1. 12 Oct, 2005 1 commit
    • ramil@mysql.com's avatar
      memcpy_overlap() removed, as · 58c64aea
      ramil@mysql.com authored
        1. it's wrong to use memcpy() for overlapped areas;
        2. we use it only once.
      During merge to 4.1 will remove a memcpy_overlap() call 
      from strings/ctype-tis620.c as well in order to fix 
      bug #10836: ctype_tis620 test failure with ICC-compiled binaries on IA64.
      58c64aea
  2. 11 Oct, 2005 2 commits
  3. 09 Oct, 2005 2 commits
  4. 03 Oct, 2005 1 commit
  5. 29 Sep, 2005 3 commits
  6. 27 Sep, 2005 1 commit
  7. 23 Sep, 2005 1 commit
  8. 19 Sep, 2005 1 commit
  9. 18 Sep, 2005 1 commit
  10. 15 Sep, 2005 1 commit
  11. 14 Sep, 2005 1 commit
  12. 13 Sep, 2005 1 commit
  13. 12 Sep, 2005 2 commits
  14. 08 Sep, 2005 3 commits
  15. 07 Sep, 2005 1 commit
  16. 06 Sep, 2005 2 commits
  17. 02 Sep, 2005 3 commits
  18. 01 Sep, 2005 2 commits
  19. 31 Aug, 2005 1 commit
  20. 30 Aug, 2005 2 commits
  21. 29 Aug, 2005 2 commits
  22. 28 Aug, 2005 4 commits
  23. 24 Aug, 2005 2 commits
    • jani@a193-229-222-105.elisa-laajakaista.fi's avatar
      fe977b0c
    • heikki@hundin.mysql.fi's avatar
      ha_innodb.cc: · 9f72d504
      heikki@hundin.mysql.fi authored
        Fix bug #12779 : never give a row count estimate of 0 to the MySQL query optimizer, as then left join optimizer may beleive it KNOWS that the table is empty; note that this fix may change query optimization of many other queries where one table is empty; note that the proper fix would be to make the query optimizer to know that the row count estimates it receives really are just estimates, it cannot assume they are certain
      9f72d504