June 2010 - developers - lists.mariadb.org

[Maria-developers] Rev 2792: Fixed memory management problem (Item can't contain other Item due to way of Items destruction). in file:///home/bell/maria/bzr/work-maria-5.3-scache2/
by sanja＠askmonty.org 08 Jun '10

08 Jun '10

At file:///home/bell/maria/bzr/work-maria-5.3-scache2/ ------------------------------------------------------------ revno: 2792 revision-id: sanja(a)askmonty.org-20100608093610-efg156vu4mi4u9pp parent: sanja(a)askmonty.org-20100608074734-1m60ib2tac7y9m33 committer: sanja(a)askmonty.org branch nick: work-maria-5.3-scache2 timestamp: Tue 2010-06-08 12:36:10 +0300 message: Fixed memory management problem (Item can't contain other Item due to way of Items destruction). === modified file 'sql/item_cmpfunc.cc' --- a/sql/item_cmpfunc.cc 2010-05-31 21:25:54 +0000 +++ b/sql/item_cmpfunc.cc 2010-06-08 09:36:10 +0000 @@ -1737,13 +1737,16 @@ not_null_tables_cache|= args[1]->not_null_tables(); const_item_cache&= args[1]->const_item(); DBUG_ASSERT(scache == NULL); + DBUG_ASSERT(value_for_scache == NULL); if (args[0]->cols() ==1 && thd->variables.optimizer_switch & OPTIMIZER_SWITCH_SUBQUERY_CACHE && !(sub->engine->uncacheable() & (UNCACHEABLE_RAND | UNCACHEABLE_SIDEEFFECT))) { sub->depends_on.push_front((Item**)&cache); - scache= new Subquery_cache_tmptable(thd, sub->depends_on, &result); + value_for_scache= new Item_bool_cache; + scache= new Subquery_cache_tmptable(thd, sub->depends_on, + value_for_scache); } fixed= 1; return FALSE; @@ -1851,8 +1854,8 @@ /* put result in the cache */ if (scache) { - result.set(tmp, null_value); - scache->put_value(&result); + value_for_scache->set(tmp, null_value); + scache->put_value(value_for_scache); } DBUG_RETURN(tmp); } @@ -1876,6 +1879,7 @@ delete scache; scache= 0; } + value_for_scache= 0; DBUG_VOID_RETURN; } === modified file 'sql/item_cmpfunc.h' --- a/sql/item_cmpfunc.h 2010-05-31 21:25:54 +0000 +++ b/sql/item_cmpfunc.h 2010-06-08 09:36:10 +0000 @@ -241,7 +241,7 @@ /* Subquery cache */ Subquery_cache *scache; /* result representation for the subquery cache */ - Item_bool_cache result; + Item_bool_cache *value_for_scache; bool save_cache; /* Stores the value of "NULL IN (SELECT ...)" for uncorrelated subqueries: @@ -252,7 +252,8 @@ my_bool result_for_null_param; public: Item_in_optimizer(Item *a, Item_in_subselect *b): - Item_bool_func(a, my_reinterpret_cast(Item *)(b)), cache(0), scache(NULL), + Item_bool_func(a, my_reinterpret_cast(Item *)(b)), cache(0), + scache(NULL), value_for_scache(NULL), save_cache(0), result_for_null_param(UNKNOWN) {} bool fix_fields(THD *, Item **); === modified file 'sql/item_subselect.cc' --- a/sql/item_subselect.cc 2010-05-31 21:25:54 +0000 +++ b/sql/item_subselect.cc 2010-06-08 09:36:10 +0000 @@ -34,10 +34,10 @@ Item_subselect::Item_subselect(): Item_result_field(), value_assigned(0), thd(0), substitution(0), - engine(0), old_engine(0), scache(0), used_tables_cache(0), - have_to_be_excluded(0), const_item_cache(1), inside_first_fix_fields(0), - done_first_fix_fields(FALSE), eliminated(FALSE), engine_changed(0), - changed(0), is_correlated(FALSE) + engine(0), old_engine(0), scache(0), value_for_scache(0), + used_tables_cache(0), have_to_be_excluded(0), const_item_cache(1), + inside_first_fix_fields(0), done_first_fix_fields(FALSE), + eliminated(FALSE), engine_changed(0), changed(0), is_correlated(FALSE) { with_subselect= 1; reset(); @@ -121,6 +121,7 @@ delete scache; scache= 0; } + value_for_scache= 0; reset(); value_assigned= 0; DBUG_VOID_RETURN; @@ -129,7 +130,7 @@ void Item_singlerow_subselect::cleanup() { DBUG_ENTER("Item_singlerow_subselect::cleanup"); - value= 0; row= 0; + value= 0; row= 0; null_value_item= 0; Item_subselect::cleanup(); DBUG_VOID_RETURN; } @@ -572,7 +573,7 @@ Item_singlerow_subselect::Item_singlerow_subselect(st_select_lex *select_lex) - :Item_subselect(), value(0) + :Item_subselect(), value(0), null_value_item(0) { DBUG_ENTER("Item_singlerow_subselect::Item_singlerow_subselect"); init(select_lex, new select_singlerow_subselect(this)); @@ -760,6 +761,7 @@ (uint)depends_on.elements, (uint)test(thd->variables.optimizer_switch & OPTIMIZER_SWITCH_SUBQUERY_CACHE))); engine->fix_length_and_dec(row= &value); + DBUG_ASSERT(scache == NULL); if (depends_on.elements && optimizer_flag(thd, OPTIMIZER_SWITCH_SUBQUERY_CACHE) && !(engine->uncacheable() & (UNCACHEABLE_RAND | @@ -839,6 +841,22 @@ DBUG_RETURN(NULL); } + +/** + Puts NULL value as result in the cache +*/ + +void Item_singlerow_subselect::put_null_value_in_scache() +{ + if (!value_for_scache) + { + value_for_scache= new Item_bool_cache; + value_for_scache->set(0, TRUE); // NULL + } + DBUG_ASSERT(value_for_scache->null_value); + scache->put_value(value_for_scache); +} + double Item_singlerow_subselect::val_real() { Item *cached_value; @@ -870,7 +888,7 @@ reset(); DBUG_PRINT("info", ("error: %u", (uint)err)); if (scache && !err) - scache->put_value(&const_null_value); + put_null_value_in_scache(); DBUG_RETURN(0); } } @@ -906,7 +924,7 @@ reset(); DBUG_PRINT("info", ("error: %u", (uint)err)); if (scache && !err) - scache->put_value(&const_null_value); + put_null_value_in_scache(); DBUG_RETURN(0); } } @@ -942,7 +960,7 @@ reset(); DBUG_PRINT("info", ("error: %u", (uint)err)); if (scache && !err) - scache->put_value(&const_null_value); + put_null_value_in_scache(); DBUG_RETURN(0); } } @@ -979,7 +997,7 @@ reset(); DBUG_PRINT("info", ("error: %u", (uint)err)); if (scache && !err) - scache->put_value(&const_null_value); + put_null_value_in_scache(); DBUG_RETURN(0); } } @@ -1016,7 +1034,7 @@ reset(); DBUG_PRINT("info", ("error: %u", (uint)err)); if (scache && !err) - scache->put_value(&const_null_value); + put_null_value_in_scache(); DBUG_RETURN(0); } } @@ -1108,13 +1126,16 @@ max_columns= engine->cols(); /* We need only 1 row to determine existence */ unit->global_parameters->select_limit= new Item_int((int32) 1); + + DBUG_ASSERT(scache == NULL); + DBUG_ASSERT(value_for_scache == NULL); if (substype() == EXISTS_SUBS && depends_on.elements && optimizer_flag(thd, OPTIMIZER_SWITCH_SUBQUERY_CACHE) && !(engine->uncacheable() & (UNCACHEABLE_RAND | UNCACHEABLE_SIDEEFFECT))) { - DBUG_ASSERT(scache == NULL); - scache= new Subquery_cache_tmptable(thd, depends_on, &result); + value_for_scache= new Item_bool_cache; + scache= new Subquery_cache_tmptable(thd, depends_on, value_for_scache); DBUG_PRINT("info", ("cache: 0x%lx", (ulong) scache)); } DBUG_VOID_RETURN; @@ -1141,8 +1162,8 @@ if (scache) { - result.set(value, FALSE); - scache->put_value(&result); + value_for_scache->set(value, FALSE); + scache->put_value(value_for_scache); } DBUG_RETURN((double) value); @@ -1170,8 +1191,8 @@ if (scache) { - result.set(value, FALSE); - scache->put_value(&result); + value_for_scache->set(value, FALSE); + scache->put_value(value_for_scache); } DBUG_RETURN(value); @@ -1213,8 +1234,8 @@ if (scache) { - result.set(value, FALSE); - scache->put_value(&result); + value_for_scache->set(value, FALSE); + scache->put_value(value_for_scache); } str->set((ulonglong)value,&my_charset_bin); @@ -1257,8 +1278,8 @@ if (scache) { - result.set(value, FALSE); - scache->put_value(&result); + value_for_scache->set(value, FALSE); + scache->put_value(value_for_scache); } int2my_decimal(E_DEC_FATAL_ERROR, value, 0, decimal_value); @@ -1287,8 +1308,8 @@ if (scache) { - result.set(value, FALSE); - scache->put_value(&result); + value_for_scache->set(value, FALSE); + scache->put_value(value_for_scache); } DBUG_RETURN(value != 0); === modified file 'sql/item_subselect.h' --- a/sql/item_subselect.h 2010-05-31 21:25:54 +0000 +++ b/sql/item_subselect.h 2010-06-08 09:36:10 +0000 @@ -60,8 +60,8 @@ subselect_engine *old_engine; /* subquery cache */ Subquery_cache *scache; - /* null consrtant for caching */ - Item_null const_null_value; + /* subquery cache value for NULL and TRUE/FALSE subqueries */ + Item_bool_cache *value_for_scache; /* cache of used external tables */ table_map used_tables_cache; /* allowed number of columns (1 for single value subqueries) */ @@ -217,10 +217,15 @@ { protected: Item_cache *value, **row; + /* null value for subquery cache value */ + Item_null *null_value_item; + + void put_null_value_in_scache(); public: Item_singlerow_subselect(st_select_lex *select_lex); - Item_singlerow_subselect() :Item_subselect(), value(0), row (0) {} + Item_singlerow_subselect() :Item_subselect(), value(0), row (0), + null_value_item(0) {} void cleanup(); subs_type substype() { return SINGLEROW_SUBS; } @@ -284,8 +289,6 @@ { protected: bool value; /* value of this item (boolean: exists/not-exists) */ - /* result representation for the subquery cache */ - Item_bool_cache result; public: Item_exists_subselect(st_select_lex *select_lex);

1 0

[Maria-developers] Build Slaves
by Adam M. Dutko 08 Jun '10

08 Jun '10

Is the adutko-ultrasparc3 buildbot too slow? would it be missed?

2 1

[Maria-developers] Rev 2791: bugfixes lost in moving between trees in file:///home/bell/maria/bzr/work-maria-5.3-scache2/
by sanja＠askmonty.org 08 Jun '10

08 Jun '10

At file:///home/bell/maria/bzr/work-maria-5.3-scache2/ ------------------------------------------------------------ revno: 2791 revision-id: sanja(a)askmonty.org-20100608074734-1m60ib2tac7y9m33 parent: sanja(a)askmonty.org-20100607075845-lo3tcaiuk54qqlw0 committer: sanja(a)askmonty.org branch nick: work-maria-5.3-scache2 timestamp: Tue 2010-06-08 10:47:34 +0300 message: bugfixes lost in moving between trees === modified file 'sql/item.cc' --- a/sql/item.cc 2010-05-31 21:25:54 +0000 +++ b/sql/item.cc 2010-06-08 07:47:34 +0000 @@ -5152,6 +5152,10 @@ int Item_field::save_in_field(Field *to, bool no_conversions) { + /* if it is external field */ + if (unlikely(depended_from)) + return save_field_in_field(field, &null_value, to, no_conversions); + return save_field_in_field(result_field, &null_value, to, no_conversions); } @@ -6359,7 +6363,7 @@ int Item_ref::save_in_field(Field *to, bool no_conversions) { int res; - if (result_field) + if (result_field && !depended_from) return save_field_in_field(result_field, &null_value, to, no_conversions); res= (*ref)->save_in_field(to, no_conversions); null_value= (*ref)->null_value;

1 0

[Maria-developers] Question about License...
by Mateusz Matan 08 Jun '10

08 Jun '10

Welcome, We designed a custom Storage Engine (CLDB) for MySQL/MariaDB which now pass all premature tests. I wonder, how can we solve license problem... We would not share our sources, but we will just use shared library (*.so)... Which licenses should our customer who bought MariaDB with CLDB use to stay legally... If it is possible, can we use in corporation use MariaDB on GPL and just sell CLDB (storage engine) on other license (which one) ?? Thank You for Your quick reply... -- ___________________________________________________________ Mateusz Matan IT Security R&D Department, C/C++ programmer ComArch S.A., Al. Jana Pawła II 41d, 31-864 Kraków tel: (+48 12) 684 8411 e-mail: Mateusz.Matan(a)comarch.pl

1 0

Re: [Maria-developers] [Commits] Rev 2797: MWL#90, code movearound to unify merged and non-merged semi-join materialization processing in file:///home/psergey/dev/maria-5.3-subqueries-r12/
by Sergey Petrunya 07 Jun '10

07 Jun '10

Hi Monty, On Fri, Jun 04, 2010 at 05:36:59PM +0300, Michael Widenius wrote: > Note that this is not a full review, but just a quick scan of some of > the things in the commit. (One suspicious thing found...) Thanks for the feedback! Reply summary: - This commit was primarily to get a buildbot run, hence the presence of loads of commented-out old code and lack of real comments. I'll address this a bit later. - The suspicious thing confirmed and fixed. - Style suggestions followed. > >>>>> "Sergey" == Sergey Petrunya <psergey(a)askmonty.org> writes: > > <cut> > > Sergey> === modified file 'sql/item_cmpfunc.cc' > Sergey> --- a/sql/item_cmpfunc.cc 2010-05-25 06:32:15 +0000 > Sergey> +++ b/sql/item_cmpfunc.cc 2010-06-04 13:40:57 +0000 > Sergey> @@ -5733,6 +5733,8 @@ > Sergey> It's a field from an materialized semi-join. We can substitute it only > Sergey> for a field from the same semi-join. > Sergey> */ > Sergey> +#if 0 > Sergey> + psergey3:remove: > > Please don't ever use #if 0; Instead use something like: > > #ifdef LEFT_FOR_TESTING_WILL_BE_REMOVED_BY_PSERGEY_SOON > > Even better to just remove the code (after all, we can always find it > in bzr) > > <cut> > Sergey> - if (item->field->table->reginfo.join_tab >= first) > Sergey> + //if (item->field->table->reginfo.join_tab >= first) > > Same here; Don't leave the old code around > > <cut> > > Sergey> +bool join_tab_execution_startup(JOIN_TAB *tab) > Sergey> { > Sergey> + DBUG_ENTER("join_tab_execution_startup"); > Sergey> Item_in_subselect *in_subs; > > Please put DBUG_ENTER after all declarations. > (So that we have same layout in C and C++) > > <cut> > > Sergey> +#if 0 > > Replace with proper #if or remove code > <cut> > > Sergey> +++ b/sql/sql_select.cc 2010-06-04 13:40:57 +0000 > Sergey> @@ -1008,15 +1006,26 @@ > Sergey> /* > Sergey> Permorm the the optimization on fields evaluation mentioned above > Sergey> for all on expressions. > Sergey> - */ > Sergey> - for (JOIN_TAB *tab= join_tab + const_tables; tab < join_tab + tables ; tab++) > Sergey> + */ > Sergey> + > Sergey> { > Sergey> - if (*tab->on_expr_ref) > Sergey> + List_iterator<JOIN_TAB_RANGE> it(join_tab_ranges); > Sergey> + JOIN_TAB_RANGE *jt_range; > Sergey> + bool first= TRUE; > > Wouldn't it be better to set first to const_tables and then to 0 ? > > Sergey> + while ((jt_range= it++)) > Sergey> { > Sergey> + for (JOIN_TAB *tab= jt_range->start + (first ? const_tables : 0); > > If you do the above, then you can just do 'jt_range->start + first' here > > Sergey> + tab < jt_range->end; tab++) > Sergey> + { > Sergey> + if (*tab->on_expr_ref) > Sergey> + { > Sergey> + *tab->on_expr_ref= substitute_for_best_equal_field(*tab->on_expr_ref, > Sergey> + tab->cond_equal, > Sergey> + map2table); > Sergey> + (*tab->on_expr_ref)->update_used_tables(); > Sergey> + } > Sergey> + } > Sergey> + first= FALSE; > Sergey> } > Sergey> } > > A comment for the above outer loop would be nice. > (It's not obvious why only the first element in join_tab_ranges has > const tables) > > Sergey> @@ -1026,6 +1035,7 @@ > Sergey> { > Sergey> conds=new Item_int((longlong) 0,1); // Always false > Sergey> } > Sergey> + > Sergey> if (make_join_select(this, select, conds)) > Sergey> { > Sergey> zero_result_cause= > Sergey> @@ -1289,7 +1299,8 @@ > Sergey> if (need_tmp || select_distinct || group_list || order) > Sergey> { > Sergey> for (uint i = const_tables; i < tables; i++) > Sergey> - join_tab[i].table->prepare_for_position(); > Sergey> + table[i]->prepare_for_position(); > Sergey> + > > Isn't table[] in other order than join_tab? > (I thought that only join_tab has the const tables first) Right. Will fix this. > <cut> > > Sergey> +JOIN_TAB *first_linear_tab(JOIN *join, bool after_const_tables) > Sergey> +{ > Sergey> + JOIN_TAB *first= join->join_tab; > Sergey> + if (after_const_tables) > Sergey> + first += join->const_tables; > > remove space before '+' > > Sergey> + if (first < join->join_tab + join->top_jtrange_tables) > Sergey> + return first; > Sergey> + else > Sergey> + return NULL; > Sergey> +} > > Better to do: > > return (first < join->join_tab + join->top_jtrange_tables) ? first : 0; > > Or: > > if (first < join->join_tab + join->top_jtrange_tables) > return first; > return NULL; > Changed. > <cut> > > Sergey> +JOIN_TAB *next_linear_tab(JOIN* join, JOIN_TAB* tab, bool include_bush_roots) //psergey2: added > > Remove comments; It's trival to see that the function was added :) > > Sergey> +{ > Sergey> + if (include_bush_roots && tab->bush_children) > Sergey> + return tab->bush_children->start; > Sergey> + > Sergey> + if (tab->last_leaf_in_bush) > Sergey> + tab= tab->bush_root_tab; > Sergey> + > Sergey> + if (tab->bush_root_tab) > Sergey> + return ++tab; > > Add an assert before if (tab->last_leaf_in_bush): > > DBUG_ASSERT(!tab->last_leaf_in_bush || tab->bush_root_tab); > Just to declare that the above code is safe! Done. > Sergey> + > Sergey> + if (++tab == join->join_tab + join->top_jtrange_tables /*join->join_tab_ranges.head()->end*/) > > Move the comment to previous row and make it more clear what you are testing > (The current comment doesn't tell me much) > > Sergey> + return NULL; > Sergey> + > Sergey> + if (!include_bush_roots && tab->bush_children) > Sergey> + { > Sergey> + tab= tab->bush_children->start; > Sergey> + } > Sergey> + return tab; > > Why not do: > > return ((!include_bush_roots && tab->bush_children) ? > tab->bush_children->start : tab); > > Sergey> + if ((start? tab: ++tab) == join->join_tab_ranges.head()->end) > Sergey> + return NULL; /* End */ > > I think the above code would be more clear if you would do: > > if (start) > tab++; > if (...) > > This makes it clear that the ++tab is not just for the test but also > for future usage of tab. Changed. > Sergey> + > Sergey> + if (tab->bush_children) > Sergey> + return tab->bush_children->start; > Sergey> + > Sergey> + return tab; > > Could be combined with ? > > Sergey> +} > Sergey> + > Sergey> + > Sergey> +static Item *null_ptr= NULL; > > Can we make this const, so that if anyone tried to change this memory > location we would get an exception? Done > <cut> > > Sergey> DBUG_RETURN(TRUE); /* purecov: inspected */ > > Sergey> join_tab= parent->join_tab_reexec; > Sergey> + //psergey2: hopefully this is ok: > Sergey> + // join_tab_ranges.head()->start= join_tab; > Sergey> + // join_tab_ranges.head()->end= join_tab + 1; > > Better to know than to hope :) It wasn't ok actually, already changed. > > (sorry, don't have time to look at the rest now) BR Sergey -- Sergey Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] New (by Knielsen test): Replication API for stacked event generators (120)
by worklog-noreply＠askmonty.org 07 Jun '10

07 Jun '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Replication API for stacked event generators CREATION DATE..: Mon, 07 Jun 2010, 13:13 SUPERVISOR.....: Knielsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 120 (http://askmonty.org/worklog/?tid=120) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: A part of the replication project, MWL#107. Events are produced by event Generators. Examples are - Generation of statement-based replication events - Generation of row-based events - Generation of PBXT engine-level replication events and maybe reading of events from relay log on slave may also be an example of generating events. Event generators can be stacked, and a generator may defer event generation to the next one down the stack. For example, the row-level replication event generator may defer DLL to the statement-level replication event generator. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Knielsen test): Replication API for stacked event generators (120)
by worklog-noreply＠askmonty.org 07 Jun '10

07 Jun '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Replication API for stacked event generators CREATION DATE..: Mon, 07 Jun 2010, 13:13 SUPERVISOR.....: Knielsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 120 (http://askmonty.org/worklog/?tid=120) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: A part of the replication project, MWL#107. Events are produced by event Generators. Examples are - Generation of statement-based replication events - Generation of row-based events - Generation of PBXT engine-level replication events and maybe reading of events from relay log on slave may also be an example of generating events. Event generators can be stacked, and a generator may defer event generation to the next one down the stack. For example, the row-level replication event generator may defer DLL to the statement-level replication event generator. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): New replication APIs (107)
by worklog-noreply＠askmonty.org 07 Jun '10

07 Jun '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: New replication APIs CREATION DATE..: Mon, 15 Mar 2010, 13:55 SUPERVISOR.....: Knielsen IMPLEMENTOR....: Knielsen COPIES TO......: Sergei CATEGORY.......: Server-Sprint TASK ID........: 107 (http://askmonty.org/worklog/?tid=107) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 50 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Mon, 07 Jun 2010, 12:11)=-=- High Level Description modified. --- /tmp/wklog.107.old.31097 2010-06-07 12:11:57.000000000 +0000 +++ /tmp/wklog.107.new.31097 2010-06-07 12:11:57.000000000 +0000 @@ -7,3 +7,6 @@ https://lists.launchpad.net/maria-developers/msg01998.html +Wiki page for the project: + + http://askmonty.org/wiki/ReplicationProject -=-=(Knielsen - Mon, 29 Mar 2010, 07:33)=-=- Research and design discussions: Galera, 2pc/XA, group commit, multi-engine transactions. Worked 14 hours and estimate 0 hours remain (original estimate increased by 14 hours). -=-=(Knielsen - Wed, 24 Mar 2010, 10:39)=-=- Design discussions Worked 11 hours and estimate 0 hours remain (original estimate increased by 11 hours). -=-=(Knielsen - Mon, 15 Mar 2010, 14:28)=-=- Research into the problem, and discussions on phone/mailing list Worked 25 hours and estimate 0 hours remain (original estimate increased by 25 hours). -=-=(Guest - Mon, 15 Mar 2010, 14:18)=-=- High-Level Specification modified. --- /tmp/wklog.107.old.9086 2010-03-15 14:18:18.000000000 +0000 +++ /tmp/wklog.107.new.9086 2010-03-15 14:18:18.000000000 +0000 @@ -1 +1,43 @@ +Current ideas/status after discussions on the mailing list: + + - Implement a set of plugin APIs and use them to move all of the existing + MySQL replication into a (set of) plugins. + + - Design the APIs so that they can support full MySQL replication, but also + so that they do not hardcode assumptions about how this replication + implementation is done, and so that they will be suitable for other types of + replication (Tungsten, Galera, parallel replication, ...). + + - APIs need to include the concept of a global transaction ID. Need to + determine the extent to which the semantics of such ID will be defined + by the API, and to which extend it will be defined by the plugin + implementations. + + - APIs should properly support reliable crash-recovery with decent + performance (eg. not require multiple mandatory fsync()s per commit, and + not make group commit impossible). + + - Would be nice if the API provided facilities for implementing good + consistency checking support (mainly checking master tables against slave + tables is hard here I think, but also applying wrong binlog data and + individual event checksums). + + +Steps to make this more concrete: + + - Investigate the current MySQL replication, and list all of the places where + a plugin implementation will need to connect/hook into the MySQL server. + * handler::{write,update,delete}_row() + * Statement execution + * Transaction start/commit + * Table open + * Query safe/not/safe for statement based replication + * Statement-based logging details (user variables, random seed, etc.) + * ... + + - Use this list to make an initial sketch of the set of APIs we need. + + - Use the list to determine the feasibility of this project and the level of + detail in the API needed to support a full replication implementation as a + plugin. -=-=(Serg - Mon, 15 Mar 2010, 14:13)=-=- Observers changed: Sergei DESCRIPTION: This is a top-level task for the project of designing a new set of replication APIs for MariaDB. This task is for the initial discussion of what to do and where to focus. The project is started in this email thread: https://lists.launchpad.net/maria-developers/msg01998.html Wiki page for the project: http://askmonty.org/wiki/ReplicationProject HIGH-LEVEL SPECIFICATION: Current ideas/status after discussions on the mailing list: - Implement a set of plugin APIs and use them to move all of the existing MySQL replication into a (set of) plugins. - Design the APIs so that they can support full MySQL replication, but also so that they do not hardcode assumptions about how this replication implementation is done, and so that they will be suitable for other types of replication (Tungsten, Galera, parallel replication, ...). - APIs need to include the concept of a global transaction ID. Need to determine the extent to which the semantics of such ID will be defined by the API, and to which extend it will be defined by the plugin implementations. - APIs should properly support reliable crash-recovery with decent performance (eg. not require multiple mandatory fsync()s per commit, and not make group commit impossible). - Would be nice if the API provided facilities for implementing good consistency checking support (mainly checking master tables against slave tables is hard here I think, but also applying wrong binlog data and individual event checksums). Steps to make this more concrete: - Investigate the current MySQL replication, and list all of the places where a plugin implementation will need to connect/hook into the MySQL server. * handler::{write,update,delete}_row() * Statement execution * Transaction start/commit * Table open * Query safe/not/safe for statement based replication * Statement-based logging details (user variables, random seed, etc.) * ... - Use this list to make an initial sketch of the set of APIs we need. - Use the list to determine the feasibility of this project and the level of detail in the API needed to support a full replication implementation as a plugin. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): New replication APIs (107)
by worklog-noreply＠askmonty.org 07 Jun '10

07 Jun '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: New replication APIs CREATION DATE..: Mon, 15 Mar 2010, 13:55 SUPERVISOR.....: Knielsen IMPLEMENTOR....: Knielsen COPIES TO......: Sergei CATEGORY.......: Server-Sprint TASK ID........: 107 (http://askmonty.org/worklog/?tid=107) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 50 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Mon, 07 Jun 2010, 12:11)=-=- High Level Description modified. --- /tmp/wklog.107.old.31097 2010-06-07 12:11:57.000000000 +0000 +++ /tmp/wklog.107.new.31097 2010-06-07 12:11:57.000000000 +0000 @@ -7,3 +7,6 @@ https://lists.launchpad.net/maria-developers/msg01998.html +Wiki page for the project: + + http://askmonty.org/wiki/ReplicationProject -=-=(Knielsen - Mon, 29 Mar 2010, 07:33)=-=- Research and design discussions: Galera, 2pc/XA, group commit, multi-engine transactions. Worked 14 hours and estimate 0 hours remain (original estimate increased by 14 hours). -=-=(Knielsen - Wed, 24 Mar 2010, 10:39)=-=- Design discussions Worked 11 hours and estimate 0 hours remain (original estimate increased by 11 hours). -=-=(Knielsen - Mon, 15 Mar 2010, 14:28)=-=- Research into the problem, and discussions on phone/mailing list Worked 25 hours and estimate 0 hours remain (original estimate increased by 25 hours). -=-=(Guest - Mon, 15 Mar 2010, 14:18)=-=- High-Level Specification modified. --- /tmp/wklog.107.old.9086 2010-03-15 14:18:18.000000000 +0000 +++ /tmp/wklog.107.new.9086 2010-03-15 14:18:18.000000000 +0000 @@ -1 +1,43 @@ +Current ideas/status after discussions on the mailing list: + + - Implement a set of plugin APIs and use them to move all of the existing + MySQL replication into a (set of) plugins. + + - Design the APIs so that they can support full MySQL replication, but also + so that they do not hardcode assumptions about how this replication + implementation is done, and so that they will be suitable for other types of + replication (Tungsten, Galera, parallel replication, ...). + + - APIs need to include the concept of a global transaction ID. Need to + determine the extent to which the semantics of such ID will be defined + by the API, and to which extend it will be defined by the plugin + implementations. + + - APIs should properly support reliable crash-recovery with decent + performance (eg. not require multiple mandatory fsync()s per commit, and + not make group commit impossible). + + - Would be nice if the API provided facilities for implementing good + consistency checking support (mainly checking master tables against slave + tables is hard here I think, but also applying wrong binlog data and + individual event checksums). + + +Steps to make this more concrete: + + - Investigate the current MySQL replication, and list all of the places where + a plugin implementation will need to connect/hook into the MySQL server. + * handler::{write,update,delete}_row() + * Statement execution + * Transaction start/commit + * Table open + * Query safe/not/safe for statement based replication + * Statement-based logging details (user variables, random seed, etc.) + * ... + + - Use this list to make an initial sketch of the set of APIs we need. + + - Use the list to determine the feasibility of this project and the level of + detail in the API needed to support a full replication implementation as a + plugin. -=-=(Serg - Mon, 15 Mar 2010, 14:13)=-=- Observers changed: Sergei DESCRIPTION: This is a top-level task for the project of designing a new set of replication APIs for MariaDB. This task is for the initial discussion of what to do and where to focus. The project is started in this email thread: https://lists.launchpad.net/maria-developers/msg01998.html Wiki page for the project: http://askmonty.org/wiki/ReplicationProject HIGH-LEVEL SPECIFICATION: Current ideas/status after discussions on the mailing list: - Implement a set of plugin APIs and use them to move all of the existing MySQL replication into a (set of) plugins. - Design the APIs so that they can support full MySQL replication, but also so that they do not hardcode assumptions about how this replication implementation is done, and so that they will be suitable for other types of replication (Tungsten, Galera, parallel replication, ...). - APIs need to include the concept of a global transaction ID. Need to determine the extent to which the semantics of such ID will be defined by the API, and to which extend it will be defined by the plugin implementations. - APIs should properly support reliable crash-recovery with decent performance (eg. not require multiple mandatory fsync()s per commit, and not make group commit impossible). - Would be nice if the API provided facilities for implementing good consistency checking support (mainly checking master tables against slave tables is hard here I think, but also applying wrong binlog data and individual event checksums). Steps to make this more concrete: - Investigate the current MySQL replication, and list all of the places where a plugin implementation will need to connect/hook into the MySQL server. * handler::{write,update,delete}_row() * Statement execution * Transaction start/commit * Table open * Query safe/not/safe for statement based replication * Statement-based logging details (user variables, random seed, etc.) * ... - Use this list to make an initial sketch of the set of APIs we need. - Use the list to determine the feasibility of this project and the level of detail in the API needed to support a full replication implementation as a plugin. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Rev 2790: bugfixes in file:///home/bell/maria/bzr/work-maria-5.3-scache2/
by sanja＠askmonty.org 07 Jun '10

07 Jun '10

At file:///home/bell/maria/bzr/work-maria-5.3-scache2/ ------------------------------------------------------------ revno: 2790 revision-id: sanja(a)askmonty.org-20100607075845-lo3tcaiuk54qqlw0 parent: sanja(a)askmonty.org-20100531212554-oal32d5v360l6cul committer: sanja(a)askmonty.org branch nick: work-maria-5.3-scache2 timestamp: Mon 2010-06-07 10:58:45 +0300 message: bugfixes === modified file 'mysql-test/r/subquery_cache.result' --- a/mysql-test/r/subquery_cache.result 2010-05-31 21:25:54 +0000 +++ b/mysql-test/r/subquery_cache.result 2010-06-07 07:58:45 +0000 @@ -588,4 +588,28 @@ Subquery_cache_hit 0 Subquery_cache_miss 4 drop table t1; +#test of sql_big_tables switch and outer table reference in subquery with grouping +set option sql_big_tables=1; +CREATE TABLE t1 (a INT PRIMARY KEY, b INT); +INSERT INTO t1 VALUES (1,1),(2,1),(3,2),(4,2),(5,3),(6,3); +SELECT (SELECT t1_outer.a FROM t1 AS t1_inner GROUP BY b LIMIT 1) FROM t1 AS t1_outer; +(SELECT t1_outer.a FROM t1 AS t1_inner GROUP BY b LIMIT 1) +1 +2 +3 +4 +5 +6 +drop table t1; +set option sql_big_tables=0; +#test of function reference to outer query +set local group_concat_max_len=400; +create table t2 (a int, b int); +insert into t2 values (1,1), (2,2); +select b x, (select group_concat(x) from t2) from t2; +x (select group_concat(x) from t2) +1 1,1 +2 2,2 +drop table t2; +set local group_concat_max_len=default; set optimizer_switch='subquery_cache=default'; === modified file 'mysql-test/t/subquery_cache.test' --- a/mysql-test/t/subquery_cache.test 2010-05-31 21:25:54 +0000 +++ b/mysql-test/t/subquery_cache.test 2010-06-07 07:58:45 +0000 @@ -201,4 +201,20 @@ show status like "subquery_cache%"; drop table t1; +--echo #test of sql_big_tables switch and outer table reference in subquery with grouping +set option sql_big_tables=1; +CREATE TABLE t1 (a INT PRIMARY KEY, b INT); +INSERT INTO t1 VALUES (1,1),(2,1),(3,2),(4,2),(5,3),(6,3); +SELECT (SELECT t1_outer.a FROM t1 AS t1_inner GROUP BY b LIMIT 1) FROM t1 AS t1_outer; +drop table t1; +set option sql_big_tables=0; + +--echo #test of function reference to outer query +set local group_concat_max_len=400; +create table t2 (a int, b int); +insert into t2 values (1,1), (2,2); +select b x, (select group_concat(x) from t2) from t2; +drop table t2; +set local group_concat_max_len=default; + set optimizer_switch='subquery_cache=default'; === modified file 'sql/table.cc' --- a/sql/table.cc 2010-05-31 21:25:54 +0000 +++ b/sql/table.cc 2010-06-07 07:58:45 +0000 @@ -5187,10 +5187,16 @@ key_part_info->store_length= key_part_info->length; if ((*reg_field)->real_maybe_null()) + { key_part_info->store_length+= HA_KEY_NULL_LENGTH; + keyinfo->key_length+= HA_KEY_NULL_LENGTH; + } if ((*reg_field)->type() == MYSQL_TYPE_BLOB || (*reg_field)->real_type() == MYSQL_TYPE_VARCHAR) + { key_part_info->store_length+= HA_KEY_BLOB_LENGTH; + keyinfo->key_length+= HA_KEY_BLOB_LENGTH; // ??? + } key_part_info->type= (uint8) (*reg_field)->key_type(); key_part_info->key_type =

1 0