maria-developers team mailing list archive

[Igor Babaev <igor@xxxxxxxxxxxx>]

On 06/22/2011 06:05 AM, Sergei Golubchik wrote:
> Hi.
> 
> Here I'll list changesets that cause conflicts and ask what to do.
> 
> 1. http://tinyurl.com/63tey39 vs. http://tinyurl.com/5v2pluf
> 
> it's a fix for bug#49322. both changes are similar, return is moved
> down. But Igor put it after the while() loop, while Ole John - in the
> loop, as soon a non-match is found. What difference does it make, I mean
> the earlier return, what consequences?

Sorry for the delay with the reply: it took me a significant amount of
time to figure out whether Ole's patch covered all cases. It does. His
patch saves on condition checks within unnecessary loops, mine on
possible shortcuts. It would be optimal to join them together like this:

<<
    while (join_tab->first_unmatched && found)
    {
      /*
        The while condition is always false if join_tab is not
        the last inner join table of an outer join operation.
      */
      JOIN_TAB *first_unmatched= join_tab->first_unmatched;
      /*
        Mark that a match for current outer table is found.
        This activates push down conditional predicates attached
        to the all inner tables of the outer join.
      */
      first_unmatched->found= 1;
      for (JOIN_TAB *tab= first_unmatched; tab <= join_tab; tab++)
      {
        /* Check all predicates that has just been activated. */
        /*
          Actually all predicates non-guarded by first_unmatched->found
          will be re-evaluated again. It could be fixed, but, probably,
          it's not worth doing now.
        */
        if (tab->select_cond && !tab->select_cond->val_int())
        {
          /* The condition attached to table tab is false */
          if (tab == join_tab)
          {
            found= 0;
            if (tab->table->reginfo.not_exists_optimize)
              return NESTED_LOOP_NO_MORE_ROWS;
          else
          {
            /*
              Set a return point if rejected predicate is attached
              not to the last table of the current nest level.
            */
            join->return_tab= tab;
            return (tab->table->reginfo.not_exists_optimize) ?
              NESTED_LOOP_NO_MORE_ROWS : NESTED_LOOP_OK;
          }
        }
      }
      /*
        Check whether join_tab is not the last inner table
        for another embedding outer join.
      */
      if ((first_unmatched= first_unmatched->first_upper) &&
          first_unmatched->last_inner != join_tab)
        first_unmatched= 0;
      join_tab->first_unmatched= first_unmatched;
    }

>>

Yet I think it should be done with a separate commit.

That's why I would advise for 1:

Take my complete patch (as it was pushed into maria-5.1 and later merged
into 5.3) and add Ole's test cases.

> 
> 2. http://tinyurl.com/69lwrp9 - a big chunk is moved around in
> test_if_skip_sort_order()
> 
> that's all, the rest of the file I've merged already.
> 
> Regards,
> Sergei

As to 2, ask Monty because he changed a lot of code in this procedure
when merging with the latest 5.1.

Regards,
Igor.