- developers - lists.mariadb.org

[Maria-developers] WL#243 New (by Psergey): Support >63 tables in a select
by worklog-noreply＠askmonty.org 24 Oct '11

24 Oct '11

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Support >63 tables in a select CREATION DATE..: Mon, 24 Oct 2011, 18:38 SUPERVISOR.....: IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 243 (http://askmonty.org/worklog/?tid=243) VERSION........: WorkLog-4.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Quote: > Number of joins in a view >63 > MariaDB has a limit of 63 tables joined in a select statement. This > needs to be extended in order to handle the many number of attributes > that may be connected to an anchor. Preferably a much larger number than > 63 in order to be 'future proof'. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v4.0.0)

1 0

[Maria-developers] WL#242 New (by Psergey): Parameterized VIEWs
by worklog-noreply＠askmonty.org 24 Oct '11

24 Oct '11

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Parameterized VIEWs CREATION DATE..: Mon, 24 Oct 2011, 16:57 SUPERVISOR.....: IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 242 (http://askmonty.org/worklog/?tid=242) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Quote: > We need views which can take a specified number of parameters of a > specified type. For example, the syntax could be something like: > > CREATE VIEW exampleView (@timepoint DATE) AS > SELECT FROM WHERE colDate <= @timepoint; > > Implementation could be done as simple as doing a search and replace > with the variable and the parameter given when the view is used, such > as: > > SELECT * FROM exampleView('2001-01-01'); ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v4.0.0)

1 0

Re: [Maria-developers] [Commits] Rev 3246: Fix bug lp:869036
by Sergei Golubchik 21 Oct '11

21 Oct '11

Hi, Timour! On Oct 21, timour(a)askmonty.org wrote: I don't understand, why do you need map_ptr_array (working memory provided by the caller). Looks like unnecessary complication to me. I'd do something like for (j= 0; j < bitmap_count; j++) DBUG_ASSERT(end_bit < bitmap_array[j]->n_bits); start_idx= start_bit/8/sizeof(my_bitmap_map); end_idx= end_bit/8/sizeof(my_bitmap_map); for (i= start_idx; i < end_idx; i++) { cur_res= ~0; for (j= 0; cur_res && j < bitmap_count; j++) cur_res &= bitmap_array[j]->bitmap[i]; if (cur_res) return TRUE; } cur_res= last_word_mask(end_bit); for (j= 0; cur_res && j < bitmap_count; j++) cur_res &= bitmap_array[j]->bitmap[end_idx+1]; return cur_res != 0; } and created last_word_mask() function by extracting the relevant part of create_last_word_mask() to a separate inline function. Besides, it may be that your implementation has a bug - you use a last word mask from the shortest bitmap, while you should've built it from end_bit. Now you risk taking more bits into account that end_bit dictates. > === modified file 'mysys/my_bitmap.c' > --- a/mysys/my_bitmap.c 2011-05-02 17:58:45 +0000 > +++ b/mysys/my_bitmap.c 2011-10-21 09:44:36 +0000 > @@ -410,6 +410,88 @@ void bitmap_intersect(MY_BITMAP *map, co > } > } > > + > +/* > + Check if there is some bit index between start_bit and end_bit, such that > + this is bit is set for all bitmaps in bitmap_list. > + > + SYNOPSIS > + bitmap_exists_intersection() > + bitmpap_array [in] a set of MY_BITMAPs > + map_ptr_array [in] wokring memory provided by the caller, same size as > + bitmpap_array > + bitmap_count [in] number of elements in bitmpap_array > + start_bit [in] beginning (inclusive) of the range of bits to search > + end_bit [in] end (inclusive) of the range of bits to search, must be > + no bigger than the bits of the shortest bitmap. > + > + NOTES > + This function assumes that for at least one of the bitmaps in bitmap_array all > + bits outside the range [start_bit, end_bit] are 0. As a result is not > + necessary to take care of the bits outside the range [start_bit, end_bit]. > + > + RETURN > + TRUE if an intersecion exists > + FALSE no intersection > +*/ > + > +my_bool bitmap_exists_intersection(const MY_BITMAP **bitmap_array, > + const my_bitmap_map **map_ptr_array, > + uint bitmap_count, > + uint start_bit, uint end_bit) > +{ > + const MY_BITMAP *cur_bitmap; > + my_bitmap_map cur_res, last_word_mask_inverted= 0; > + uint i, j, min_n_bits= UINT_MAX; > + uint end_word_idx= end_bit / 32; > + > + DBUG_ASSERT(bitmap_count && end_bit >= start_bit); > + for (j= 0; j < bitmap_count; j++) > + { > + cur_bitmap= bitmap_array[j]; > + DBUG_ASSERT(end_bit <= cur_bitmap->n_bits); > + /* > + Get the word that contains 'start_bit' so that intersection begins > + from that word instead of the first word of the bitmap. > + */ > + map_ptr_array[j]= cur_bitmap->bitmap + start_bit / 32; > + /* > + Find and save the last_word_mask of the shortest bitmap that > + contains end_bit in its last word. > + */ > + if (no_words_in_map(bitmap_array[j]) - 1 == end_word_idx && > + bitmap_array[j]->n_bits < min_n_bits) > + { > + min_n_bits= bitmap_array[j]->n_bits; > + last_word_mask_inverted= ~bitmap_array[j]->last_word_mask; > + } > + } > + > + /* Find the first word where the intersection of all bits is non-empty. */ > + for (i= 0; i <= end_word_idx; i++) > + { > + cur_res= 0xFFFFFFFF; > + for (j= 0; j < bitmap_count; j++) > + { > + cur_res &= *(map_ptr_array[j]); > + ++map_ptr_array[j]; > + } > + /* Clear not relevant bits using the longest last_word_mask. */ > + if (last_word_mask_inverted && i == end_word_idx) > + cur_res&= last_word_mask_inverted; > + /* > + If there is at least one non-zero bit, there was at least one sub-row > + with all bits set, that is one NULL-subrow. > + */ > + if (cur_res) > + return TRUE; > + } > + > + /* No intersection was found. */ > + return FALSE; > +} Regards, Sergei

1 0

[Maria-developers] optimizer_switch flags for next 5.3 release
by Sergei Petrunia 19 Oct '11

19 Oct '11

Hi, Written down notes from discussion about which flags should be enabled by default in the next 5.3 release: - Derived tables and keys in them should be enabled - BKA should not be enabled (doesn't make sense for 'default' query loads) - semi-joins should be enabled - Materialization (base) should be enabled - extra strategies for NULL-aware materialization - ??? - ICP should be enabled - Hash join could be enabled but in the next minor release. - Cost-based MRR should not be enabled as cost model is not tuned. when some feature is considered stable enough, its owner should apply for testing with {$current_optimizer_switch_default+new_feature}, and once that is passed, change the default in the main branch. BR Sergey -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

Re: [Maria-developers] Compiler warning about assigned but not used variables fixed
by Sergei Golubchik 18 Oct '11

18 Oct '11

Hi, Sanja! On Oct 17, sanja(a)askmonty.org wrote: > message: > Compiler warning about assigned but not used variables fixed. Pretty much ok. But please change ifdef's to __attribute__((unused)), as discussed. And see below about the commented out code, looks like you forgot to fix or uncomment it. > === modified file 'client/mysqltest.cc' > --- a/client/mysqltest.cc 2011-05-28 02:11:32 +0000 > +++ b/client/mysqltest.cc 2011-10-17 19:38:32 +0000 > @@ -5238,7 +5238,10 @@ void do_connect(struct st_command *comma > int con_port= opt_port; > char *con_options; > my_bool con_ssl= 0, con_compress= 0; > - my_bool con_pipe= 0, con_shm= 0; > + my_bool con_pipe= 0; > +#ifdef HAVE_SMEM > + my_bool con_shm= 0; > +#endif try not to overuse #ifdef's, (*) prefer __attribute__((unused)) instead. it is less error-prone and results in a more maintainable and readable code. > struct st_connection* con_slot; > > static DYNAMIC_STRING ds_connection_name; > === modified file 'mysys/ma_dyncol.c' > --- a/mysys/ma_dyncol.c 2011-09-22 09:04:00 +0000 > +++ b/mysys/ma_dyncol.c 2011-10-17 19:38:32 +0000 > @@ -1916,16 +1915,18 @@ dynamic_column_update_many(DYNAMIC_COLUM > } > > /* if (new_offset_size != offset_size) then we have to rewrite header */ > + /* > + Now we only copy. TODO: fix > + > header_delta_sign= new_offset_size - offset_size; > data_delta_sign= 0; > - copy= FALSE; > for (i= 0; i < add_column_count; i++) > { > - /* This is the check for increasing/decreasing */ > + // This is the check for increasing/decreasing > DELTA_CHECK(header_delta_sign, plan[i].hdelta, copy); > DELTA_CHECK(data_delta_sign, plan[i].ddelta, copy); > } > - > + */ why did you comment this out? > calc_param(&new_entry_size, &new_header_size, > new_offset_size, new_column_count); > > === modified file 'sql/sql_show.cc' > --- a/sql/sql_show.cc 2011-10-11 10:55:42 +0000 > +++ b/sql/sql_show.cc 2011-10-17 19:38:32 +0000 > @@ -1274,7 +1274,9 @@ int store_create_info(THD *thd, TABLE_LI > handler *file= table->file; > TABLE_SHARE *share= table->s; > HA_CREATE_INFO create_info; > +#ifdef WITH_PARTITION_STORAGE_ENGINE > bool show_table_options= FALSE; > +#endif same here, and below. > bool foreign_db_mode= (thd->variables.sql_mode & (MODE_POSTGRESQL | > MODE_ORACLE | > MODE_MSSQL | > @@ -1507,7 +1509,9 @@ int store_create_info(THD *thd, TABLE_LI > packet->append(STRING_WITH_LEN("\n)")); > if (!(thd->variables.sql_mode & MODE_NO_TABLE_OPTIONS) && !foreign_db_mode) > { > +#ifdef WITH_PARTITION_STORAGE_ENGINE > show_table_options= TRUE; > +#endif > /* > Get possible table space definitions and append them > to the CREATE TABLE statement > > === modified file 'sql/table.cc' > --- a/sql/table.cc 2011-09-02 12:10:10 +0000 > +++ b/sql/table.cc 2011-10-17 19:38:32 +0000 > @@ -5689,7 +5689,8 @@ int update_virtual_fields(THD *thd, TABL > { > DBUG_ENTER("update_virtual_fields"); > Field **vfield_ptr, *vfield; > - int error= 0; > + int error __attribute__ ((unused)); > + error= 0; FYI, you can still use initializer and __attribute__ together: int error __attribute__ ((unused)) = 0; there is no need to change the above, either way is ok. > if (!table || !table->vfield) > DBUG_RETURN(0); > > === modified file 'storage/xtradb/buf/buf0buf.c' > --- a/storage/xtradb/buf/buf0buf.c 2011-05-10 15:17:43 +0000 > +++ b/storage/xtradb/buf/buf0buf.c 2011-10-17 19:38:32 +0000 > @@ -3895,7 +3895,7 @@ buf_page_io_complete( > enum buf_io_fix io_type; > const ibool uncompressed = (buf_page_get_state(bpage) > == BUF_BLOCK_FILE_PAGE); > - enum buf_flush flush_type; > + //enum buf_flush flush_type; please send all your XtraDB changes to Percona > mutex_t* block_mutex; > Regards, Sergei (*) see for example http://www.google.com/search?q=ifdef+considered+harmful

2 2

Re: [Maria-developers] MariaDB 5.3.2 beta is being released
by Daniel Bartholomew 14 Oct '11

14 Oct '11

On Thu, 13 Oct 2011 11:13:07 -0400 Daniel Bartholomew <dbart(a)montyprogram.com> wrote: Daniel> I'm uploading the release to the primary mirror now. Is there Daniel> anything in particular we should be sure to put into the Daniel> Release Notes? The MariaDB 5.3.2 files have been mirrored to all but one of the mirrors, so I've activated the download, release notes, and changelog pages. http://kb.askmonty.org/en/mariadb-532-release-notes http://kb.askmonty.org/en/mariadb-532-changelog http://downloads.askmonty.org/mariadb/5.3.2 On the release notes page, I mentioned the Feedback plugin. Is there anything else that should be mentioned? Please let me know ASAP, as I will be sending out the announce email shortly. Thanks. -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

[Maria-developers] Feature request: Add support for a syntax check for SQL scripts
by Raphael Vullriede 14 Oct '11

14 Oct '11

Hi, we have a lot of developers writing a lot of SQL scripts against a lot of different databases. These scripts usually consists of mixed DDL (e.g. change a column definition) and DML (e.g. adding new master data). Since our applications are constantly under development we'd like to add these scripts to our continuous integration environment. For that it would be great to have a feature that checks the script against a given database without actually executing it. It should not only check the SQL syntax but also if all mentioned tables, columns etc. exists in the given database. It is not possible to use transactions for that since DDL is (currently) not transactional. I'm thinking of something like this: SET DISABLE_EXECUTION="TRUE"; -- here come the sql statements SET DISABLE_EXECUTION="FALSE"; I've tried to write such a check with an external tool but as it turned out this is virtually impossible and has a lot of loopholes. Do you think it would be possible to implement something like this in the server? Thanks, Raphael

5 7

[Maria-developers] FW:MariaDB 5.3.2 beta is being released
by Sergei Petrunia 13 Oct '11

13 Oct '11

----- Forwarded message from Sergei Petrunia <psergey(a)askmonty.org> ----- Date: Thu, 13 Oct 2011 12:55:36 +0400 From: Sergei Petrunia <psergey(a)askmonty.org> To: maria-docs(a)lists.launchpad.net, maria-dev(a)lists.launchpad.net, community(a)askmonty.org Subject: MariaDB 5.3.2 beta is being released Hello, MariaDB 5.3.2 beta is being built. The tip changeset is: revno: 3231 tags: mariadb-5.3.2-beta revision-id: psergey(a)askmonty.org-20111012201551-t0nswetvvpp2a4sm parent: psergey(a)askmonty.org-20111012174056-ov0wq91z6z4zrajd committer: Sergey Petrunya <psergey(a)askmonty.org> branch nick: 5.3-push19 timestamp: Thu 2011-10-13 00:15:51 +0400 message: Remove garbage comment The buildnumber is 255, tarbuildnum=1289. BR Sergey -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog ----- End forwarded message ----- -- BR Sergey -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] FW: [pstoev@askmonty.org: Re: MariaDB 5.3.2 beta is being released
by Sergei Petrunia 13 Oct '11

13 Oct '11

(Forwarding as this was sent to wrong 'maria-dev' address): ----- Forwarded message from Philip Stoev <pstoev(a)askmonty.org> ----- From: Philip Stoev <pstoev(a)askmonty.org> To: Sergei Petrunia <psergey(a)askmonty.org> Cc: maria-docs(a)lists.launchpad.net, maria-dev(a)lists.launchpad.net, community(a)askmonty.org Subject: Re: MariaDB 5.3.2 beta is being released Date: Thu, 13 Oct 2011 12:01:32 +0300 Organization: Monty Program AB X-Mailer: Microsoft Outlook Express 6.00.2900.5931 Windows packages have been built and are available at a new location: http://i7.askmonty.org/packages/ Philip Stoev ----- Original Message ----- From: "Sergei Petrunia" <psergey(a)askmonty.org> To: <maria-docs(a)lists.launchpad.net>; <maria-dev(a)lists.launchpad.net>; <community(a)askmonty.org> Sent: Thursday, October 13, 2011 11:55 AM Subject: MariaDB 5.3.2 beta is being released > Hello, > > MariaDB 5.3.2 beta is being built. The tip changeset is: > > revno: 3231 > tags: mariadb-5.3.2-beta > revision-id: psergey(a)askmonty.org-20111012201551-t0nswetvvpp2a4sm > parent: psergey(a)askmonty.org-20111012174056-ov0wq91z6z4zrajd > committer: Sergey Petrunya <psergey(a)askmonty.org> > branch nick: 5.3-push19 > timestamp: Thu 2011-10-13 00:15:51 +0400 > message: > Remove garbage comment > > > The buildnumber is 255, tarbuildnum=1289. > > BR > Sergey > -- > Sergei Petrunia, Software Developer > Monty Program AB, http://askmonty.org > Blog: http://s.petrunia.net/blog > ----- End forwarded message ----- -- BR Sergey -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] rqg-perpush-replication fails
by Sergei Petrunia 13 Oct '11

13 Oct '11

Hi Kristian, I've got RQG replication tests fail as follows for 5.3: http://buildbot.askmonty.org/buildbot/builders/rqg-perpush-replication/buil… http://buildbot.askmonty.org/buildbot/builders/rqg-perpush-replication-chec… 111013 2:36:57 [ERROR] /home/buildbot/buildbot_basedir/rqg-perpush-replication-checksum/build/sql/mysqld: unknown option '--binlog-annotate-rows-events' 111013 3:30:24 [ERROR] /home/buildbot/buildbot_basedir/rqg-perpush-replication/build/sql/mysqld: unknown variable 'binlog_annotate_rows_events=1' This seems like a trivial failure that was overlooked? BR Sergey -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1