Home - Waterfall Grid T-Grid Console Builders Recent Builds Buildslaves Changesources - JSON API - About

Console View


Categories: connectors experimental galera main
Legend:   Passed Failed Warnings Failed Again Running Exception Offline No data

connectors experimental galera main
Dmitry Shulga
MDEV-30645: CREATE TRIGGER FOR { STARTUP | SHUTDOWN }

Follow-up to fix issue with starting server on server with broken table mysql.event
Vladislav Vaintroub
MDEV-37788 fix uintNkorr/intNstore macros to avoid unaligned access

Replaced macros with optimized inline functions, that use memcpy
instead of unaligned access (translates to single instruction
on architectures we support). Used compiler byteswap intrinsics
where appropriate.

Removed byte_order_generic.h as well as Intel specific headers
byte_order_generic_{x86|x86_64}.h

Added some casts to places where intNstore writes into "const char*".
Daniel Black
wsrep_info plugin config file created
Mohammad Tafzeel Shams
MDEV-22186: Add innodb_trim_core_file to exclude InnoDB memory from cores

Problem:
There is no control to include or exclude large InnoDB memory regions
such as the buffer pool and log buffer from core files, which can lead to
unnecessarily large core dumps or prevent capturing useful memory state.

Solution:
Introduce a new global dynamic system variable innodb_trim_core_file that
allows excluding large InnoDB memory regions (buffer pool and log buffer)
from core dumps using MADV_DONTDUMP where supported. When enabled and
@@core_file=ON, the server marks relevant memory regions to be omitted
from core files. The madvise state is dynamically updated when the
variable changes.

Additionally, a test helper script is introduced to inspect
/proc/<pid>/smaps and verify whether large memory regions are marked
for trimming.

- innodb_trim_core_file
  Global boolean system variable controlling whether InnoDB memory
  regions should be excluded from core dumps. Has no effect if
  @@core_file=OFF. Default: ON on non-debug builds with MADV_DONTDUMP
  support, OFF otherwise.

- innobase_should_trim_core_dump()
  Determines whether InnoDB memory allocations should be marked with
  MADV_DONTDUMP, based on @@core_file and @@innodb_trim_core_file.

- innodb_srv_trim_core_file_update()
  Update hook that sets srv_trim_core_file and triggers runtime updates
  of madvise state for both the buffer pool and log buffers.

- srv_trim_core_file
  Global variable backing @@innodb_trim_core_file.

- buf_pool_t::buf_pool_should_madvise_dont_dump
  Tracks the current madvise state for the buffer pool memory.

- buf_pool_t::madvise_update_dump()
  Re-evaluates whether the buffer pool should be excluded from core
  dumps and applies ut_dontdump() or ut_dodump() accordingly.

- buf_pool_t::create()
  Initializes madvise state based on innobase_should_trim_core_dump()
  and conditionally applies MADV_DONTDUMP to the buffer pool memory.

- buf_pool_t::resize()
  Reapplies MADV_DONTDUMP to newly allocated memory regions when
  resizing if innodb_trim_core_file=ON

- log_t::log_buf_should_madvise_dont_dump
  Tracks the madvise state for the InnoDB log buffers.

- log_t::madvise_dontdump()
  Marks log buffers and resize buffers with MADV_DONTDUMP.

- log_t::madvise_dodump()
  Restores inclusion of log buffers in core dumps using MADV_DODUMP.

- log_t::madvise_update_dump()
  Updates the madvise state for log buffers when
  @@innodb_trim_core_file changes.

- log_t::attach()
  Call ut_dodump() for log_t::buf & log_t::flush_buf on successful
  allocation and innodb_trim_core_file=OFF

- log_t::resize_start()
  Call ut_dodump() for log_t::resize_buf & log_t::resize_flush_buf on
  successful allocation and innodb_trim_core_file=OFF

- check_core_dump_trim.inc
  Test helper that inspects /proc/<pid>/smaps to detect memory regions
  with the dd VmFlag and determine whether the core file is expected
  to be trimmed based on buffer pool and log buffer sizes.
Vladislav Vaintroub
Revert "MDEV-38989 main.ctype_utf16le SEGV in Ubuntu 26.04 (x86-64-v3)"

This reverts commit 6318d5001597fff28ea7ba7f957d66edd7e3e954.

uint2korr is now fixed in MDEV-37788, thus partial workarounds are not necessary anymore.
Sergei Petrunia
Cleanup: remove read-side structures that have write-side copies

Remove:
-class range_context_for_replay;
-class irc_context_for_replay;
-class rir_context_for_replay;
Mohammad Tafzeel Shams
MDEV-22186: Add innodb_trim_core_file to exclude InnoDB memory from cores

Problem:
There is no control to include or exclude large InnoDB memory regions
such as the buffer pool and log buffer from core files, which can lead to
unnecessarily large core dumps or prevent capturing useful memory state.

Solution:
Introduce a new global dynamic system variable innodb_trim_core_file that
allows excluding large InnoDB memory regions (buffer pool and log buffer)
from core dumps using MADV_DONTDUMP where supported. When enabled and
@@core_file=ON, the server marks relevant memory regions to be omitted
from core files. The madvise state is dynamically updated when the
variable changes.

Additionally, a test helper script is introduced to inspect
/proc/<pid>/smaps and verify whether large memory regions are marked
for trimming.

- innodb_trim_core_file
  Global boolean system variable controlling whether InnoDB memory
  regions should be excluded from core dumps. Has no effect if
  @@core_file=OFF. Default: ON on non-debug builds with MADV_DONTDUMP
  support, OFF otherwise.

- innobase_should_trim_core_dump()
  Determines whether InnoDB memory allocations should be marked with
  MADV_DONTDUMP, based on @@core_file and @@innodb_trim_core_file.

- innodb_srv_trim_core_file_update()
  Update hook that sets srv_trim_core_file and triggers runtime updates
  of madvise state for both the buffer pool and log buffers.

- srv_trim_core_file
  Global variable backing @@innodb_trim_core_file.

- buf_pool_t::buf_pool_should_madvise_dont_dump
  Tracks the current madvise state for the buffer pool memory.

- buf_pool_t::madvise_update_dump()
  Re-evaluates whether the buffer pool should be excluded from core
  dumps and applies ut_dontdump() or ut_dodump() accordingly.

- buf_pool_t::create()
  Initializes madvise state based on innobase_should_trim_core_dump()
  and conditionally applies MADV_DONTDUMP to the buffer pool memory.

- buf_pool_t::resize()
  Reapplies MADV_DONTDUMP to newly allocated memory regions when
  resizing if innodb_trim_core_file=ON

- log_t::log_buf_should_madvise_dont_dump
  Tracks the madvise state for the InnoDB log buffers.

- log_t::madvise_dontdump()
  Marks log buffers and resize buffers with MADV_DONTDUMP.

- log_t::madvise_dodump()
  Restores inclusion of log buffers in core dumps using MADV_DODUMP.

- log_t::madvise_update_dump()
  Updates the madvise state for log buffers when
  @@innodb_trim_core_file changes.

- log_t::attach()
  Call ut_dodump() for log_t::buf & log_t::flush_buf on successful
  allocation and innodb_trim_core_file=OFF

- log_t::resize_start()
  Call ut_dodump() for log_t::resize_buf & log_t::resize_flush_buf on
  successful allocation and innodb_trim_core_file=OFF

- check_core_dump_trim.inc
  Test helper that inspects /proc/<pid>/smaps to detect memory regions
  with the dd VmFlag and determine whether the core file is expected
  to be trimmed based on buffer pool and log buffer sizes.
Sergei Petrunia
Small cleanups. Remove table_context_for_replay::ddl
Alexander Barkov
MDEV-38933 Implement basic low level functions for DBMS_SQL

Adding SQL functions:
- DBMS_SQL_COMMAND(ps_name)
- DBMS_SQL_BIND_PARAM_BY_NAME(ps_name,pos,value)
- DBMS_SQL_COLUMN_VALUE(c,pos,dst)
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions

Optimize big endian (and some little endian) Innodb access functions by
using my_htobeN/my_betohN and memcpy.
Vladislav Vaintroub
MDEV-37788 review fixes - optimize uint3korr, sint3korr, and intNstore

uint3korr - Reuse uint2korr, to remove one MOV instruction
(of previously 3) on current compilers.

sint3korr - Reuse uint3korr, and replace the conditional sign-extension
with a branchless shift-based sign extension

int3store, int5store, int6store - reuse int2store and int4store.
Vladislav Vaintroub
Revert "MDEV-38989 main.ctype_utf16le SEGV in Ubuntu 26.04 (x86-64-v3)"

This reverts commit 6318d5001597fff28ea7ba7f957d66edd7e3e954.

uint2korr is now fixed in MDEV-37788, thus partial workarounds are not necessary anymore.
Daniel Black
deb: no wsrep_on until cluster address set
Marko Mäkelä
fixup! 1d57f26b50436d3e59ccce9a2b8cd7a9836c07ac

Skip log_checkpoint_low() if innodb_log_recovery_target is set
Vladislav Vaintroub
MDEV-37788 fix uintNkorr/intNstore macros to avoid unaligned access

Replaced macros with optimized inline functions, that use memcpy
instead of unaligned access (translates to single instruction
on architectures we support). Used compiler byteswap intrinsics
where appropriate.

Removed byte_order_generic.h as well as Intel specific headers
byte_order_generic_{x86|x86_64}.h

Added some casts to places where intNstore writes into "const char*".
Yuchen Pei
MDEV-38732 Correctly construct queries with NULL uservar values in spider

The subsequent assertion failure in end_statement is likely caused by
HA_ERR_OUT_OF_MEM not correctly handled, but it should not return an
HA_ERR_OUT_OF_MEM here in the first place.
Sergei Petrunia
Fix for sub-statements: handle PROCEDURE also.
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions

Optimize big endian (and some little endian) Innodb access functions by
using my_htobeN/my_betohN and memcpy.
Alexey Botchkov
MDEV-37262 XMLTYPE: validation.
  13.03
Marko Mäkelä
fixup! 32b7005d3463f95a03bf1dd9e7be4b59f3f9f591
bsrikanth-mariadb
MDEV-38803: Rename 2 files

change opt_store_replay_context.h and .cc files to
opt_context_store_replay.h and .cc respectively
Marko Mäkelä
fixup! 1d57f26b50436d3e59ccce9a2b8cd7a9836c07ac

log_t::write_checkpoint(): Properly resume set_archive().
Sergei Petrunia
Further code cleanups. Don't print empty objects for VIEWs.
forkfun
MDEV-38967: STR_TO_DATE() cannot be used in indexed virtual columns

Update STR_TO_DATE() to accept an optional third argument for the locale.
DATE_FORMAT() and STR_TO_DATE() return NULL when the locale argument is NULL;
issue a warning and fall back to the session @@lc_time_names for invalid locale names.
STR_TO_DATE() can now be used in indexed virtual columns while having all 3 arguments,
but not allowed when 2 as it depends then on the session @@lc_time_names
Marko Mäkelä
fixup! 8ee3ddfb46654f2dc49f5b3f141a83e85b833a11

innodb_log_recovery_target fixes
bsrikanth-mariadb
MDEV-38805: use charset utf8mb4, for client-server communication

REPLACE INTO statements would now use my_charset_utf8mb4_bin conversion
Also, the sql_script file structure would now look like

SET NAMES utf8mb4;
SET var1=value1;
.
.
CREATE DATABASE 'name' if NOT..;
USE DATABASE 'name';
CREATE TABLE ...;
.
REPLACE INTO ..;
.
.
set @opt_context='...';
.
.

Additionally, made few changes to the documentation comments
Sergei Petrunia
Make the context to be recorded even when optimizer trace is off.
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions

Optimize big endian (and some little endian) Innodb access functions by
using my_htobeN/my_betohN and memcpy.
Sergei Petrunia
Fix optimizer context recording and sub-statements:

Sub-statements don't start or stop recording.
Marko Mäkelä
fixup! 1d57f26b50436d3e59ccce9a2b8cd7a9836c07ac

Allow the last checkpoint slot to be overwritten after recovery
Brian White
fix build with WITH_EXTRA_CHARSETS=none again
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions

Optimize big endian (and some little endian) Innodb access functions by
using my_htobeN/my_betohN and memcpy.
Daniel Black
MDEV-26896 Enable -Wconversion globally
Jan Lindström
MDEV-38895 : Regression on MDL conflict handling

The reason for this timeout is a regression introduced in commit
https://github.com/MariaDB/server/commit/e40277d29b7c531e1ed6b3bed7ecfc8cfeff4c7e.
Where there is incorrect condition wsrep_thd_is_BF(granted_thd, false)
because if granted thread is BF it may not be BF-killed.

In this patch following has been changed
    * Split wsrep_handle_mdl_conflict to smaller easier to understand parts
    * Improved code comments
    * Improved debug logging to contain thread ids and necessary information
    * Corrected and simplified conditions so that BF threads are not killed
    * Added few debug assertions to verify that BF threads are not killed
    * Code cleanups so that tab-characters are not used
    * Existing test cases already test rest of the cases
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions (MSVC, GCC)

Optimize big endian (and some little endian) Innodb access functions
Vladislav Vaintroub
MDEV-39027 suboptimal code for InnoDB big endian access functions

Optimize big endian (and some little endian) Innodb access functions by
using my_htobeN/my_betohN and memcpy.
Sergei Petrunia
Fix msan/valgrind failure (actually harmless)