before calling lock_trx_handle_wait_low
and release them after that.
row_search_for_mysql
Remove lock sys and trx mutex taking and releasing.
trx_rollback_to_savepoint_for_mysql_low
trx_commit_in_memory
Clean up victim variable.
ignore FK-prelocked tables when looking for write-prelocked tables
with auto-increment to complain about "Statement is unsafe because
it invokes a trigger or a stored function that inserts into an
AUTO_INCREMENT column"
special cases:
* change systemd detection to use CHECK_LIBRARY_EXISTS at least once,
to have it detected by build_depends.cmake
* similarly, use find_library for pam
* unixODBC is weird, libodbc.so is in the unixODBC package, not
in the unixODBC-devel, where normally all .so files belong.
Packaging bug? As a workaround, use find_file(sql.h) instead of
find_path(sql.h) to make sure that /usr/include/sql.h (not /usr/include)
is cached by cmake, and later build_depends.cmake will select
unixODBC-devel, as a package owning /usr/include/sql.h file.
automatic BuildRequires for source RPM: for every FILEPATH and
"Have library XXX" cached variable, detect what rpm package it comes from
and add it to the list of dependencies.
That is, the source RPM will BuildRequire all those packages that
were found by cmake when the source RPM was built. Presumably, our
CMakeLists.txt won't check for libraries that aren't needed for a build.
It supports libraries/executables/files found with
FIND_LIBRARY
FIND_FILE
FIND_PROGRAM
CHECK_LIBRARY_EXISTS
create source RPM cpack-way
when building binary packages, this source rpm will use
same BUILD_CONFIG and WITH_SSL values that were used when
creating the source RPM.
Only do it for a reasonably new cmake, where
source rpms are known to work (3.10.2 is ok, 3.5.2 is not).
And force a shorter CPACK_RPM_BUILD_SOURCE_DIRS_PREFIX so that
a source rpm could be built from a standard location in /usr/src
now we can afford it. Fix -Werror errors. Note:
* old gcc is bad at detecting uninit variables, disable it.
* time_t is int or long, cast it for printf's
