1. 10 Mar, 2020 3 commits
  2. 09 Mar, 2020 1 commit
  3. 04 Mar, 2020 2 commits
  4. 03 Mar, 2020 1 commit
  5. 27 Feb, 2020 2 commits
  6. 26 Feb, 2020 3 commits
  7. 20 Feb, 2020 2 commits
  8. 09 Feb, 2020 1 commit
  9. 02 Feb, 2020 1 commit
  10. 30 Jan, 2020 1 commit
  11. 27 Jan, 2020 1 commit
    • Sachin's avatar
      LAG free alter · 6da5b791
      Sachin authored
      BINLOG_SPLIT_ALTER added
      now master and slave has same sequence
      6da5b791
  12. 08 Nov, 2019 11 commits
    • Marko Mäkelä's avatar
      Cleanup: Replace xdes_set_bit() · bce71a29
      Marko Mäkelä authored
      The XDES_CLEAN_BIT is always set for every element of
      the page allocation bitmap in the extent descriptor pages.
      Do not bother touching it, to avoid redundant writes.
      
      xdes_set_free<bool free>(): Replaces xdes_set_bit().
      bce71a29
    • Marko Mäkelä's avatar
      Cleanup: Remove redundant XDES_FREE_BIT parameters · 5ed54e78
      Marko Mäkelä authored
      The page allocation bitmaps in the extent descriptor pages
      contain two bits per page: XDES_FREE_BIT and XDES_CLEAN_BIT,
      which is unused. Simplify read access.
      
      xdes_is_free(descr,mtr): Remove. Use !xdes_get_n_used(descr) instead.
      
      xdes_is_free(): Replaces xdes_get_bit(), xdes_mtr_get_bit().
      
      xdes_find_free(): Replaces xdes_find_bit().
      
      fsp_seg_inode_page_get_nth_inode(): Remove the redundant parameters
      physical_size, mtr.
      
      fsp_seg_inode_page_find_used(), fsp_seg_inode_page_find_free():
      Remove the redundant parameter mtr.
      5ed54e78
    • Marko Mäkelä's avatar
      Merge 10.4 into 10.5 · 74b7d018
      Marko Mäkelä authored
      74b7d018
    • Marko Mäkelä's avatar
      Use uint16_t for FIL_PAGE_TYPE · b5ef7ffa
      Marko Mäkelä authored
      Since commit 5d596064
      fil_page_type_is_index() expects uint16_t, not ulint.
      b5ef7ffa
    • Marko Mäkelä's avatar
      15b713ca
    • Marko Mäkelä's avatar
      MDEV-12353 preparation: Remove redundant writes · a6d614fb
      Marko Mäkelä authored
      fsp_alloc_seg_inode_page(): Ever since
      commit 3926673c
      all newly allocated pages are zero-initialized.
      Assert that this is the case for the FSEG_ID fields.
      (Side note: before that fix, other parts of the pages
      could contain nonzero garbage.)
      
      btr_store_big_rec_extern_fields(): Remove the redundant initialization
      of the most significant 32 bits of BTR_EXTERN_LEN. InnoDB never supported
      BLOBs that are longer than 4GiB. In fact, dtuple_convert_big_rec()
      would write emit an error message if a clustered index record tuple would
      exceed 1,000,000,000 bytes in length.
      a6d614fb
    • Marko Mäkelä's avatar
      Cleanup: Do not pass mtr_t* as NULL · 5b72e813
      Marko Mäkelä authored
      PageConverter::update_index_page(): Invoke lower-level writes directly.
      5b72e813
    • Marko Mäkelä's avatar
      5d596064
    • Marko Mäkelä's avatar
      Merge 10.4 into 10.5 · 52246dff
      Marko Mäkelä authored
      52246dff
    • Marko Mäkelä's avatar
      Cleanup: Remove mach_read_ulint() · 78d0d2cd
      Marko Mäkelä authored
      The function mach_read_ulint() is a wrapper for the lower-level
      functions mach_read_from_1(), mach_read_from_2(), mach_read_from_8().
      Invoke those functions directly, for better readability of the code.
      
      mtr_t::read_ulint(), mtr_read_ulint(): Remove. Yes, we will lose the
      ability to assert that the read is covered by the mini-transaction.
      We would still check that on writes, and any writes that
      wrongly bypass mini-transaction logging would likely be caught by
      stress testing with Mariabackup.
      78d0d2cd
    • Marko Mäkelä's avatar
      MDEV-17138 follow-up: Use MLOG_MEMSET for writing FIL_NULL · 8a5eb414
      Marko Mäkelä authored
      Always use the MLOG_MEMSET record for writing FIL_NULL,
      because it is more compact.
      8a5eb414
  13. 07 Nov, 2019 5 commits
    • Aleksey Midenkov's avatar
      db565439
    • Aleksey Midenkov's avatar
      MDEV-17553 Enable setting start datetime for interval partitioned history of... · 1e73d7d6
      Aleksey Midenkov authored
      MDEV-17553 Enable setting start datetime for interval partitioned history of system versioned tables
      
      * Explicit STARTS syntax
      * SHOW CREATE
      * Default STARTS rounding depending on INTERVAL type
      * Warn when STARTS timestamp is later than query time
      * Fix uninitialized Lex->create_last_non_select_table under
        mysql_unpack_partition()
      
      Default STARTS rounding depending on INTERVAL type
      
      If STARTS clause is omitted, default one is assigned with value
      derived from query timestamp. The rounding is done on STARTS value
      depending on INTERVAL type:
      
      SECOND: no rounding is done;
      MINUTE: timestamp seconds is set to 0;
      HOUR: timestamp seconds and minutes are set to 0;
      DAY, WEEK, MONTH and YEAR: timestamp seconds, minutes and hours are
      set to 0 (the date of rotation is kept as current date).
      1e73d7d6
    • Marko Mäkelä's avatar
      Merge 10.4 into 10.5 · 77e8a311
      Marko Mäkelä authored
      A conflict between MDEV-19514 (b42294bc)
      and MDEV-20934 (d7a24017)
      was resolved. We will not invoke the function ibuf_delete_recs()
      from ibuf_merge_or_delete_for_page(). Instead, we will add that
      logic to the function ibuf_read_merge_pages().
      77e8a311
    • Oleksandr Byelkin's avatar
      Merge 10.4 into 10.5 · 3ad37ed0
      Oleksandr Byelkin authored
      3ad37ed0
    • Marko Mäkelä's avatar
      Merge 10.3 into 10.4 · c24ec3ce
      Marko Mäkelä authored
      c24ec3ce
  14. 06 Nov, 2019 6 commits
    • Marko Mäkelä's avatar
      MDEV-20934: Make the test more robust · 352e7667
      Marko Mäkelä authored
      Due to MDEV-12288, the slow shutdown in MariaDB 10.3 will include
      resetting the DB_TRX_ID for all inserted records. This might
      cause the 60-second shutdown_server timeout to be exceeded.
      Let us wait for the purge to complete before initiating slow shutdown.
      352e7667
    • Marko Mäkelä's avatar
      Merge 10.3 into 10.4 · 928abd69
      Marko Mäkelä authored
      928abd69
    • Marko Mäkelä's avatar
      Merge 10.2 into 10.3 · 908ca466
      Marko Mäkelä authored
      908ca466
    • Marko Mäkelä's avatar
      Follow-up to 792c9f9a · 90451a59
      Marko Mäkelä authored
      dict_index_add_to_cache(): Make the 'index' a reference to a pointer,
      so that the caller will avoid the expensive call to
      dict_index_get_if_in_cache_low().
      90451a59
    • Marko Mäkelä's avatar
      Merge 10.1 to 10.2 · 8688ef22
      Marko Mäkelä authored
      8688ef22
    • Marko Mäkelä's avatar
      MDEV-20934 Infinite loop on innodb_fast_shutdown=0 with inconsistent change buffer · d7a24017
      Marko Mäkelä authored
      Due to a data corruption bug that may have occurred a long time earlier
      (possibly involving physical backup and MySQL Bug #69122, which was
      addressed in commit f166ec71)
      it seems possible that the InnoDB change buffer might end up containing
      entries, while no buffered changes exist according to the change buffer
      bitmap pages in the .ibd files.
      
      ibuf_delete_recs(): New function, to be invoked on slow shutdown only.
      Remove all buffered changes for a specific page.
      
      ibuf_merge_or_delete_for_page(): If the change buffer bitmap is clean
      and a slow shutdown is in progress, invoke ibuf_delete_recs().
      We do not want to do that during normal operation, due to the additional
      overhead that is involved. The bitmap page should be consistent with
      the change buffer in the first place.
      d7a24017