- developers - lists.mariadb.org

[Maria-developers] Json Explain
by Peter Laursen 30 Sep '12

30 Sep '12

Is Explain output in Json format planned in MariaDB (refer http://glebshchepa.blogspot.in/2012/04/optimizer-new-explain-formatjson.html) ? If so when? and will the implementation details and output format be fully identical to Oracle/MySQL 5.6? (I did not find it mentioned here: http://monty-says.blogspot.dk/2012/09/now-is-good-time-to-be-part-of-future… http://blog.mariadb.org/mariadb-directions/ .. sorry if I overlooked it somewhere) -- Peter -- Webyog

4 8

Re: [Maria-developers] [Commits] Rev 3578: MDEV-495 backport --ignore-db-dir
by Sergei Golubchik 25 Sep '12

25 Sep '12

Hi, Holyfoot! On Sep 25, holyfoot(a)askmonty.org wrote: > revno: 3578 > revision-id: holyfoot(a)askmonty.org-20120925094024-yv0wpbl6k7b6bgg8 > parent: sanja(a)montyprogram.com-20120920094859-lcrxe8lr1mtg8h10 > committer: Alexey Botchkov <holyfoot(a)askmonty.org> > branch nick: mdev495 > timestamp: Tue 2012-09-25 14:40:24 +0500 > message: > MDEV-495 backport --ignore-db-dir. > The feature was backported from MySQL 5.6. > Some code was added to make commands as > SELECT * FROM ignored_db.t1; > CALL ignored_db.proc(); > USE ignored_db; > to take that option into account. Just a few comments: 1. add a test case for CREATE DATABASE, CREATE DATABASE ignored_db should fail. 2. I'd prefer to get an error "db not found" instead of "incorrect db name", is it difficult to do? 3. add a test for the following: lost+found is ignored, but CREATE DATABASE `lost+found` should succeed! (indeed, it creates a directory lost@002bfound, which is not ignored), and later SELECT, CALL, USE, SHOW DATABASES, should show that new db. > === added file 'mysql-test/r/ignore_db_dirs_basic.result' > --- a/mysql-test/r/ignore_db_dirs_basic.result 1970-01-01 00:00:00 +0000 > +++ b/mysql-test/r/ignore_db_dirs_basic.result 2012-09-25 09:40:24 +0000 > @@ -0,0 +1,28 @@ > +select @@ignore_db_dirs; > +@@ignore_db_dirs > +e,lost+found,.mysqlgui,ignored_db > +# Check that SHOW DATABASES ignores all directories from > +# @@ignore_db_dirs and all directories with names starting > +# with '.' > +SHOW DATABASES; > +Database > +information_schema > +#mysql50#.otherdir > +mtr > +mysql > +test > +USE ignored_db; > +ERROR 42000: Incorrect database name 'ignored_db' > +SELECT * FROM ignored_db.t1; > +ERROR 42000: Incorrect database name 'ignored_db' > +CALL ignored_db.p1(); > +ERROR 42000: Incorrect database name 'ignored_db' > +SELECT COUNT(*) FROM INFORMATION_SCHEMA.SCHEMATA WHERE SCHEMA_NAME='ignored_db'; > +COUNT(*) > +1 > +SET @@global.ignore_db_dirs = 'aha'; > +ERROR HY000: Variable 'ignore_db_dirs' is a read only variable > +SET @@local.ignore_db_dirs = 'aha'; > +ERROR HY000: Variable 'ignore_db_dirs' is a read only variable > +SET @@ignore_db_dirs = 'aha'; > +ERROR HY000: Variable 'ignore_db_dirs' is a read only variable When merging to 5.5 you'll need to move this test to sys_vars suite (otherwise sys_vars.all_vars will fail) > === modified file 'sql/sql_show.cc' > --- a/sql/sql_show.cc 2012-08-24 13:39:34 +0000 > +++ b/sql/sql_show.cc 2012-09-25 09:40:24 +0000 > @@ -379,6 +379,239 @@ bool mysqld_show_privileges(THD *thd) > } > > > +/** Hash of LEX_STRINGs used to search for ignored db directories. */ > +static HASH ignore_db_dirs_hash; > + > +/** > + An array of LEX_STRING pointers to collect the options at > + option parsing time. > +*/ > +static DYNAMIC_ARRAY ignore_db_dirs_array; I don't think that sql_show.cc is where all this code really belongs, but ok, as you like. Regards, Sergei

1 0

[Maria-developers] FYI: More ha_cassandra benchmark data
by Sergei Petrunia 24 Sep '12

24 Sep '12

Hi! For the record, posting some ha_cassandra benchmarks: <contents> Run #1: re-run previous benchmarks Run #2: evaluate max. network performance Run #3: Load two dbt3 databases at once Run #4: load two datasets into different nodes Results Conclusions </contents> === Run #1: re-run previous benchmarks === My results with - m1.large mysqld node (mysqld uses optimized binary) - Cassandra on 2*m1.large nodes - Loading data with the standard LOAD DATA INFILE command. - %CPU mysqld =14 sar -n DEV 10: 03:04:41 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 03:04:51 PM lo 0.00 0.00 0.00 0.00 0.00 0.00 0.00 03:04:51 PM eth0 1575.20 156.60 81.47 4458.13 0.00 0.00 0.00 On cassandra node: - jsvc %CPU 70-170... 02:51:00 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 02:51:05 PM lo 0.00 0.00 0.00 0.00 0.00 0.00 0.00 02:51:05 PM eth0 8325.20 7333.80 3376.25 1696.74 0.00 0.00 0.00 Total load time: time: 19 min. === Run #2: evaluate max. network performance === for comparison, sending /dev/zero on one host to /dev/null on another with nc: 03:39:24 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 03:39:34 PM lo 0.00 0.00 0.00 0.00 0.00 0.00 0.00 03:39:34 PM eth0 16756.70 3688.50 851.91 87477.61 0.00 0.00 0.00 sending in both directions: SQL: 03:45:36 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 03:45:46 PM lo 0.00 0.00 0.00 0.00 0.00 0.00 0.00 03:45:46 PM eth0 63308.20 4473.50 73002.80 81484.77 0.00 0.00 0.00 Cassandra: 03:47:32 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 03:47:42 PM lo 0.20 0.20 0.01 0.01 0.00 0.00 0.00 03:47:42 PM eth0 60741.50 15241.70 84110.07 70712.32 0.00 0.00 0.00 === Run #3: Load two dbt3 databases at once === - Loading two DBT-3 databases at once. - The databases have their own key spaces, and mysql tables - Still, both loaders connect to the same cassandra host. Cassandra host: 04:27:57 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 04:28:07 PM eth0 9979.10 8984.00 6006.77 2621.08 0.00 0.00 0.00 04:29:07 PM eth0 9181.80 7315.00 6416.25 2686.83 0.00 0.00 0.00 04:29:17 PM eth0 10277.00 9629.60 6047.00 2672.02 0.00 0.00 0.00 04:30:23 PM CPU %user %nice %system %iowait %steal %idle 04:30:33 PM all 34.56 24.80 24.71 0.18 0.05 15.69 04:30:53 PM all 32.08 11.20 24.25 1.31 0.08 31.07 04:31:03 PM all 35.59 22.55 22.28 0.27 0.00 19.31 it has periodical slow-downs like this: 04:32:38 PM all 53.25 1.30 19.05 3.03 0.00 23.38 04:32:39 PM all 65.84 10.40 21.78 0.00 0.00 1.98 04:32:40 PM all 37.14 33.81 18.10 0.95 0.00 10.00 04:32:41 PM all 42.65 25.98 26.96 0.00 0.00 4.41 04:32:42 PM all 33.17 39.42 19.23 0.00 0.00 8.17 04:32:43 PM all 36.89 36.89 19.90 0.00 0.00 6.31 04:32:44 PM all 39.81 35.55 13.74 0.00 0.00 10.90 04:32:45 PM all 34.45 33.97 22.01 0.00 0.00 9.57 04:32:46 PM all 35.58 36.06 21.15 0.00 0.00 7.21 04:32:47 PM all 31.05 35.16 15.07 0.00 0.00 18.72 04:32:48 PM all 41.31 26.29 20.66 0.00 0.00 11.74 04:32:49 PM all 39.02 32.20 25.37 0.00 0.00 3.41 04:32:50 PM all 26.64 34.50 13.97 0.00 0.00 24.89 04:32:51 PM all 56.93 27.72 12.87 0.00 0.00 2.48 04:32:52 PM all 71.14 15.42 11.44 0.00 0.00 1.99 04:32:53 PM all 64.22 8.82 23.04 0.00 0.00 3.92 04:32:54 PM all 76.73 1.98 19.31 0.00 0.00 1.98 04:32:55 PM all 61.24 0.00 30.14 0.96 0.00 7.66 04:32:56 PM all 29.20 0.00 32.40 0.00 0.00 38.40 04:32:57 PM all 30.29 0.00 29.88 0.00 0.00 39.83 04:32:58 PM all 23.19 0.00 20.29 0.00 0.36 56.16 04:32:59 PM all 23.34 0.00 16.38 0.00 0.35 59.93 04:33:00 PM all 20.07 0.00 21.90 0.00 0.00 58.03 04:33:01 PM all 24.20 0.00 18.86 0.00 0.00 56.94 04:33:02 PM all 20.07 0.00 22.18 0.00 0.70 57.04 Is this a compaction? mysqld host: 04:29:01 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 04:29:01 PM eth0 1416.20 247.90 75.43 6578.11 0.00 0.00 0.00 04:29:11 PM eth0 1175.00 199.60 62.49 5292.67 0.00 0.00 0.00 04:29:21 PM eth0 1074.60 186.90 57.21 4952.19 0.00 0.00 0.00 04:30:30 PM CPU %user %nice %system %iowait %steal %idle 04:30:40 PM all 7.77 0.00 0.56 0.00 0.05 91.62 04:30:50 PM all 8.20 0.00 0.60 0.00 0.30 90.89 04:31:00 PM all 8.01 0.00 0.66 0.00 0.05 91.28 Conclusions: - mysqld is still idle most of the time. - Cassandra is busier. Sometimes, it is 1014/Sep/12 12:50 PM0% busy. - network use betwen mysqld and cassandra is 2x more, which is expected Run time: 24 minutes for both (compare to 19 minutes) === Run #4: load two datasets into different nodes === ==== Network ==== SQL node: 05:44:21 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 05:44:41 PM eth0 1178.10 210.10 62.81 5510.03 0.00 0.00 0.00 05:44:51 PM eth0 1062.10 200.90 57.53 5152.63 0.00 0.00 0.00 05:45:01 PM eth0 965.80 194.00 52.53 5012.62 0.00 0.00 0.00 Cassandra1: 05:44:19 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 05:44:49 PM eth0 10159.00 8659.20 4443.26 1798.75 0.00 0.00 0.00 05:44:59 PM eth0 9018.20 7750.10 3982.21 1578.30 0.00 0.00 0.00 05:45:09 PM eth0 8741.30 7521.70 3921.69 1541.55 0.00 0.00 0.00 Cassandra2: 05:44:56 PM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s 05:45:06 PM eth0 9305.60 7906.00 4121.10 1622.92 0.00 0.00 0.00 05:45:16 PM eth0 9186.50 7828.80 4258.12 1681.78 0.00 0.00 0.00 05:45:26 PM eth0 9675.70 8283.10 4306.89 1698.37 0.00 0.00 0.00 ==== CPU ==== SQL node: 05:46:45 PM CPU %user %nice %system %iowait %steal %idle 05:46:55 PM all 8.39 0.00 0.70 0.00 0.40 90.51 05:47:05 PM all 7.65 0.00 0.74 0.00 0.70 90.91 05:47:15 PM all 8.76 0.00 0.64 0.00 0.89 89.70 cassandra 1: 05:47:01 PM CPU %user %nice %system %iowait %steal %idle 05:47:11 PM all 27.28 28.24 27.55 0.00 0.05 16.89 05:47:21 PM all 31.45 24.44 27.56 1.38 0.05 15.13 05:47:31 PM all 29.46 19.60 23.19 2.35 0.04 25.36 05:47:41 PM all 38.47 0.04 14.21 0.15 0.07 47.06 05:47:51 PM all 31.04 9.92 17.97 0.39 0.08 40.61 05:48:01 PM all 24.73 28.07 21.77 0.26 0.04 25.12 05:48:11 PM all 19.79 11.67 19.23 1.83 0.07 47.41 05:48:21 PM all 20.29 0.00 11.07 0.06 0.13 68.44 05:48:31 PM all 24.08 0.00 11.96 0.13 0.13 63.69 Cassandra 2: 05:46:56 PM CPU %user %nice %system %iowait %steal %idle 05:47:06 PM all 31.97 27.14 26.91 0.09 0.05 13.85 05:47:16 PM all 32.79 20.43 23.39 2.34 0.04 21.01 05:47:26 PM all 26.43 27.07 27.29 0.14 0.09 18.98 05:47:36 PM all 17.72 8.59 12.00 2.97 0.21 58.52 05:47:46 PM all 13.63 0.00 6.77 0.06 0.30 79.23 05:47:56 PM all 27.19 3.92 8.52 0.11 0.25 60.01 05:48:06 PM all 19.94 27.83 18.47 0.13 0.08 33.54 05:48:16 PM all 13.69 16.01 14.24 2.69 0.15 53.22 05:48:26 PM all 18.13 0.00 9.57 0.25 0.38 71.66 === Results === The first client failed with ERROR 1928 (HY000) at line 145: Internal error: 'TimedOutException: Default TException.' real 4m3.463s user 0m0.008s sys 0m0.000s The second completed after: real 19m57.875s (which matches load time for loading one dataset). === Conclusions === - need to fix the Timeout error (filed as MDEV-535) - CPU time spent by mysqld is negligible now (I don't know whether this comes from using faster CPU for SQL node, or using a release build of mariadb, or both) - A confirmation of previous findings: if we use only one connection, the speed at which we can insert the data is very limited (see: with two connections, we have inserted 2x data in 20% more time. This is where we used two connections to connect to the *same* node) - I am not sure whether I should just start using multiple connections, or also automatically route the data (both writes and reads) to the right node? - need to check out Hector client. It seems to be able to connect to multiple nodes of Cassandra cluster. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] Updates to mytop (from MariaDB 5.5 distribution)
by Michael Widenius 21 Sep '12

21 Sep '12

Hi! I have done some extensions to mytop as part of porting, fixing and enhancing it for MariaDB. I have now merged my changes with 1.91 and thought it was time to share the changes with you. Sorry for not doing it sooner! I also have added Jean Weisbuch patches for --full-queries and reading of .my.cnf Here is a diff of my changes. Hope you find some of them useful! If you have any questions, feel free to ask! Best regards, Monty Creator of MySQL and MariaDB

1 0

[Maria-developers] Cassandra SE: problems including it into release
by Sergei Petrunia 21 Sep '12

21 Sep '12

Hello, I've figured there may be an additional challenge in getting Cassandra storage engine into a *MariaDB release*. The problem is packaging. <contents> Cassandra SE packaging problems Need Thrift library Need recent gcc How did Sphinx SE manage to avoid a similar mess? Possible way out </contents> == Cassandra SE packaging problems == === Need recent gcc === C++ code generated by Thrift uses boost, and so one needs a recent gcc to compile it with MySQL. We have the same issue with oqgraph. See MDEV-501 for details. I am not aware of crashing problems that ograph+boost combination has. Still, recent-gcc requirement may limit the range of platforms we can build for. For example, I never managed to compile on Ubuntu 9.04. === Need Thrift library === Cassandra SE uses Thrift protocol to talk to server. In order to do this, it needs to link againist libthrift.so. There are no Thrift packages in either Fedora or Ubuntu (actually I haven't seen Thrift packages for any OS). Thrift homepage only offers source tarballs. The source tarball uses automake and compiles into 1. thrift compiler binary (we dont actually need it. Source code generated by thrift needs to be modified anyway.) 2. thrift libraries for a dozen of languages, including libthrift.so, which we do need. == How did Sphinx SE manage to avoid a similar mess? == I've asked myself a question how do we manage to include Sphinx SE without having to mess with Sphinx packages. The answer is: Sphinx SE doesn't need any sphinx libraries, either at compile time, or at runtime. It implements its client protocol fully, except for INSERT commands, for which it uses the available mysql client library to talk to Sphinx server using mysql client-server protocol. == An unlikely but possible way out == Cassandra 1.2 (the next version that is promised to be released "in Q4 2012" will introduce native network protocol. A week ago they've pushed first documentation draft. The native network protocol should allow to do everything that Thrift API allows, and more. There are no C++ clients yet, though. If I am faced with a choice betwen 1. implementing a relatively simple network protocol 2. Making Thrift packages for various platforms (and potentially convincing downstream maintainers to accept them) then I will definitely pick #1 (I have implemented network protocols in the past). There is a big risk, though: native protocol support is very new in Cassandra, they pushed it about a week ago. It was pushed into their "trunk" which is under development. I don't know if the code in their trunk is sufficiently stable for us to develop againist. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

3 3

Re: [Maria-developers] New to MariaDB
by Giacomo Picchiarelli 21 Sep '12

21 Sep '12

Thanks! At beginning Il'll try to do some benchmarks! On 13/09/2012 22:11, Michael Widenius wrote: > Hi! > >>>>>> "Sergei" == Sergei Golubchik <serg(a)askmonty.org> writes: > Sergei> Hi, Giacomo! > Sergei> On Aug 11, Giacomo Picchiarelli wrote: >>> Hi, >>> >>> I'm new to MariaDB and I'm reading the knowledge base. >>> I'm interested in studying Aria Engine (and contribute, if I can :) ), >>> are there tasks for newbie developers? > Sergei> I'm afraid, there are no tasks in the Aria engine right now :( > > Sergei> If you're interested to look at other tasks - here is the list of task, > Sergei> feel free to pick anything from it: > > Sergei> https://mariadb.atlassian.net/secure/IssueNavigator.jspa?requestId=11300 > > Actually, there are some open bugs on Aria: > > https://mariadb.atlassian.net/browse/MDEV-333 > > You can go to: > https://mariadb.atlassian.net/secure/QuickSearch.jspa > And search for "Aria" > > Same goes for lanchpad: > > https://bugs.launchpad.net/maria?field.searchtext=aria&search=Search&field.… > > Note that many of the bugs in launchpad for Aria are already fixed. > However it would be quite valuable if you could help us go trough > them and mark which still are open and try to fix them yourself. > > Another thing that can be done on Aria is to run some benchmarks on it > and try to find ways to improve the performance on it. > > Regards, > Michael Widenius > -- Giacomo Picchiarelli Skype: g.picchiarelli

2 1

Re: [Maria-developers] [Commits] Rev 3577: MDEV-521 fix. in file:///home/bell/maria/bzr/work-maria-5.3-MDEV-521/
by Sergei Petrunia 20 Sep '12

20 Sep '12

Hi Sanja, Ok to push. On Thu, Sep 20, 2012 at 12:49:03PM +0300, sanja(a)montyprogram.com wrote: > At file:///home/bell/maria/bzr/work-maria-5.3-MDEV-521/ > > ------------------------------------------------------------ > revno: 3577 > revision-id: sanja(a)montyprogram.com-20120920094859-lcrxe8lr1mtg8h10 > parent: timour(a)askmonty.org-20120917081346-nqd08d0v27w8zp5c > committer: sanja(a)montyprogram.com > branch nick: work-maria-5.3-MDEV-521 > timestamp: Thu 2012-09-20 12:48:59 +0300 > message: > MDEV-521 fix. > > After pullout item during single row subselect transformation it should be fixed properly. > === modified file 'mysql-test/r/subselect.result' > --- a/mysql-test/r/subselect.result 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/r/subselect.result 2012-09-20 09:48:59 +0000 > @@ -6163,5 +6163,20 @@ a ( 3, 3 ) NOT IN ( SELECT NULL, NULL ) > 3 NULL > set optimizer_switch=@mdev367_optimizer_switch; > DROP TABLE t1; > +# > +# MDEV-521 single value subselect transformation problem > +# > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +DROP TABLE t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'mysql-test/r/subselect_no_mat.result' > --- a/mysql-test/r/subselect_no_mat.result 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/r/subselect_no_mat.result 2012-09-20 09:48:59 +0000 > @@ -6162,6 +6162,21 @@ a ( 3, 3 ) NOT IN ( SELECT NULL, NULL ) > 3 NULL > set optimizer_switch=@mdev367_optimizer_switch; > DROP TABLE t1; > +# > +# MDEV-521 single value subselect transformation problem > +# > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +DROP TABLE t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set optimizer_switch=default; > > === modified file 'mysql-test/r/subselect_no_opts.result' > --- a/mysql-test/r/subselect_no_opts.result 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/r/subselect_no_opts.result 2012-09-20 09:48:59 +0000 > @@ -6158,6 +6158,21 @@ a ( 3, 3 ) NOT IN ( SELECT NULL, NULL ) > 3 NULL > set optimizer_switch=@mdev367_optimizer_switch; > DROP TABLE t1; > +# > +# MDEV-521 single value subselect transformation problem > +# > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +DROP TABLE t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set @optimizer_switch_for_subselect_test=null; > > === modified file 'mysql-test/r/subselect_no_scache.result' > --- a/mysql-test/r/subselect_no_scache.result 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/r/subselect_no_scache.result 2012-09-20 09:48:59 +0000 > @@ -6169,6 +6169,21 @@ a ( 3, 3 ) NOT IN ( SELECT NULL, NULL ) > 3 NULL > set optimizer_switch=@mdev367_optimizer_switch; > DROP TABLE t1; > +# > +# MDEV-521 single value subselect transformation problem > +# > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +DROP TABLE t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set optimizer_switch=default; > > === modified file 'mysql-test/r/subselect_no_semijoin.result' > --- a/mysql-test/r/subselect_no_semijoin.result 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/r/subselect_no_semijoin.result 2012-09-20 09:48:59 +0000 > @@ -6158,6 +6158,21 @@ a ( 3, 3 ) NOT IN ( SELECT NULL, NULL ) > 3 NULL > set optimizer_switch=@mdev367_optimizer_switch; > DROP TABLE t1; > +# > +# MDEV-521 single value subselect transformation problem > +# > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +f1 > +u1 > +u2 > +DROP TABLE t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set @optimizer_switch_for_subselect_test=null; > > === modified file 'mysql-test/t/subselect.test' > --- a/mysql-test/t/subselect.test 2012-08-25 06:15:57 +0000 > +++ b/mysql-test/t/subselect.test 2012-09-20 09:48:59 +0000 > @@ -5229,5 +5229,18 @@ set optimizer_switch=@mdev367_optimizer_ > > DROP TABLE t1; > > +--echo # > +--echo # MDEV-521 single value subselect transformation problem > +--echo # > +CREATE TABLE t1 (f1 char(2), PRIMARY KEY (f1)) ENGINE=MyISAM; > +INSERT INTO t1 VALUES ('u1'),('u2'); > + > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > +FLUSH TABLES; > +SELECT a.* FROM t1 a WHERE ( SELECT EXISTS ( SELECT 1 FROM t1 b WHERE b.f1 = a.f1 ) ); > + > +# Cleanup > +DROP TABLE t1; > + > --echo # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'sql/item_subselect.cc' > --- a/sql/item_subselect.cc 2012-09-17 08:13:46 +0000 > +++ b/sql/item_subselect.cc 2012-09-20 09:48:59 +0000 > @@ -1007,11 +1007,9 @@ Item_singlerow_subselect::select_transfo > } > substitution= select_lex->item_list.head(); > /* > - as far as we moved content to upper level, field which depend of > - 'upper' select is not really dependent => we remove this dependence > + as far as we moved content to upper level we have to fix dependences & Co > */ > - substitution->walk(&Item::remove_dependence_processor, 0, > - (uchar *) select_lex->outer_select()); > + substitution->fix_after_pullout(select_lex->outer_select(), &substitution); > } > DBUG_RETURN(false); > } > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] Server-id patch
by Kristian Nielsen 18 Sep '12

18 Sep '12

Robert Hodges <robert.hodges(a)continuent.com> writes: > I think we need to take this out to one of the email threads. I hope other > people would be interested in these problems. Agree, I've quoted liberally and Cc:'ed maria-developers@ > On Aug 31, 2012, at 6:14 AM PDT, Kristian Nielsen wrote: >> Robert Hodges <robert.hodges(a)continuent.com> writes: >> >>> 1.) The server-id of which I was speaking is a very simple patch to allow us >>> to do the following client command: 'set session server-id=X' where X is >>> whatever we choose. Currently MySQL replication sets the server ID from an >>> internal structure (THD as I recall), which means that applications like >>> Tungsten cannot get the server-id into the binlog when applying updates on a >>> slave. Server-id is essential to prevent replication loops. Monty and I >>> discussed this feature briefly over dinner in April. I hacked in support >>> using a UDF about a year ago but it would be better to have this become a >>> dynamic session variable. >> Yes, it does not seem too hard to do (though as always some care is needed to >> introduce it in a clean way). >> >> I created an issue for this with some descriptions: >> >> https://mariadb.atlassian.net/browse/MDEV-500 > SET SESSION pseudo_server_id=236; > > Then the server_id 236 would to go into the binlog. This seems like a good > idea and avoids accidentally reseting the server_id, which is a pretty > significant variable. Seems there is hope we can have this in 10.0. >>> 2.) I agree with your criticisms of the MySQL 5.6 replication design. It's >>> surprisingly complex and looks as if it will have a lot of bugs. I felt at >> >> Yes. They need this complexity because their parallel replication (which is >> similar to Tungsten parallel replication I think?) can arbitrarily reorder >> transactions in a slave binlog compared to the order in the master binlog. > We just use the sequential log. The slaves are complex but it does not > pollute the log. In fact, everything up to the point of applying to the > slave is purely sequential. I think their design got this wrong. Well, the MySQL --log-slave-updates work by generating completely new events on the slave as the received events are executed by the SQL thread(s). Do you mean that Tungsten does not do this, but uses the original binlog with original events from the master? How is multi-master (one slave replicating from several masters) handled (if it is handled) ? >>> the time that the better approach on transaction IDs would be to put enough >>> information into logs to enable applications to compute a vector clock on >>> the stream of binlog events coming from multiple servers. This is a >> >> Interesting, I though about a concept of "vector clock" before, but did not >> know the terminology. >> >> Aha, so your point is that even though we will apply changes from multiple >> upstream master servers in parallel on a single slave, with a vector clock, we >> can still order them on a slave according to which transactions on one server >> where visible to which transactions on the other. And stuff like >> that. Interesting! >> >> But note that to be truly "theoretically sound", all queries (ie. SELECT) as >> well as all applications that interact directly or indirectly with multiple >> servers need to be included in the vector clock. So it is not without >> complexity. > Are you thinking about holding locks? For practical purposes it's enough to > do writes only. My point is that you have a global ID that is partially > ordered by originating server. Vector clocks are the traditional way of > handling that. This is the basis for most mechanisms of conflict > resolution--you check to see if the clocks are ordered and if they are not > (because two thing happened at the same time, hence are not comparable), it > signals a conflict that must be reconciled. Well, I was mostly just reacting to the phrase "truly theoretically sound". If readers are not considered in the vector clock, then they are able to observe different states of the database at different nodes in the cluster. But as you say, that is probably acceptable. >> This does not immediately apply to the MySQL design, as they work with a >> single master (applying in parallel transactions in different databases), and >> as they arbitrarily reorder between different databases, without any attempt >> to maintain inter-database consistency. But of course, my point is that this >> design is fundamentally wrong. >> >>> theoretically sound approach to enable both synchronous and asynchronous >>> multi-master replication. (Galera incidentally uses approximately this >>> approach for sync replication as you may be aware.) If you give us server >>> IDs we don't need anything else. Oops, I forgot. Making the binlog truly >>> crash safe in every possible case is also very helpful. >> >> I think it is getting closer to being truly crash safe - can you mention the >> situations where it is not? And btw., are you interested in the recent testing >> we (especially Elena) made for testing such crash safety, and the fixes made >> for it? We found some fairly interesting (and scary) stuff... > Actually your (MP's?) fix for group commit is the most important > improvement. Can we turn sync_binlog back on without killing performance? Yes, well at least performance is substantially improved with sync_binlog=1. With a parallel load you can see improvements of 10x or more, and even single-threaded load is improved in 10.0. >> I have been thinking about this for some time. Basically, the correct approach >> is to make the binlog into a true transaction log using standard textbook >> techniques. Pre-allocate the log, make it page based (this will also improve >> performance). Do ping-pong writes at the end so that we can recover even from >> torn pages. Stuff like that. Would be cool to do, but so far I have not really >> seen much interest for something like this. > It would be a lot of work. If group commit is fixed, then most on-premise > people can solve enough problems with RAID and BBU that they don't need much > more. Right, agree, I thought you had something more subtle than sync_binlog=0 in mind. (BTW, I fixed a couple binlog crash recovery bugs recently in MariaDB, which also affect MySQL). >>> 3.) I read the email you cited. I need some time to study it in order to >>> comment intelligently. The most immediate feedback is that I doubt you want >>> to make assumptions about how slaves will apply transactions. The role of >>> the master is to provide local transactions in serial order. Looking across >>> a set of masters you then have a set of easily identifiable streams of >>> partially ordered transactions. That in turn provides a mathematically >>> tractable model for reasoning about things like update order, dependencies, >>> and potential conflicts. Any metadata you can supply about dependencies, >>> e.g., which schema(s) transactions actually touch is useful but may or may >>> not be relevant on slaves depending on the application. It's up to slaves >>> to make decisions on reordering. >> >> But think about a three-level multi-master topology. Servers A and B are >> masters for server C, which is itself a master for D. The DBA/application >> developer will declare that transactions on A and B are independent, so C can >> apply them in any order, and in parallel. Clearly on D we will want to also >> apply them in parallel, and the only way we can do that is to use the >> different server ids A and B in the stream of events in the binlog of C. >> >> So now consider a simple two-level topology with just master C and slave >> D. Why not still let C put different "server_id"s in different events, as a >> way to tell server D that events are independent? For example, if the DBA runs >> a long-running ALTER TABLE, he can ensure that no transactions touch this >> table while it runs, and just SET SESSION SERVER_ID to something else, and it >> can run in parallel on D without delaying normal transactions. >> >> Or think about MySQL 5.6 style parallel replication, where the DBA declares >> that on server A, transactions against different databases are to be >> considered independent and applied in parallel on the slave. While on server B >> such transactions are not independent. Now we want to setup slave C to >> replicate from both A and B. Clearly, C is the wrong place to configure >> whether transactions in different databases are to be considered independent - >> this is a property of how the transactions are generated, not where they are >> replicated. But if we just let the server_id of transactions on master A be a >> hash of the affected database, then MySQL 5.6 style parallel replication just >> falls out of multi-master for tree, without any of their complexity. > I have not ever seen this topology arise in practice within MySQL, as people > tend to keep things pretty homogeneous. It seems to arise mostly in data > warehouse loading where multiple applications load data into the same store > for reporting. These are more heterogeneous but arallel information from > MySQL is of limited value here--data warehouses in Vertica or Hadoop don't > have constraints and you can just jam data into them in parallel. Ok. >> It is just a matter of not artificially restricting transactions generated on >> a given server to have just one server_id value. Which is in fact exactly what >> you asked me about originally with SET SESSION SERVER_ID :) >> >> On the other hand, the slave is free to try to discover more opportunities for >> independence / parallelism if it can, and use them. But automatically >> discovering opportunities for parallelism on the slave is a very hard problem >> in general, as I am sure you agree ... > Meanwhile, I don't like the idea of using server_id in the way you describe > because it overloads the meaning. Server_id just tells us which server > actually committed the transaction in the first place. The reason we need > the session variable was to be able to signal this properly. I was not > intending to use it more broadly. Ok, I've noted your concern. It would be interesting to learn how you use the knowledge that one mysqld instance can have at most one server id in generated transactions? > More generally, I believe that if you want to allow programmers to declare > that things are independent it would be best to add SQL-level features. For > instance, Facebook has a patch that prevents transactions from spanning > schemas. This is a very useful feature, because it ensures that all updates > in one schema are independent from others. That in turn means you can > parallelize without being afraid of deadlocks (a big problem in this space) > or constraint violations. > > You could generalize this into the notion of a shard, which is a scope for > independent transactions and presumably also referential integrity. This > would be a very useful feature for SQL databases. I don't know if you have > read it but Pat Helland has a great paper on this topic called "Life beyond > Distributed Transactions: an Apostate’s Opinion." It was visionary at the > time (and still is) and has never been properly implemented in relational > DBMS to my knowledge. Well, independent schemas are nice, but not enough. There are many applications that need the slave to apply transactions in parallel even within a single table. - Kristian.

1 0

[Maria-developers] Review/help needed with MDEV-532
by Kristian Nielsen 14 Sep '12

14 Sep '12

Hi Serg! I've pushed the MDEV-181/MDEV-225/MDEV-232 to 10.0-base, thanks for review! As you remember, we discussed extending it with a separate patch for doing the checkpoint inside InnoDB in an asynchroneous manner. I have created MDEV-532 for this. It was not hard to create the patch, and it seems to work. But I could use some help with one last test failure that I am somewhat mystified by: rpl.rpl_binlog_index 'row' w7 [ fail ] Found warnings/errors in server log file! Test ended at 2012-09-13 15:44:01 line 120913 15:43:48 [ERROR] mysqld: Table './test/t1' is marked as crashed and should be repaired 120913 15:43:48 [Warning] Checking table: './test/t1' 120913 15:43:48 [ERROR] test.t1: 1 client is using or hasn't closed the table properly I assume this is related to how I create a background thread for the binlog, something that I do wrong or fail to do, so any suggestions on where to look would be appreciated. I have pushed the code here: lp:~maria-captains/maria/10.0-knielsen - Kristian.

1 1

[Maria-developers] New to MariaDB
by Giacomo Picchiarelli 13 Sep '12

13 Sep '12

Hi, I'm new to MariaDB and I'm reading the knowledge base. I'm interested in studying Aria Engine (and contribute, if I can :) ), are there tasks for newbie developers? Thank you. Regards. -- Giacomo Picchiarelli <gpicchiarelli(a)gmail.com> Skype: g.picchiarelli

2 1

Re: [Maria-developers] IDE debugging Mariadb linux build
by md＠rpzdesign.com 10 Sep '12

10 Sep '12

Timour: I respect your technical position, but I do not get my paychecks from monty. My total budgeted time to deal with "issues" related to Aria is exactly zero, so I am not going to try to ramp up learning hardware GDB breakpoints, etc. My broadcast was to enable the many other non-montyprogram developers a chance to load the project into the IDE/debugger so they can tweak or modify something and then contribute back into the community. They likely won't contribute if the only thing they can use is VI and GDB command line. Fortunately, Monty has provided good tested code over the years and that is why Maria will succeed, I do not have to do everything myself, but when I run into a problem, I have a chance at fixing it. Netbeans makes it easy to use the CmakeLists.txt file to load the project. Basically, open source gives me all the power and not ALL the responsibility, bloody magic really. Keep up the good work, marco On 9/10/2012 1:26 AM, Timour Katchaounov wrote: > Hi Marco, > >> I can report to all the developers for MariaDB on linux >> that Netbeans 7.1.2 is working VERY well on Fedora 14 - 64 bit. > > Netbeans has worked well for MySQL already few years ago, I tried it. > The main problem for me with both Netbeans and Eclipse was that both > have limited GDB support. I couldn't find a way to set hardware > breakpoints, neither to inspect easily arrays. AFAIR it was not possible > to evaluate expressions within the debugger, not sure about conditional > breakpoints. > > All the above features are indispensable when working on server bugs, or > analyzing how things work. If you have found a way to overcome these > limitations, then these IDEs might be useful for server development. > > > Timour > > >> >> Vladislav indicated he used only Windows Visual studio and OSX Xcode. >> >> It has a great Find It Grep support and the debugger is easy to set >> breakpoints and run the code. >> >> Marco >> >> On 9/7/2012 3:55 AM, Vladislav Vaintroub wrote: >>> >>>> -----Original Message----- >>>> From: maria-developers- >>>> bounces+wlad=montyprogram.com(a)lists.launchpad.net [mailto:maria- >>>> developers-bounces+wlad=montyprogram.com(a)lists.launchpad.net] On >>>> Behalf Of md(a)rpzdesign.com >>>> Sent: Freitag, 7. September 2012 03:00 >>>> To: maria-developers(a)lists.launchpad.net >>>> Subject: [Maria-developers] IDE debugging Mariadb linux build >>>> >>>> I am using cmake -G"CodeBlocks - Unix Makefiles" to load the 5.5.25 >>>> project into CodeBlocks for building and editing. >>>> >>>> Does anybody use KDevelop4 or Codeblocks to build mariadb? >>> I used KDevelop some time ago (half a year or so), and Codeblocks >>> long-long >>> ago (still with vanilla MySQL) >>> >>>> If so, how do you set the parameters up so you can run the server in >>>> gdb debug? >>> Debugging and building are slightly different things. CMake does not >>> have >>> builtin support for IDE-debugging, so you need to use a semi-manual >>> method >>> . >>> >>> I found the below method useful >>> 1. Switch to build directory >>> 2. cd mysql-test >>> 3. perl mysql-test-run.pl --start 1st --manual-debug >>> >>> This will create database directory, config file, and write path for >>> mysqld >>> executable together working directory and args to the console. You can >>> copy/paste it into IDE, where debug parameters for executable are >>> stored (I >>> do not know btw where KDevelop4 has parameters for debugging, I just >>> describe the process I used in Visual Studio and Xcode ) >>> >>>> Is there a way to run mysqld without mysql_safe monitor daemon in >>>> front >>>> of it? >>> Yes, for example the one above. >>> >>> >>> >>> >> >> >> _______________________________________________ >> Mailing list: https://launchpad.net/~maria-developers >> Post to : maria-developers(a)lists.launchpad.net >> Unsubscribe : https://launchpad.net/~maria-developers >> More help : https://help.launchpad.net/ListHelp >

1 0

Re: [Maria-developers] [knielsen@knielsen-hq.org] [Commits] Rev 3459: MDEV-232: Remove one fsync() from commit phase. in http://bazaar.launchpad.net/~maria-captains/maria/5.5
by Kristian Nielsen 09 Sep '12

09 Sep '12

Sergei Golubchik <serg(a)askmonty.org> writes: > Hi, Kristian! > > On Jun 22, Kristian Nielsen wrote: >> Can you review below patch for mdev-232 (At your convenience, not urgent)? > > Here's the review. Mostly questions, I have to confess :) Thanks a lot for your review! I've tried to answer all your questions below, and for many I have tried to clarify the patch, as noted. Do you want to see the changes for when I implement async commit_checkpoint_request() in InnoDB/XtraDB, before I push it to 10.0? >> === modified file 'sql/slave.cc' >> --- a/sql/slave.cc 2012-06-22 09:40:40 +0000 >> +++ b/sql/slave.cc 2012-06-22 10:20:29 +0000 >> @@ -1751,13 +1751,12 @@ when it try to get the value of TIME_ZON >> /* Announce MariaDB slave capabilities. */ >> DBUG_EXECUTE_IF("simulate_slave_capability_none", goto after_set_capability;); >> { >> - const char *q= >> - DBUG_EVALUATE_IF("simulate_slave_capability_old_53", >> - "SET @mariadb_slave_capability=" >> - STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_ANNOTATE), >> - "SET @mariadb_slave_capability=" >> - STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_MINE)); >> - if (mysql_real_query(mysql, q, strlen(q))) >> + int rc= DBUG_EVALUATE_IF("simulate_slave_capability_old_53", >> + mysql_real_query(mysql, STRING_WITH_LEN("SET @mariadb_slave_capability=" >> + STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_ANNOTATE))), >> + mysql_real_query(mysql, STRING_WITH_LEN("SET @mariadb_slave_capability=" >> + STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_MINE)))); >> + if (rc) > > Why? Monty asked me to not use strlen() which I had introduced due to a request from you in review of an earlier worklog. And you wrote that you did not care much one way or the other (neither do I), so I reverted the change back to the original without strlen(). (Maybe I should have done it as a separate changeset, but I took this short cut since MDEV 225/181/232 are kind of tied together and will likely be pushed together.) >> === modified file 'storage/xtradb/handler/ha_innodb.cc' >> --- a/storage/xtradb/handler/ha_innodb.cc 2012-06-15 12:54:23 +0000 >> +++ b/storage/xtradb/handler/ha_innodb.cc 2012-06-22 10:20:29 +0000 >> @@ -485,8 +486,11 @@ static MYSQL_THDVAR_ULONG(lock_wait_time >> static MYSQL_THDVAR_ULONG(flush_log_at_trx_commit, PLUGIN_VAR_OPCMDARG, >> "Set to 0 (write and flush once per second)," >> " 1 (write and flush at each commit)" >> - " or 2 (write at commit, flush once per second).", >> - NULL, NULL, 1, 0, 2, 0); >> + " or 2 (write at commit, flush once per second)." >> + " Can also be set to 3 to write and flush at each commit, even when this is" >> + " unnecessary because of binlog crash recovery; should not normally be used" >> + " due to performance degration, but provided for completeness.", > > would you like to improve the description on =1 value? For example, > 1 (write and flush on prepare or on commit) > 3 (write and flush both on prepare and on commit, usually it is redundant) Yes, good idea. Here is my suggestion: "Controls the durability/speed trade-off for commits." " Set to 0 (write and flush redo log to disk only once per second)," " 1 (flush to disk at each commit)," " 2 (write to kernel at commit but flush to disk only once per second)" " or 3 (flush to disk twice per commit, usually redundant)." " 1 and 3 guarantees that after a crash, committed transactions will" " not be lost and will be consistent with the binlog and other transactional" " engines. 2 can get inconsistent and lose transactions if there is a" " power failure or kernel crash but not if mysqld crashes. 0 has no" " guarantees in case of crash. 0 and 2 can be faster than 1 or 3.", >> /*****************************************************************//** >> +Handle a commit checkpoint request from server layer. >> +We simply flush the redo log immediately and do the notify call.*/ >> +static >> +void >> +innobase_checkpoint_request( >> + handlerton *hton, >> + void *cookie) >> +{ >> + log_buffer_flush_to_disk(); >> + ha_commit_checkpoint_notify(hton, cookie); >> +} >> + >> +/*****************************************************************//** > > What about making innobase_checkpoint_request() implementation that flushed > in the background? At least to serve as an example of the correct usage of your > API, if not for performance. And to test your code for cases when the notification > isn't coming directly from the checkpoint_request method. > > it looks like it's very easy to make innodb to invoke ha_commit_checkpoint_notify > from the background flusher thread (the one that syncs once a second). > you only need to store the cookie in the log_sys, and invoke > ha_commit_checkpoint_notify after the flush when this stored cookie isn't NULL. Yes, I would like to do this, I will try. It might even help performance, I have seen several blog posts mentioning problems with stalls at binlog rotation, and this might help. As you say, this should be easy. I need to store a list of cookies and their associated LSN (not just a single cookie), but this is infrequently executed so just a simple linked list with malloc/free of elements will work fine. >> +void >> +ha_commit_checkpoint_notify(handlerton *hton, void *cookie) > > are there other ha_xxx methods that a storage engine calls? > > typically server calls methods, like, ha_commit_trans, and they invoke storage > engine methods. Not the other way around. It is exactly for this reason I used names > like "trans_register_ha" - inverse of ha_register_trans - to emphasize that this > method is invoked by the engine to call into the server, not by the server to > call into the engine. Ah, thanks, good point. I suggest "commit_checkpoint_notify_ha()", what do you think? >> === modified file 'sql/log.h' >> --- a/sql/log.h 2012-06-22 09:46:28 +0000 >> +++ b/sql/log.h 2012-06-22 10:20:29 +0000 >> @@ -359,9 +373,14 @@ class MYSQL_QUERY_LOG: public MYSQL_LOG >> Ids start from BINLOG_COOKIE_START; the value BINLOG_COOKIE_DUMMY is special >> meaning "no binlog" (we cannot use zero as that is reserved for error return >> from log_and_order). >> + The low-order bit is reserved for an error flag. >> */ >> -#define BINLOG_COOKIE_DUMMY 1 >> -#define BINLOG_COOKIE_START 2 >> +#define BINLOG_COOKIE_ERROR_FLAG 1 >> +#define BINLOG_COOKIE_ID_INCREMENT (BINLOG_COOKIE_ERROR_FLAG*2) >> +#define BINLOG_COOKIE_DUMMY (BINLOG_COOKIE_ID_INCREMENT*1) > > Is that right? BINLOG_COOKIE_DUMMY == BINLOG_COOKIE_ID_INCREMENT ? It is "right", but I agree it is incomprehensible. I will improve it. The thing is, I need to store a number of different values in the "cookie" used with log_and_order() and unlog(): - 0, which is the error return value from log_and_order() - A pair (binlog_id, error_flag) - A special dummy binlog_id meaning that nothing was logged to the binlog. I will change it to this, what do you think? : /* We assign each binlog file an internal ID, used to identify them for unlog(). The IDs start from 0 and increment for each new binlog created. In unlog() we need to know the ID of the binlog file that the corresponding transaction was written into. We also need a special value for a corner case where there is no corresponding binlog id (since nothing was logged), as well as an error flag to mark that unlog() must return failure. We use the following macros to pack all of this information into the single ulong available with log_and_order() / unlog(). Note that we cannot use the value 0, as that is reserved as error return value from log_and_order(). */ #define BINLOG_COOKIE_ERROR_RETURN 0 #define BINLOG_COOKIE_DUMMY_ID 1 #define BINLOG_COOKIE_BASE 2 #define BINLOG_COOKIE_DUMMY(error_flag) \ ( (BINLOG_COOKIE_DUMMY_ID<<1) | ((error_flag)&1) ) #define BINLOG_COOKIE_MAKE(id, error_flag) \ ( (((id)+BINLOG_COOKIE_BASE)<<1) | ((error_flag)&1) ) #define BINLOG_COOKIE_GET_ERROR_FLAG(c) ((c) & 1) #define BINLOG_COOKIE_GET_ID(c) ( ((ulong)(c)>>1) - BINLOG_COOKIE_BASE ) #define BINLOG_COOKIE_IS_DUMMY(c) \ ( ((ulong)(c)>>1) == BINLOG_COOKIE_DUMMY_ID ) Then the code in log.cc can use these macros, and not have to depend on strange details on how to pack different values into a single ulong. I think the resulting code is much clearer, now BINLOG_COOKIE_* is only used in log_and_order() and unlog() functions, thanks! >> class MYSQL_BIN_LOG: public TC_LOG, private MYSQL_LOG >> @@ -464,15 +492,6 @@ class MYSQL_BIN_LOG: public TC_LOG, priv >> uint file_id; >> uint open_count; // For replication >> int readers_count; >> - bool need_start_event; >> - /* >> - no_auto_events means we don't want any of these automatic events : >> - Start/Rotate/Stop. That is, in 4.x when we rotate a relay log, we don't >> - want a Rotate_log event to be written to the relay log. When we start a >> - relay log etc. So in 4.x this is 1 for relay logs, 0 for binlogs. >> - In 5.0 it's 0 for relay logs too! >> - */ >> - bool no_auto_events; > > Why? Because it's some old code which isn't used nowadays? Yes. This was dead code. >> === modified file 'sql/log.cc' >> --- a/sql/log.cc 2012-06-22 09:46:28 +0000 >> +++ b/sql/log.cc 2012-06-22 10:20:29 +0000 >> @@ -3597,12 +3623,40 @@ bool MYSQL_BIN_LOG::reset_logs(THD* thd) >> + /* >> We are going to nuke all binary log files. >> - So first wait until all pending binlog checkpoints have completed. >> + Without binlog, we cannot XA recover prepared-but-not-committed >> + transactions in engines. So force a commit checkpoint first. >> + >> + Note that we take and immediately release LOCK_commit_ordered. This has >> + the effect to ensure that any on-going group commit (in >> + trx_group_commit_leader()) has completed before we request the checkpoint, >> + due to the chaining of LOCK_log and LOCK_commit_ordered in that function. >> + (We are holding LOCK_log, so no new group commit can start). >> + >> + Without this, it is possible (though perhaps unlikely) that the RESET >> + MASTER could run in-between the write to the binlog and the >> + commit_ordered() in the engine of some transaction, and then a crash >> + later would leave such transaction not recoverable. >> */ >> + mysql_mutex_lock(&LOCK_commit_ordered); >> + mysql_mutex_unlock(&LOCK_commit_ordered); >> + >> + mark_xids_active(current_binlog_id, 1); >> + do_checkpoint_request(current_binlog_id); > > why do_checkpoint_request ? There may be transactions in the current binlog file that are not yet flushed to disk in the engine. If we proceed in RESET MASTER to delete the binlog file, and then crash, then we could lose such transactions. So we request a checkpoint for the current binlog file (we already did so for any previous binlog files). And then wait for this and any other pending checkpoint to complete. Only then can we safely delete binlog files. >> @@ -3731,6 +3780,31 @@ bool MYSQL_BIN_LOG::reset_logs(THD* thd) >> + /* >> + Remove all entries in the xid_count list except the last. >> + Normally we will just be deleting all the entries that we waited for to >> + drop to zero above. But if we fail during RESET MASTER for some reason >> + then we will not have created any new log file, and we may keep the last >> + of the old entries. > > I'm not sure I understand this comment. > What's the point of keeping the last of the old entries, if old binlog > files are deleted and binary logging is probably stopped (when RESET MASTER fails > it's kinda serious)? The code in sql/log.cc assumes that binlog_xid_count_list is never empty. So even in the error case we must be sure to leave at least one entry in there. So the comment is to explain why we do not simply delete everything in the list at the start, and then insert the new entry on top of that. If we get an error half-way, error handling is just easier by doing the deletion at the end, where we know what we can delete and what not. It is true that RESET MASTER failure is quite serious, and if we cannot open a new binlog after deleting the old ones we are indeed in trouble. And there are very likely legacy bugs lurking in this error case. But with this code, at least no new error handling issues are introduced, and the rest of the code is clean and can rely on having at least one entry in the list. > If some engine confirms that everything below a certain cookie of > a particular binlog_id is flushed. Does it automatically mean that > all previous binlog_id's are flushed too in this engine? Yes. This is the relevant comment in handler.h: The TC will periodically (eg. once per binlog rotation) call commit_checkpoint_request(). When this happens, the engine must arrange for all transaction that have completed commit_ordered() to be durably flushed to disk (this does not include transactions that might be in the middle of executing commit_ordered()). When such flush has completed, the engine must call commit_checkpoint_notify_ha(), passing back the opaque "cookie". commit_ordered() calls complete in the same order as transactions are written in the binlog. And we request a checkpoint for a binlog file only when all transactions in it have completed commit_ordered(). So when an engine has flushed all transactions that completed commit_ordered(), this includes everything in the given binlog, as well as anything in previous binlogs. >> @@ -5493,7 +5656,27 @@ int MYSQL_BIN_LOG::rotate(bool force_rot >> if (!write_incident_already_locked(current_thd)) >> flush_and_sync(0); >> >> - *check_purge= true; >> + /* >> + We failed to rotate - so we have to decrement the xid_count back that >> + we incremented before attempting the rotate. >> + */ > > why you don't use mark_xid_done() here? We cannot call mark_xid_done() here, as we are holding LOCK_log - and mark_xid_done() also takes (sometimes) LOCK_log. >> @@ -5920,22 +6115,29 @@ MYSQL_BIN_LOG::write_binlog_checkpoint_e >> */ >> if (!ev.write(&log_file) && !flush_and_sync(0)) >> { >> - bool check_purge= false; >> signal_update(); >> - rotate(false, &check_purge); >> - if (check_purge) >> - purge(); > > Why did you remove rotate/purge from here? Binlog rotation triggers the writing of checkpoint event in the first place. So it does not seem right that writing checkpoint event can trigger binlog rotation. This could in principle lead to infinite rotation of new binlogs, though one would need a _very_ small max binlog size to trigger that. >> +int TC_LOG_MMAP::unlog(ulong cookie, my_xid xid) >> +{ >> + pending_cookies *full_buffer= NULL; >> + DBUG_ASSERT(*(my_xid *)(data+cookie) == xid); >> + >> + /* >> + Do not delete the entry immediately, as there may be participating storage >> + engines which implement commit_checkpoint_request(), and thus have not yet >> + flushed the commit durably to disk. >> + >> + Instead put it in a queue - and periodically, we will request a checkpoint >> + from all engines and delete a whole batch at once. >> + */ >> + mysql_mutex_lock(&LOCK_pending_checkpoint); >> + if (pending_checkpoint == NULL) >> + { >> + uint32 size= sizeof(*pending_checkpoint); >> + if (!(pending_checkpoint= >> + (pending_cookies *)my_malloc(size, MYF(MY_ZEROFILL)))) >> + { >> + my_error(ER_OUTOFMEMORY, MYF(0), size); >> + mysql_mutex_unlock(&LOCK_pending_checkpoint); >> + return 1; >> + } >> + pending_checkpoint->cookies[pending_checkpoint->count++]= cookie; > > Hm, should this be under if (pending_checkpoint == NULL) ? No :-( That seems completely, utterly broken :-( So this means that I must have not tested this code at all. Which makes me sad. Of course, with PBXT dead, I am not sure how it is even possible to test it... ? Thanks for spotting. >> + /* >> + We do an extra increment and notify here - this ensures that >> + things work also if there are no engines at all that support >> + commit_checkpoint_request. >> + */ >> + ++full_buffer->pending_count; >> + ha_commit_checkpoint_request(full_buffer, mmap_do_checkpoint_callback); >> + commit_checkpoint_notify(full_buffer); > > I'm sorry, I don't understand how you handle the case when some engines don't > support commit_checkpoint_request If an engine supports commit_checkpoint_request, then ha_commit_checkpoint_request() will ask the engine to do a checkpoint; and pending_count will be incremented for each such engine to ensure that we wait for each engine to finish the checkpoint before we delete any entries from the on-disk log. If an engine does not support commit_checkpoint_request, then we will not do anything with that engine before deleting entries. But such engines are required to make their transactions durable already in their commit() method, so deletion is always safe. If _no_ engines support ha_commit_checkpoint_request then we would not get any callback to commit_checkpoint_notify() - so we add one extra such a call ourselves, at the end. Also, the increment of pending_count and decrement in the extra commit_checkpoint_notify() call protects against this race: - We call engine A's commit_checkpoint_request() - Before we have time to call engine B's commit_checkpoint_request(), engine A does commit_checkpoint_notify(), so pending_count drops to zero and we delete entries even though B has not had a chance to do a checkpoint yet - We call engine B's commit_checkpoint_request with a void *cookie pointing to freed memory. Hope this clarifies things. >> @@ -7869,6 +8196,11 @@ int TC_LOG_BINLOG::recover(LOG_INFO *lin >> if (!binlog_checkpoint_found) >> break; >> first_round= false; >> + DBUG_EXECUTE_IF("xa_recover_expect_master_bin_000004", >> + if (0 != strcmp("./master-bin.000004", binlog_checkpoint_name) && >> + 0 != strcmp(".\\master-bin.000004", binlog_checkpoint_name)) > > Are you sure the name is always lowercase even on case-insensitive > filesystems? I'd use strcasecmp to avoid the need to think about it :) > Something like > > if (0 != strcasecmp("master-bin.000004", binlog_checkpoint_name+2)) > > (right, unsafe if strlen(binlog_checkpoint_name) < 2, but it's just a test case) I don't understand - this has nothing to do with file system case sensitivity. The string is generated inside log.cc based on what mysql-test-run.pl passes to mysqld for the --log-bin option, I don't see how this could change case. And I thought it was clearer to explicitly show the two possible cases for the full file name (otherwise the offset 2 in the comparison is completely magic). But as you said, this is just a test case, I'm happy to change it if you insists (I just did not understand the reason to change it). BTW, as I remember this was for something that was quite hard to test in mysql-test-run. Since recovery happens early in the server startup, before anything mysqltest can do. So I could not think of anything better than to inject an error insert with a configuration option. - Kristian.

4 9

[Maria-developers] MariaDB 5.5.27 Release Prep
by Daniel Bartholomew 07 Sep '12

07 Sep '12

All, I've prepped the release for MariaDB 5.5.27. We're now waiting for the mirrors to seed the files before turning on the downloads page and sending out announcements. Missing at the moment are our Windows builds because of an issue with the Windows build server. Hopefully those will be ready soon. Draft Release Notes and Changelog are located here: http://kb.askmonty.org/en/mariadb-5527-release-notes/ http://kb.askmonty.org/en/mariadb-5527-changelog/ If there is anything of particular note which should be mentioned in the release notes, please let me know. Thanks. -- Daniel Bartholomew Google+ - http://gplus.to/dbart Twitter - http://twitter.com/daniel_bart MariaDB: An Enhanced Drop-in Replacement for MySQL Website - http://mariadb.org Twitter - http://twitter.com/mariadb Google+ - http://google.com/+mariadb Facebook - http://fb.com/MariaDB.dbms Knowledgebase - http://kb.askmonty.org Monty Program - http://montyprogram.com

1 0

[Maria-developers] EXPLAIN ANALYZE
by Sergei Petrunia 03 Sep '12

03 Sep '12

Hello Igor, I've got EXPLAIN ANALYZE to work for the first queries, see details at http://kb.askmonty.org/en/explain-analyze/ I need input re whether the shown information is what you needed for 10.0 works. Also, we need to decide what should be shown in these columns for join buffering/subqueries (let's discuss this on the optimizer call tomorrow). BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] MDEV-406: EXPLAIN ANALYZE or a different syntax?
by sergii＠pisem.net 03 Sep '12

03 Sep '12

Hi. MDEV-406. The chosen syntax is "EXPLAIN ANALYZE". Pro: 1. PostgreSQL uses this syntax. Contra: 1. Full name is EXPLAIN ANALYZE SELECT. But statements are at most two words in MySQL: SHOW PROCESSLIST, CREATE TABLE, etc. Even the task description talks about EXPLAIN ANALYZE. Having EXPLAIN SELECT and (in the future) EXPLAIN UPDATE and EXPLAIN DELETE, EXPLAIN ANALYZE looks like it's EXPLAIN ANALYZE TABLE. But it is not. 2. In SQL the stament is a verb. That's the spirit of the standard. SELECT something. INSERT something. CREATE something, SHOW... We've deviated from it with the HANDLER command, unfortunately. Still, EXPLAIN ANALYZE SELECT is very unnatural English. EXPLAIN SELECT is better, because "SELECT" is perceived as a noun, as "SELECT statement". So EXPLAIN SELECT means "explain the SELECT statement". But EXPLAIN ANALYZE SELECT just doesn't map into English at all. Possible variants (keep in mind that instead of SELECT one might later have UPDATE and DELETE): EXPLAIN ANALYZED SELECT ... ANALYZE SELECT .... The last syntax is stragely consistent with EXPLAIN: We used to have both EXPLAIN table_name and EXPLAIN SELECT ... So, extending ANALYZE TABLE table_name to ANALYZE SELECT ... follows this weird historical pattern. (which is, of course, no argument, just a funny coincidence) Regards, Sergei

1 0

[Maria-developers] Please test: Cassandra batched-inserts, Batched Key Access
by Sergei Petrunia 31 Aug '12

31 Aug '12

Hi Elena, I've pushed "initial" support for batched INSERT and BKA into 5.5-cassandra, and updated http://kb.askmonty.org/en/cassandra-storage-engine to cover batching. Could you please test it? Cassandra uses "hashed" mode with BKA (see docs), which haven't been used for a while (previous user was Falcon storage engine), so it is possible that something malfunctions in BKA. Names/meaning of control variables and status counter names are preliminary and may change in the future. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

Re: [Maria-developers] [Commits] Rev 3565: MDEV-492: fixed incorrect error check. in file:///home/bell/maria/bzr/work-maria-5.3-MDEV-492/
by Sergei Golubchik 29 Aug '12

29 Aug '12

Hi, Oleksandr! On Aug 29, Oleksandr Byelkin wrote: > >> === modified file 'sql/sql_update.cc' > >> --- a/sql/sql_update.cc 2012-04-26 17:21:37 +0000 > >> +++ b/sql/sql_update.cc 2012-08-27 14:29:26 +0000 > >> @@ -872,7 +872,7 @@ int mysql_update(THD *thd, > >> id= thd->arg_of_last_insert_id_function ? > >> thd->first_successful_insert_id_in_prev_stmt : 0; > >> > >> - if (error < 0) > >> + if (error < 0 && (!thd->is_error())) > >> { > >> char buff[MYSQL_ERRMSG_SIZE]; > >> my_snprintf(buff, sizeof(buff), ER(ER_UPDATE_INFO), (ulong) found, (ulong) updated, > > Nope, this is not good. > > The bug happens because the error from remove_eq_conds() is ignored. > > You need to catch and process this error as soon as possible. > > While in your patch you let the update work all the way till the end. > > > When "update works all the way" it catched inside update loops, but was > that if update never make any update it ignores errors. > > So actually it should be my fix or both fixes, because errors > (especially such unexpected as "out of resources" could happened almost > anywhere, so issuing OK to the client without checking error state is a > mistake. I know. Before sending a review I actually tried to get a crash by generating an error in different places of the UPDATE. E.g. in the loop: INSERT t1 VALUES (1,2); UPDATE .... or in fill_record(): UPDATE t1 SET a=COLUMN_CREATE(...); or with double my_error(): ... WHERE COLUMN_CREATE(...) OR COLUMN_CREATE(...); all these cases were correctly taken care of, no crash, no ok packet. So, I think, it's enough to fix only that early failure that you have found. Regards, Sergei

1 0

Re: [Maria-developers] [Commits] Rev 3565: MDEV-492: fixed incorrect error check. in file:///home/bell/maria/bzr/work-maria-5.3-MDEV-492/
by Sergei Golubchik 28 Aug '12

28 Aug '12

Hi, Sanja! On Aug 27, sanja(a)askmonty.org wrote: > message: > MDEV-492: fixed incorrect error check. > === modified file 'mysql-test/t/errors.test' > --- a/mysql-test/t/errors.test 2012-03-12 07:56:56 +0000 > +++ b/mysql-test/t/errors.test 2012-08-27 14:29:26 +0000 > @@ -85,3 +85,12 @@ INSERT INTO t2 VALUES (1,0) ON DUPLICATE > INSERT INTO t2(a,b) VALUES (1,0) ON DUPLICATE KEY UPDATE > b=(SELECT VALUES(a)+2 FROM t1); > DROP TABLE t1, t2; > + > +--echo # > +--echo # MDEV-492: incorrect error check before sending OK in mysql_update > +--echo # > +CREATE TABLE t1 (a CHAR(3), b BLOB); > +--error ER_DYN_COL_DATA > +UPDATE t1 SET a = 'new' > +WHERE COLUMN_CREATE( 1, 'v', 1, 'w' ) IS NULL; > +drop table t1; > > === modified file 'sql/sql_update.cc' > --- a/sql/sql_update.cc 2012-04-26 17:21:37 +0000 > +++ b/sql/sql_update.cc 2012-08-27 14:29:26 +0000 > @@ -872,7 +872,7 @@ int mysql_update(THD *thd, > id= thd->arg_of_last_insert_id_function ? > thd->first_successful_insert_id_in_prev_stmt : 0; > > - if (error < 0) > + if (error < 0 && (!thd->is_error())) > { > char buff[MYSQL_ERRMSG_SIZE]; > my_snprintf(buff, sizeof(buff), ER(ER_UPDATE_INFO), (ulong) found, (ulong) updated, Nope, this is not good. The bug happens because the error from remove_eq_conds() is ignored. You need to catch and process this error as soon as possible. While in your patch you let the update work all the way till the end. This patch is better, it aborts much earlier: === modified file 'sql/sql_update.cc' --- sql/sql_update.cc 2012-08-09 15:22:00 +0000 +++ sql/sql_update.cc 2012-08-28 19:05:33 +0000 @@ -406,7 +406,7 @@ int mysql_update(THD *thd, table->file->info(HA_STATUS_VARIABLE | HA_STATUS_NO_LOCK); select= make_select(table, 0, 0, conds, 0, &error); - if (error || !limit || + if (error || !limit || thd->is_error() || (select && select->check_quick(thd, safe_update, limit))) { delete select; Regards, Sergei

1 0

Re: [Maria-developers] [Commits] Rev 3353: Performed re-factoring and re-structuring of the code for mwl#248: in file:///home/igor/maria/maria-5.5-mwl248-changes/
by Sergei Golubchik 28 Aug '12

28 Aug '12

Hi, Davi! On Jul 31, Davi Arnaut wrote: > Also, I would like to point out that my main argument a few e-mails > ago is that you cannot memset/bzero a non-POD type in C++ without > occurring in undefined/implementation-defined behavior, which was > ignored. This significantly lowers my confidence in MariaDB being a > high quality code base (save for some developers who tend to get it). :) After your first email I've checked that code. Perhaps I've missed something, but it did look like all bzero-ed structures in the igor's patch were of POD types. Regards, Sergei

2 3

[Maria-developers] MariaDB 5.3.8 Release Prep
by Daniel Bartholomew 28 Aug '12

28 Aug '12

All, The MariaDB 5.3.8 release is almost ready. The files have been uploaded and should be mirrored soon. I expect to send out the release announcement at around 10am U.S. Eastern time tomorrow (28 Aug 2012). As usual, I've created preliminary Release Notes and Changelog pages: http://kb.askmonty.org/en/mariadb-538-release-notes/ http://kb.askmonty.org/en/mariadb-538-changelog/ Please let me know ASAP if there are any particularly noteworthy items in this release which I should mention in the Release Notes. Thanks. -- Daniel Bartholomew Google+ - http://gplus.to/dbart Twitter - http://twitter.com/daniel_bart MariaDB: An Enhanced Drop-in Replacement for MySQL Website - http://mariadb.org Twitter - http://twitter.com/mariadb Google+ - http://google.com/+mariadb Facebook - http://fb.com/MariaDB.dbms Knowledgebase - http://kb.askmonty.org Monty Program - http://montyprogram.com

1 0

[Maria-developers] table_share->field vs table->field
by Markus Pilman 22 Aug '12

22 Aug '12

Hi all, I ran into a strange problem: I am working on a storage engine for MariaDB. To access field information I usually do as written in the documentation: for (Field** field = table->field; *field; ++field) { /* some code here */ } This works fine in most cases (for writing - read handler::write_row - it seems to work always) but sometimes (I think always in the index_next method) accessing **(table->field) results in a segmentation fault (but *(table->field) is not null, but something like 0x5e6f4c6d4578244d). Strange enough, this error does not happen always. In the debugger I saw, that table_share->field points to the correct location there so I replaces table->field in the above loop with table_share->field. But this also seems to not work - it now crashes on other queries (also on updates, but after restarting MariaDB it also crashes on a full table scan). So my question: how can I access the fields in a reliable manner? I need to do that to set/get the BLOB/TEXT fields. We have a method in our handler called set_blobs that resets the pointer in the record to the blobs, which looks like the following: void ha_kvstore::set_blobs(uchar *buf, kvid key, btree::buffer &raw_data) { auto current_read_pos = raw_data.const_ptr() + table_share->reclength; for (Field** field = table->field; *field; ++field) { if (!((*field)->flags & BLOB_FLAG)) continue; (*field)->table = table; Field_blob* blob = *((Field_blob**)field); uint32 packlength = blob->pack_length_no_ptr(); bmove((*field)->ptr + packlength, &current_read_pos, sizeof(char*)); current_read_pos += blob->get_length(); } } The storage engine always crashes in this code. What are we doing wrong? Any help is appreciated! Best Markus

2 2

[Maria-developers] Introduction / Os X Build Host
by Pablito 21 Aug '12

21 Aug '12

Hello I'm Pablito Vicente, I'm from Argentina, have been using Mysql for 12 years now as a web developer ( PHP mostly ). During the weekend I was trying to download a package for Mac and I notice brew was the only choice, that I used and worked but I would like to know if I can help generating builds for the latest versions and I will try to also investigate how to create packages so people can download a DMG image for easier installation. I will follow the instructions on: https://kb.askmonty.org/en/buildbot-setup-buildbot-setup-for-macosx/ to set it up but I was wondering if this will work when the computers are turned on only for this task. I don't usually leave computers on unless that's required. I could offer builds for Mountain Lion and Snow Leopard, though I think they're probably the same. well let me know if this can be of any help. I might send some emails for some help as I've been away from sys admin tasks for years. warm regards, Pablito

1 0

Re: [Maria-developers] Fwd: [Commits] Rev 3498: Fix bug mdev-447: Wrong output from the EXPLAIN command of the test case for lp bug #714999 in file:///home/tsk/mprog/src/5.5-md447/
by Sergei Petrunia 21 Aug '12

21 Aug '12

Hi Timour Ok to push. On Mon, Aug 20, 2012 at 02:47:35PM +0300, Timour Katchaounov wrote: > Sergey, > > Could you please review this simple patch. > It is a backport of several lines from your code. > > Timour > > ------------------------------------------------------------ > revno: 3498 > revision-id: timour(a)askmonty.org-20120820114533-hznbt0z7ikdxoawq > parent: knielsen(a)knielsen-hq.org-20120817123528-eggusbbjukhxshvs > committer: timour(a)askmonty.org > branch nick: 5.5-md447 > timestamp: Mon 2012-08-20 14:45:33 +0300 > message: > Fix bug mdev-447: Wrong output from the EXPLAIN command of the test case for lp bug #714999 > > The fix backports from MWL#182: Explain running statements the logic that > saves the original JOIN_TAB array of a query plan after optimization. This > array is later used during EXPLAIN to iterate over the original JOIN plan > nodes in the cases when this plan could be changed by early subquery > execution during the optimization phase of the outer query. > > > === modified file 'mysql-test/r/subselect.result' > --- a/mysql-test/r/subselect.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect.result 2012-08-20 11:45:33 +0000 > @@ -4196,7 +4196,7 @@ INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > -2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > > === modified file 'mysql-test/r/subselect_innodb.result' > --- a/mysql-test/r/subselect_innodb.result 2012-06-06 19:26:40 +0000 > +++ b/mysql-test/r/subselect_innodb.result 2012-08-20 11:45:33 +0000 > @@ -333,7 +333,7 @@ WHERE (SELECT DISTINCT b FROM t3) > 0); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 const PRIMARY PRIMARY 4 const 1 Using where; Using index > 2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -3 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 0 Using temporary > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 1 Using temporary > SELECT * > FROM t1 > WHERE t1.a = ( > @@ -386,7 +386,7 @@ select 1 from t1 where 1 like (select 1 > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 1 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 1 > -3 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +3 SUBQUERY t1 ALL NULL NULL NULL NULL 1 Using temporary; Using filesort > select 1 from t1 where 1 like (select 1 from t1 where 1 <=> (select 1 from t1 group by a1)); > 1 > 1 > > === modified file 'mysql-test/r/subselect_mat_cost_bugs.result' > --- a/mysql-test/r/subselect_mat_cost_bugs.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect_mat_cost_bugs.result 2012-08-20 11:45:33 +0000 > @@ -148,7 +148,7 @@ FROM t2 GROUP BY f1 > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 SUBQUERY t1 system NULL NULL NULL NULL 1 > -3 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +3 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > drop table t1, t2, t3; > # > # LP BUG#715034 Item_sum_distinct::clear(): Assertion `tree != 0' failed > > === modified file 'mysql-test/r/subselect_no_mat.result' > --- a/mysql-test/r/subselect_no_mat.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect_no_mat.result 2012-08-20 11:45:33 +0000 > @@ -4200,7 +4200,7 @@ INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > -2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > > === modified file 'mysql-test/r/subselect_no_opts.result' > --- a/mysql-test/r/subselect_no_opts.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect_no_opts.result 2012-08-20 11:45:33 +0000 > @@ -4196,7 +4196,7 @@ INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > -2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > > === modified file 'mysql-test/r/subselect_no_scache.result' > --- a/mysql-test/r/subselect_no_scache.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect_no_scache.result 2012-08-20 11:45:33 +0000 > @@ -4202,7 +4202,7 @@ INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > -2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > > === modified file 'mysql-test/r/subselect_no_semijoin.result' > --- a/mysql-test/r/subselect_no_semijoin.result 2012-07-25 16:41:48 +0000 > +++ b/mysql-test/r/subselect_no_semijoin.result 2012-08-20 11:45:33 +0000 > @@ -4196,7 +4196,7 @@ INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > -2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > > === modified file 'sql/sql_select.cc' > --- a/sql/sql_select.cc 2012-08-15 06:34:18 +0000 > +++ b/sql/sql_select.cc 2012-08-20 11:45:33 +0000 > @@ -271,8 +271,11 @@ Item_equal *find_item_equal(COND_EQUAL * > bool *inherited_fl); > JOIN_TAB *first_depth_first_tab(JOIN* join); > JOIN_TAB *next_depth_first_tab(JOIN* join, JOIN_TAB* tab); > -JOIN_TAB *first_breadth_first_tab(JOIN *join); > -JOIN_TAB *next_breadth_first_tab(JOIN *join, JOIN_TAB *tab); > + > +enum enum_exec_or_opt {WALK_OPTIMIZATION_TABS , WALK_EXECUTION_TABS}; > +JOIN_TAB *first_breadth_first_tab(JOIN *join, enum enum_exec_or_opt tabs_kind); > +JOIN_TAB *next_breadth_first_tab(JOIN *join, enum enum_exec_or_opt tabs_kind, > + JOIN_TAB *tab); > > /** > This handles SELECT with and without UNION. > @@ -6649,12 +6652,12 @@ double JOIN::get_examined_rows() > { > ha_rows examined_rows; > double prev_fanout= 1; > - JOIN_TAB *tab= first_breadth_first_tab(this); > + JOIN_TAB *tab= first_breadth_first_tab(this, WALK_OPTIMIZATION_TABS); > JOIN_TAB *prev_tab= tab; > > examined_rows= tab->get_examined_rows(); > > - while ((tab= next_breadth_first_tab(this, tab))) > + while ((tab= next_breadth_first_tab(this, WALK_OPTIMIZATION_TABS, tab))) > { > prev_fanout *= prev_tab->records_read; > examined_rows+= (ha_rows) (tab->get_examined_rows() * prev_fanout); > @@ -7269,23 +7272,30 @@ prev_record_reads(POSITION *positions, u > Enumerate join tabs in breadth-first fashion, including const tables. > */ > > -JOIN_TAB *first_breadth_first_tab(JOIN *join) > +JOIN_TAB *first_breadth_first_tab(JOIN *join, enum enum_exec_or_opt tabs_kind) > { > - return join->join_tab; /* There's always one (i.e. first) table */ > + /* There's always one (i.e. first) table */ > + return (tabs_kind == WALK_EXECUTION_TABS)? join->join_tab: > + join->table_access_tabs; > } > > > -JOIN_TAB *next_breadth_first_tab(JOIN *join, JOIN_TAB *tab) > +JOIN_TAB *next_breadth_first_tab(JOIN *join, enum enum_exec_or_opt tabs_kind, > + JOIN_TAB *tab) > { > + JOIN_TAB* const first_top_tab= first_breadth_first_tab(join, tabs_kind); > + const uint n_top_tabs_count= (tabs_kind == WALK_EXECUTION_TABS)? > + join->top_join_tab_count: > + join->top_table_access_tabs_count; > if (!tab->bush_root_tab) > { > /* We're at top level. Get the next top-level tab */ > tab++; > - if (tab < join->join_tab + join->top_join_tab_count) > + if (tab < first_top_tab + n_top_tabs_count) > return tab; > > /* No more top-level tabs. Switch to enumerating SJM nest children */ > - tab= join->join_tab; > + tab= first_top_tab; > } > else > { > @@ -7309,7 +7319,7 @@ JOIN_TAB *next_breadth_first_tab(JOIN *j > Ok, "tab" points to a top-level table, and we need to find the next SJM > nest and enter it. > */ > - for (; tab < join->join_tab + join->top_join_tab_count; tab++) > + for (; tab < first_top_tab + n_top_tabs_count; tab++) > { > if (tab->bush_children) > return tab->bush_children->start; > @@ -7333,7 +7343,7 @@ JOIN_TAB *first_top_level_tab(JOIN *join > > JOIN_TAB *next_top_level_tab(JOIN *join, JOIN_TAB *tab) > { > - tab= next_breadth_first_tab(join, tab); > + tab= next_breadth_first_tab(join, WALK_EXECUTION_TABS, tab); > if (tab && tab->bush_root_tab) > tab= NULL; > return tab; > @@ -7633,6 +7643,12 @@ get_best_combination(JOIN *join) > > join->top_join_tab_count= join->join_tab_ranges.head()->end - > join->join_tab_ranges.head()->start; > + /* > + Save pointers to select join tabs for SHOW EXPLAIN > + */ > + join->table_access_tabs= join->join_tab; > + join->top_table_access_tabs_count= join->top_join_tab_count; > + > update_depend_map(join); > DBUG_RETURN(0); > } > @@ -21389,8 +21405,8 @@ static void select_describe(JOIN *join, > bool printing_materialize_nest= FALSE; > uint select_id= join->select_lex->select_number; > > - for (JOIN_TAB *tab= first_breadth_first_tab(join); tab; > - tab= next_breadth_first_tab(join, tab)) > + for (JOIN_TAB *tab= first_breadth_first_tab(join, WALK_OPTIMIZATION_TABS); tab; > + tab= next_breadth_first_tab(join, WALK_OPTIMIZATION_TABS, tab)) > { > if (tab->bush_root_tab) > { > @@ -21473,16 +21489,8 @@ static void select_describe(JOIN *join, > else > { > TABLE_LIST *real_table= table->pos_in_table_list; > - /* > - Internal temporary tables have no corresponding table reference > - object. Such a table may appear in EXPLAIN when a subquery that needs > - a temporary table has been executed, and JOIN::exec replaced the > - original JOIN with a plan to access the data in the temp table > - (made by JOIN::make_simple_join). > - */ > - const char *tab_name= real_table ? real_table->alias : > - "internal_tmp_table"; > - item_list.push_back(new Item_string(tab_name, strlen(tab_name), cs)); > + item_list.push_back(new Item_string(real_table->alias, > + strlen(real_table->alias), cs)); > } > /* "partitions" column */ > if (join->thd->lex->describe & DESCRIBE_PARTITIONS) > > === modified file 'sql/sql_select.h' > --- a/sql/sql_select.h 2012-06-06 19:26:40 +0000 > +++ b/sql/sql_select.h 2012-08-20 11:45:33 +0000 > @@ -897,6 +897,19 @@ class JOIN :public Sql_alloc > > public: > JOIN_TAB *join_tab, **best_ref; > + > + /* > + For "Using temporary+Using filesort" queries, JOIN::join_tab can point to > + either: > + 1. array of join tabs describing how to run the select, or > + 2. array of single join tab describing read from the temporary table. > + > + SHOW EXPLAIN code needs to read/show #1. This is why two next members are > + there for saving it. > + */ > + JOIN_TAB *table_access_tabs; > + uint top_table_access_tabs_count; > + > JOIN_TAB **map2table; ///< mapping between table indexes and JOIN_TABs > JOIN_TAB *join_tab_save; ///< saved join_tab for subquery reexecution > > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] [Commits] Rev 3353: Performed re-factoring and re-structuring of the code for mwl#248: in file:///home/igor/maria/maria-5.5-mwl248-changes/
by Kristian Nielsen 13 Aug '12

13 Aug '12

Davi Arnaut <davi(a)twitter.com> writes: > On Wed, Aug 1, 2012 at 12:46 AM, Kristian Nielsen > <knielsen(a)knielsen-hq.org> wrote: >> Davi Arnaut <davi(a)twitter.com> writes: >> >>> On Thu, Jul 26, 2012 at 5:50 PM, Igor Babaev <igor(a)askmonty.org> wrote: >>>> + bzero(column_stats, sizeof(Column_statistics) * cnt); >>> >>> That's not allowed for non-POD types. >> >> Can you elaborate on exactly when bzero() (or memset() or similar) is allowed >> or not allowed? Ok, so I got time to research this a bit more. The class in question is like this, the original has more members but this is the essential structure: class Column_statistics { private: static const uint Scale_factor_nulls_ratio= 100000; public: Field *min_value; private: ulong nulls_ratio; public: void set_all_nulls() { ... } }; The question is whether it is ok to allocate this with alloc_root() or similar, and initialise it with bzero(pointer, sizeof(Column_statistics)). (The issue of bzero vs. memset is a separate one, not discussed in this mail). Davi points out that this is not a POD. This is because a POD must be a C++ aggregate, and aggregates can not have private members. Unfortunately, it seems the question of whether bzero is allowed for a non-POD, is more difficult to answer. If we look to the standards, the answer is no. It is not well-defined to bzero() any data except integers. This is what the standard says. However, this is unhelpful for the case at hand, as bzero to initialise a simple struct is a well-established practice in systems programming and commonly accepted to be valid. It seems generally accepted that bzero() of a POD is acceptable. Since the only thing that makes Column_statistics not a POD is the presence of private: members, the question becomes: Should bzero initialisation of classes with private members be disallowed? I could not find any definitive reference to support such disallowance. I am not sure why private members would be disallowed in aggretate, and hence POD. My guess is because aggregates may be initialised with an initialiser list, which makes no sense for private member. Or maybe compilers could re-arrance private members to be placed after public members. Or compilers could optimise based on knowledge that no external code can modify a private member (so it does not need to be spilled over a call to pthread_mutex_lock() or similar). But none of this would give a problem with bzero-initialisation after alloc. So I do not know of a language standard/compiler argument to disallow bzero-initialisation of non-POD. ---- But I would still strongly advice to be very careful with such code, and I think Davi's original concern is similar to mine. The problem is that Column_statistics already looks a lot like a nicely encapsulated proper C++ class. It has private members, accessors, it even defines some constants. One could easily imagine that in the future, there will be the temptation to extend it even more so, like add a virtual function - and then we suddenly have a big problem. bzero() is most certainly _NOT_ ok for such class. So I would personally stay well away from this grey area, and rather - _either_ use only a simple struct with no member functions or private or anything like that - and use bzero/memcpy and such. - _or_ do a full class, but then only initialise it with proper call of contructors (whether explicitly or implicitly defined). That makes it much clearer whether a given struct class is one or the other. But I suppose the optimiser code already heavily uses a different strategy. At the least, the declaration of class Column_statistics should have a comment saying that it is initialised with bzero(), so it can not use features not compatible with this. The nice thing about Davi's suggested rule "no bzero of non-POD" is that it is simple, easy to remember, and easy to state. Which is the primary point of coding style guidelines in the first place: make the rules clear, so there is less wasted time discussing what they are. Hope this helps, - Kristian.

1 1

Re: [Maria-developers] [Commits] Rev 3353: Performed re-factoring and re-structuring of the code for mwl#248: in file:///home/igor/maria/maria-5.5-mwl248-changes/
by Michael Widenius 13 Aug '12

13 Aug '12

Hi! >>>>> "Davi" == Davi Arnaut <davi(a)twitter.com> writes: <cut> Davi> It's not well defined, that's one of the reasons why it was Davi> deprecated. It does not even exist anymore in some systems, Davi> POSIX.1-2008 removed the specification of bzero(). It's well defined in the sence that it works everywhere (at least for MySQL code). >> Better to always use the same code to reset memory than doing it many >> different ways. In MySQL code bzero() has been the offical and >> recommended way to reset memory. Davi> Not anymore, Oracle MySQL has addressed this issue. In MariaDB code that is still not the case. Personally, I find bzero() easier to read, instead of memset() as there is no extra parameter and it's also easier grep for it. I assume you agree that for all particular purposes, there is no reason to not use bzero() in MariaDB code? Regards, Monty

2 1

[Maria-developers] Thoughts on Global Transaction ID, parallel slave, and multi-source replication
by Kristian Nielsen 10 Aug '12

10 Aug '12

This is a writeup of some thoughts I have on getting Global Transaction ID in MariaDB, as well as some related thoughts on parallel slave and multi-source replication. MySQL 5.6 has global transaction ID now, and I looked into it in some details. Unfortunately, I _really_ did not like what I saw. There are two main problems, in my opinion: 1. The basic design is bad, leading to an excessively complicated implementation. 2. The implementation looks very incomplete/immature, to the point that I am amazed that anyone would push code like this into a main tree. Point 2 means that I really do not want to maintain this piece of **** in MariaDB, though we can hope that things will improve in subsequent releases. Point 1 leads me to suggest that instead of merging the MySQL 5.6 implementation, we could do our own implementation, using a design that is both much simpler, and at the same time more flexible that what is in MySQL 5.6. I will elaborate further on these points in the rest of the mail. ---- With respect to point 2, let me quote this little gem from the MySQL 5.6 code: /* Apparently this code is not being called. We need to investigate if this is a bug or this code is not necessary. /Alfranio */ DBUG_ASSERT(0); /NOTREACHED/ Now, certainly, there are dark corners of the MySQL replication code, and adding a comment/assert like this could be justified if a developer encounters it and finds it may be dead code. But this is _new_ code, added in a new 5.6 release! So the authors of the new feature, by their own statement, have no idea what the new code they are pushing into the main tree is actually doing. That is just insane. I have no idea what the replication team at MySQL is up to these days pulling stunts like this, I have to confess I am disappointed. I found many similar examples in the 5.6 code. I wrote up a more detailed analysis in TODO-171, where I elaborate more on the problems I see with the MySQL 5.6 Global Transaction ID code: https://mariadb.atlassian.net/browse/TODO-171 When the authors do not understand the code themselves, it really does not seem attractive for us to try to support and bugfix it. I also found a lot of what I think are left-over, unsused parts of an even more complicated design that was apparently abandoned (or postponed?), but left in the code, which does not make supporting this any easier. ---- To explain point 1, I first need to mention the MySQL 5.6 feature of parallel slave apply, as I believe that a major factor in the design of MySQL Global Transaction ID has been the need to interoperate with that feature. The MySQL parallel replication slave feature works so that transactions on tables in different databases can be run in parallel with each other (incidentally, there is some indication that the MySQL replication team considers this *the* meaning of parallel slave, which is sad, as it is just one piece of the general problem of replication slave scalability - it helps in some workloads, but it is useless for others). So with parallel slave, transactions on the master can be executed in a different order on the slave. This means that the crucial notion of a slaves current position in the replication stream becomes more complicated - it is not just a single point in the master binlog, rather every slave replication thread can be at a different point in the master binlog. MySQL 5.6 solves this by introducing "Gtid sets". A Gtid is a MySQL 5.6 global transaction ID - a unique, 24-byte identifier for every transaction. The state of a replication slave at any given moment is then the set Gtids of all transactions ever executed on that slave server This notion of having a set of all transactions ever executed really complicates the design. There is a need to represent such a set as basically intervals of "holes" in the Gtid sequence, to avoid the memory representation of such sets growing without bound over time. This in turns introduces the requirement that there can _never_ be holes in the sequence of Gtids generated on a master - if there ever is introduced a hole, slaves will wait around for ever waiting for a transaction containing the missing Gtid. I think recovering from errors with such a concept will be forever a nightmare. I think this basic fundament of the design is the main source of my dislike for the MySQL global transaction ID. It sounds nice on a first look, but once you start to think deeper on the problem it just does not seem to work well. ---- So I think I have a better idea of how to make Global Transaction ID work well with parallel slave based on executing transactions in different databases out-of-order. The MySQL Gtid is defined as a pair of server-id and sequence number. Every master server has a single server-id (despite no less than 128 bits being allocated for this - anyone planning a replication hierarchy with 2**128 servers?). The master generates a single stream of events in the binlog, though streams from other servers may be interleaved with it in case of multi-master. It is then the task of a slave to try to split up the single master stream in to multiple streams that can be executed in parallel. I think this is the root of the problem. The slave is the wrong place to do this splitting! It is on the master that we know how to interpret transactions (ie. that transactions in different databases are / are not independent), so doing this on the slave is harder. And two slaves can in general split differently, which makes it harder for a slave deeper in the replication hierarchy to switch to a new master (that master being itself a slave of a higher-up master), leading to the need for a complex set-of-all-executed-transactions concept. Suppose we instead make the decision on the _master_ on how to split up the replication stream for parallel apply. Instead of a single server-id for the master, we can have multiple - let us call them "Source ID". On the _master_, we will decide that two transactions in different databases are independent, and assign them different Source ID. The slave then receives multiple replication streams, and is free to execute each stream in parallel - or not, at the discretion of the DBA. Transactions _within_ a single stream are always executed in-order, so the slave has no need to remember any sets of all transactions ever executed - all it needs to remember is the sequence number within each stream it has seen - ie. last sequence number for each Source ID. This is a simple and obvious generalisation of the old concept of current binlog position. Downstream slaves see the exact same binlog sequences, each with transactions in the same order (though sequences may be interleaved differently) - so changing masters is simpler, all servers in the hierarchy have the same view. So to implement the same functionality as MySQL 5.6 parallel replication, all we need is on the master to generate Source ID as a hash of the database name (just as MySQL 5.6 assigns transactions to slave threads based on such hash of the database name). And implement a facility on the slave such that transactions with different Source ID can be executed in parallel by multiple slave threads. And now, since we decide on the master, we can provide much more flexibility for the application. We can have _some_ databases db1,db2 be executed in parallel on the slave - while other databases db3,db4 are considered dependent, and not modified out-of-order on the slave. Or the application can declare that two transactions are independent within the same database or even within the same table; for example log record inserts into a table can often be allowed to happen in different order on slaves. Or the DBA can assign a special Source ID to a long-runnign ALTER TABLE, and such ALTER TABLE can then execute independently on slaves to not stall the normal load. All we need is to provide a simple facility (eg. session variable) that applications/users can use to specify Source ID. And then the next step is multi-source replication, which is the much-request feature that one slave can replicate from two or more different masters at the same time. The different masters will just supply different Source IDs, so apart from the need to maintain multiple TCP connections, one to each master, from the rest of the code's point of view multi-source replication is no different, it is just more independent replication streams. So my suggestion is: rather than struggle to try and merge and support the MySQL 5.6 Global Transaction ID and parallel slave, let us instead implement something like the above idea, with multiple Source IDs on the master, and parallel execution of distinct replication streams on the slave. I believe this will be no more work, and the end result will be both simpler and more flexible. Thanks, - Kristian.

4 9

[Maria-developers] Cost analysis: checking if join optimization costs match the reality.
by Sergei Petrunia 09 Aug '12

09 Aug '12

Hi! I've ran this experiment: I took the scalar subquery out of the query and then tried to compare optimizer costs with execution times. == The query == explain extended select sql_calc_found_rows s_name, s_address from nation straight_join supplier where s_suppkey in (select ps_suppkey from partsupp where ps_partkey in (select p_partkey from part where p_name like 'forest%') ) and s_nationkey = n_nationkey and n_name = 'CANADA' order by s_name limit 10; +--+-----------+--------+------++-------------+-------+-------------------+----+---------------------------------------------------------+ |id|select_type|table |type ||key |key_len|ref |rows|Extra | +--+-----------+--------+------++-------------+-------+-------------------+----+---------------------------------------------------------+ | 1|PRIMARY |nation |ref ||n_name |26 |const | 1|Using where; Using index; Using temporary; Using filesort| | 1|PRIMARY |supplier|ref ||i_s_nationkey|5 |nation.n_nationkey | 207| | | 1|PRIMARY |partsupp|ref ||i_ps_suppkey |4 |supplier.s_suppkey | 50|Using index | | 1|PRIMARY |part |eq_ref||PRIMARY |4 |partsupp.ps_partkey| 1|Using where; FirstMatch(supplier) | +--+-----------+--------+------++-------------+-------+-------------------+----+---------------------------------------------------------+ +--+------------+-----------+------++-------------+-------+------------------+----+---------------------------------------------------------+ |id|select_type |table |type ||key |key_len|ref |rows|Extra | +--+------------+-----------+------++-------------+-------+------------------+----+---------------------------------------------------------+ | 1|PRIMARY |nation |ref ||n_name |26 |const | 1|Using where; Using index; Using temporary; Using filesort| | 1|PRIMARY |supplier |ref ||i_s_nationkey|5 |nation.n_nationkey| 207| | | 1|PRIMARY |<subquery2>|eq_ref||distinct_key |4 |func | 1| | | 2|MATERIALIZED|part |range ||p_name |58 |NULL |2387|Using where; Using index | | 2|MATERIALIZED|partsupp |ref ||i_ps_partkey |4 |part.p_partkey | 2|Using index | +--+------------+-----------+------++-------------+-------+------------------+----+---------------------------------------------------------+ === Execution times === FirstMatch: 0.20 - 0.30 sec. Materialization: 0.20 sec - 0.30 sec run times for both fluctuate between 0.20 and 0.30 sec, FirstMatch seems to be somewhat slower (dunno if the difference is statistically meaningful) == Optimizer costs === FirstMatch: 11029.4 Materialization: 3594.5 === Read record counters === (source data is plan counters provided below) FirstMatch: 43185 Materialization: 12354 + 9K temptable writes + 400 reads === Conclusions === - optimizer thinks FirstMatch is 3x more expensive. - In reality it is *somewhat* expensive. - Record counters say FirstMatch is 3.5x .. 1.9x more expensive, depending on how we treat temptable writes/reads ==> Overall, I see no big discrepancies. ==> Also, I have also discovered that JOIN::get_prefix_cost_and_fanout() forgets to add record_count/TIME_FOR_COMPARE, which causes (slightly) different cost for SJ-Materialization. (I have noted the difference in some previous email) === Firstmatch plan counters === MariaDB [dbt3sf1]> show table_statistics; +--------------+------------+-----------+--------------+-------------------------+ | Table_schema | Table_name | Rows_read | Rows_changed | Rows_changed_x_#indexes | +--------------+------------+-----------+--------------+-------------------------+ | dbt3sf1 | nation | 1 | 0 | 0 | | dbt3sf1 | part | 21386 | 0 | 0 | | dbt3sf1 | partsupp | 21386 | 0 | 0 | | dbt3sf1 | supplier | 412 | 0 | 0 | +--------------+------------+-----------+--------------+-------------------------+ 4 rows in set (0.00 sec) MariaDB [dbt3sf1]> show index_statistics; +--------------+------------+---------------+-----------+ | Table_schema | Table_name | Index_name | Rows_read | +--------------+------------+---------------+-----------+ | dbt3sf1 | nation | n_name | 1 | | dbt3sf1 | part | PRIMARY | 21386 | | dbt3sf1 | supplier | i_s_nationkey | 412 | | dbt3sf1 | partsupp | i_ps_suppkey | 21386 | +--------------+------------+---------------+-----------+ Total= 43185 rows read === Materialization plan counters === MariaDB [dbt3sf1]> show index_statistics; +--------------+------------+---------------+-----------+ | Table_schema | Table_name | Index_name | Rows_read | +--------------+------------+---------------+-----------+ | dbt3sf1 | nation | n_name | 1 | | dbt3sf1 | part | p_name | 2389 | | dbt3sf1 | supplier | i_s_nationkey | 412 | | dbt3sf1 | partsupp | i_ps_partkey | 9552 | +--------------+------------+---------------+-----------+ 4 rows in set (0.00 sec) MariaDB [dbt3sf1]> show table_statistics; +--------------+------------+-----------+--------------+-------------------------+ | Table_schema | Table_name | Rows_read | Rows_changed | Rows_changed_x_#indexes | +--------------+------------+-----------+--------------+-------------------------+ | dbt3sf1 | nation | 1 | 0 | 0 | | dbt3sf1 | part | 2389 | 0 | 0 | | dbt3sf1 | partsupp | 9552 | 0 | 0 | | dbt3sf1 | supplier | 412 | 0 | 0 | +--------------+------------+-----------+--------------+-------------------------+ 4 rows in set (0.00 sec) Total = 12354 rows read + 9K temptable writes + 400 reads from temptable. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 1

[Maria-developers] Cost analysis: Materialization plan
by Sergei Petrunia 09 Aug '12

09 Aug '12

Debugging with optimizer_switch='materialization=on', query: explain extended select sql_calc_found_rows s_name, s_address from nation straight_join supplier where s_suppkey in (select ps_suppkey from partsupp where ps_partkey in (select p_partkey from part where p_name like 'forest%') and ps_availqty > (select 0.5 * sum(l_quantity) from lineitem where l_partkey = ps_partkey and l_suppkey = ps_suppkey and l_shipdate >= date('1994-01-01') and l_shipdate < date('1994-01-01') + interval '1' year )) and s_nationkey = n_nationkey and n_name = 'canada' order by s_name limit 10; +----+------------------+-----------+------++-------------------+-------+---------------------------------------+----+--------+--------------------------------------------+ |id |select_type |table |type ||key |key_len|ref |rows|filtered|Extra | +----+------------------+-----------+------++-------------------+-------+---------------------------------------+----+--------+--------------------------------------------+ | 1|PRIMARY |nation |ALL ||NULL |NULL |NULL | 25| 100.00|Using where; Using temporary; Using filesort| | 1|PRIMARY |supplier |ref ||i_s_nationkey |5 |nation.n_nationkey | 180| 100.00| | | 1|PRIMARY |<subquery2>|eq_ref||distinct_key |4 |func | 1| 100.00| | | 2|MATERIALIZED |part |range ||p_name |58 |NULL |2387| 100.00|Using where; Using index | | 2|MATERIALIZED |partsupp |ref ||PRIMARY |4 |part.p_partkey | 1| 100.00|Using where | | 4|DEPENDENT SUBQUERY|lineitem |ref ||i_l_suppkey_partkey|10 |partsupp.ps_partkey,partsupp.ps_suppkey| 4| 100.00|Using where | +----+------------------+-----------+------++-------------------+-------+---------------------------------------+----+--------+--------------------------------------------+ ********* SEMI-JOIN pre-optimization ************** ===== PRE-1. part === (gdb) p position->records_read $564 = 2387 (gdb) p position->read_time $565 = 497.33246183668518 (gdb) p current_record_count / TIME_FOR_COMPARE $566 = 477.39999999999998 (gdb) p current_read_time $567 = 974.73246183668516 ===== PRE-2. partsupp === (gdb) p position->records_read $570 = 1 (gdb) p position->read_time $571 = 2390.4263969377266 (gdb) p current_record_count / TIME_FOR_COMPARE $572 = 477.39999999999998 (gdb) p current_read_time $573 = 3842.5588587744119 ## (part, partsupp) is the best plan to materialize the subquery. (gdb) p join->best_read $575 = 6229.5578587744112 (gdb) p join->record_count $576 = 2387 ## Extra expenses for semi-join (gdb) print lookup_cost $582 = 0.050000000000000003 (gdb) print write_cost $584 = 0.050000000000000003 ********* MAIN join optimization ************** ===== 1. nation === (gdb) p position->records_read $587 = 25 (gdb) p position->read_time $588 = 1 (gdb) p current_record_count / TIME_FOR_COMPARE $589 = 5 (gdb) p current_read_time $590 = 6 ===== 2. supplier === (gdb) p position->records_read $593 = 180 (gdb) p position->read_time $594 = 4525 (gdb) p current_record_count / TIME_FOR_COMPARE $595 = 900 (gdb) p current_read_time $596 = 5431 ## here it considers "part, partsupp" extension, with huge cost: ## I don't skip it: === 3.A part === (gdb) p position->records_read $599 = 2387 (gdb) p position->read_time $600 = 2237996.0782650835 (gdb) p current_record_count / TIME_FOR_COMPARE $601 = 2148300 (gdb) p current_read_time $602 = 4391727.078265084 === 4.A partsupp == (gdb) p position->records_read $605 = 1 (gdb) p position->read_time $606 = 10741500 (gdb) p current_record_count / TIME_FOR_COMPARE $607 = 2148300 (gdb) p current_read_time $608 = 17281527.078265086 > advance_sj_state > Firstmatch_picker::check_qep ### Picks first match with (gdb) print read_time $615 = 12984927.078265084 (gdb) print rec_count $616 = 4500 < Firstmatch_picker::check_qep > Sj_materialization_picker::check_qep (gdb) p prefix_cost.total_cost() $622 = 5431 # OK, same as $596 (gdb) p mat_info->materialization_cost.total_cost() $623 = 3007.1088587744121 ## This one is weird. It should be $575 + $582 + $584= 6229.6578587744098 ## I will investigate. ## The difference is not millions, but a ~3K. (gdb) print prefix_rec_count $625 = 4500 (gdb) p mat_info->lookup_cost.total_cost() $626 = 0.050000000000000003 < Sj_materialization_picker::check_qep ## Materialization is chosen as the best strategy with (gdb) print current_record_count $634 = 4500 (gdb) print current_read_time $635 = 8663.1088587744125 < advance_sj_state === 5. Finally === We hit these lines: if (join->sort_by_table && join->sort_by_table != join->positions[join->const_tables].table->table) /* We have to make a temp table */ current_read_time+= current_record_count; and get the final query plan cost of (gdb) p join->record_count $643 = 4500 (gdb) p join->best_read $644 = 13163.107858774412 #records is the same for plan in this email and in the previous. The cost is very different, however, it is 13K vs 360K. Now, when we have the breakdown of the numbers, we can evaluate what did the optimizer get wrong. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 2

[Maria-developers] Cost analysis: FirstMatch plan
by Sergei Petrunia 09 Aug '12

09 Aug '12

Hi Timour, Please find below cost analysis for the plan with FirstMatch strategy. Debugging with optimizer_switch='semijoin=off', query: explain extended select sql_calc_found_rows s_name, s_address from nation straight_join supplier where s_suppkey in (select ps_suppkey from partsupp where ps_partkey in (select p_partkey from part where p_name like 'forest%') and ps_availqty > (select 0.5 * sum(l_quantity) from lineitem where l_partkey = ps_partkey and l_suppkey = ps_suppkey and l_shipdate >= date('1994-01-01') and l_shipdate < date('1994-01-01') + interval '1' year )) and s_nationkey = n_nationkey and n_name = 'CANADA' order by s_name limit 10; +----+------------------+--------+------++-------------------+-------+---------------------------------------+----+--------------------------------------------+ |id |select_type |table |type ||key |key_len|ef |rows|Extra | +----+------------------+--------+------++-------------------+-------+---------------------------------------+----+--------------------------------------------+ | 1|PRIMARY |nation |ALL ||NULL |NULL |NULL | 25|Using where; Using temporary; Using filesort| | 1|PRIMARY |supplier|ref ||i_s_nationkey |5 |nation.n_nationkey | 180| | | 1|PRIMARY |partsupp|ref ||i_ps_suppkey |4 |supplier.s_suppkey | 39|Using where | | 1|PRIMARY |part |eq_ref||PRIMARY |4 |partsupp.ps_partkey | 1|Using where; FirstMatch(supplier) | | 4|DEPENDENT SUBQUERY|lineitem|ref ||i_l_suppkey_partkey|10 |partsupp.ps_partkey,partsupp.ps_suppkey| 4|Using where | +----+------------------+--------+------++-------------------+-------+---------------------------------------+----+--------------------------------------------+ ===== 1. nation === (gdb) p position->records_read $367 = 25 (gdb) p position->read_time $368 = 1 (gdb) p current_record_count / TIME_FOR_COMPARE $369 = 5 (gdb) p current_read_time $370 = 6 ===== 2. supplier === (gdb) p position->records_read $377 = 180 (gdb) p position->read_time $378 = 4525 (gdb) p current_record_count / TIME_FOR_COMPARE $379 = 900 (gdb) p current_read_time $380 = 5431 ## here it considers "part, partsupp" extension, with cost ## 12,984,927. I skip that, because that's not the optimal plan ===== 3. partsupp === (gdb) p position->records_read $411 = 39 (gdb) p position->read_time $412 = 180000 (gdb) p current_record_count / TIME_FOR_COMPARE $413 = 35100 (gdb) p current_read_time $414 = 220531 ===== 4. part === (gdb) p position->records_read $421 = 1 (gdb) p position->read_time $422 = 175500 (gdb) p current_record_count / TIME_FOR_COMPARE $423 = 35100 (gdb) p current_read_time $424 = 431131 > advance_sj_state() > Firstmatch_picker::check_qep > optimize_wo_join_buffering ### It doesn't make any best_access_path() calls here, because no table is ### using join buffering. ### It forgets to add (current_record_count / TIME_FOR_COMPARE), and ends ### up with a slightly lower cost of (gdb) print cost $425 = 360931 > optimize_wo_join_buffering > Firstmatch_picker::check_qep ## FirstMatch is chosen as the way we will use to eliminate duplicates.. ## At the end of advance_sj_state, we have: (gdb) p *current_read_time $426 = 360931 (gdb) p *current_record_count $427 = 4500 > advance_sj_state(). And this is what is chosen as the final query plan. -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 2

[Maria-developers] Cost analysis: check if optimizer is able to produce a correct E(#scalar_subquery executions)
by Sergei Petrunia 09 Aug '12

09 Aug '12

Let's see if the optimizer is theoretically able to attach the subquery to the right table for the FirstMatch plan: On Thu, Aug 09, 2012 at 12:39:00PM +0400, Sergei Petrunia wrote: > == Modified query == ... > > == Benchmarking == > > Number of times scalar-subquery was executed: > Materialization: 9552 times (same as before) > First-Match: 299 times (A LOT LESS!) > > Execution times: > Materialization: 5.30 sec (same as before) > First-Match: 1.72 sec (a lot less, used to be 15 sec! Now it beats > materialization) > Let's check whether the optimizer has sufficient input data to conclude that moving the scalar-subquery from table partsupp to table part is advantageous. The data about extra selectivity is basically tab->quick_condition_rows. Let's take the EXPLAIN and its analysis from the "Cost analysis: optimizer statistics vs real dataset properties" email. Relevant lines will start with 'NOTE>'. +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ |id|select_type |table |type ||key |key_len|ref |rows|filtered|Extra | +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ | 1|PRIMARY |nation |ref ||n_name |26 |const | 1| 100.00|Using where; Using index; Using temporary; Using filesort| | 1|PRIMARY |supplier|ref ||i_s_nationkey |5 |nation.n_nationkey | 251| 100.00| | | 1|PRIMARY |partsupp|ref ||i_ps_suppkey |4 |supplier.s_suppkey | 34| 100.00|Using where | | 1|PRIMARY |part |eq_ref||PRIMARY |4 |partsupp.ps_partkey | 1| 100.00|Using where; FirstMatch(supplier) | | 4|DEP. SUBQUERY|lineitem|ref ||i_l_suppkey_partkey|10 |partsupp.ps_partkey,partsupp.ps_suppkey| 3| 100.00|Using where | +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ 5 rows in set, 3 warnings (0.01 sec) === nation: === like above, 1 row. OK. NOTE> - "using where" won't filter anything out. === supplier: === see above, 251 rows, ~ok 412 real... NOTE> - "using where" won't filter anything out. === partsupp: === - 800K rows and 10K distinct ps_suppkey, which gives rec_per_key=80 (EXPLAIN shows 34) - as for data that we will hit: select count(*) from nation, supplier, partsupp where s_suppkey=ps_suppkey and s_nationkey=n_nationkey and n_name='canada'; gives 32690 rows. 32690 / 412 = 80, matches rec_per_key. NOTE> "using where" won't filter anything out (NOT TAKING the scalar-subquery into account) === part: === - eq_ref, so exactly 1 match. We know DBT-3 dataset is such that it always has one. NOTE> "using where" is "p_name like forest"! it will filter stuff out! I haven't checked, but I suppose that condition "p_name like forest%'(*) is not correlated with the table access condition, 'p_partkey=partsupp_ps_partkey'. The email 'Cost analysis: Materialization plan' shows a range access on (*), with the estimate rows=2378. Total number of rows in table part is is 200K, InnoDB's estimate is 200999. This gives selectivity of 0.0118. === How many times scalar-subquery will be evaluated === Let's take estimate numbers: 1 * 251 * 34 * 1 * 0.0118 = 100.7012 evaluations. If the estimates were perfect: 1 * 400 * 80 * 1 * 0.118 = 377.6 evaluations As mentioned at the top of this email, in reality subquery is evaluated 299 times. === Conclusion#1 === quick_condition estimates provide information that will allow the optimizer to make a correct conclusion about # of times that scalar-subquery will be evaluated after table part. What remains to be checked: check that our cost formulas will make the plan of { FirstMatch, scalar-subquery attached to table part} cheaper than { FirstMatch, scalar-subquery attached to table partsupp} and cheaper than the SJ-Materialization plan. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] Cost analysis: checking how fast cost-based predicate pushdown can be
by Sergei Petrunia 09 Aug '12

09 Aug '12

... > ********* Some benchmarking ************* > Execution times: > - FirstMatch is ~15 sec > - Materialization: ~5 sec. > > With scalar-subquery predicate replaced with TRUE: > - Materialization: ~0.20 sec. > - FirstMatch: 0.19 sec. > > Number of times the scalar subquery was executed: > - DEP. SUBQUERY: ... 24055 times > > - MATERIALIZATION: 9555 times > > If we take the cost of scalar-subquery predicate evaluation into account, > the difference should be around: > > (24055 - 9555) * 5 = 72K > > which is sufficient. Then, we've tried to imitate the effects of "cost-based predicate pushdown" (mdev-83), by making the scalar subquery predicate depend on both tables part and partsupp. In order to do that, we've also had to convert the subquery (select p_partkey from part where p_name like 'forest%') into join. This change doesn't affect the optimizer because this subquery would be converted into an inner join anyway. == Modified query == explain select sql_calc_found_rows s_name, s_address from nation straight_join supplier where s_suppkey in (select ps_suppkey from partsupp, part where ps_partkey = p_partkey and p_name like 'forest%' and ps_availqty + p_retailprice*0 > (select 0.5 * sum(l_quantity) from lineitem where l_partkey = ps_partkey and l_suppkey = ps_suppkey and l_shipdate >= date('1994-01-01') and l_shipdate < date('1994-01-01') + interval '1' year )) and s_nationkey = n_nationkey and n_name = 'CANADA' order by s_name limit 10; == Benchmarking == Number of times scalar-subquery was executed: Materialization: 9552 times (same as before) First-Match: 299 times (A LOT LESS!) Execution times: Materialization: 5.30 sec (same as before) First-Match: 1.72 sec (a lot less, used to be 15 sec! Now it beats materialization) BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] what compiler should i use when building mariadb?
by Rich Prohaska 09 Aug '12

09 Aug '12

hello, we have been using gcc 4.4 to build mysql and mariadb successfully. however, when we tried to use gcc 4.7.1 to build mariadb-5.5.25, we hit this problem: /home/tokubuild/build-tokudb-46462/mariadb-5.5.25-tokudb-46462-linux-x86_64-build/mariadb-5.5.25-tokudb-46462-src/mysys/my_context.c: Assembler messages: /home/tokubuild/build-tokudb-46462/mariadb-5.5.25-tokudb-46462-linux-x86_64-build/mariadb-5.5.25-tokudb-46462-src/mysys/my_context.c:207: Error: CFI instruction used without previous .cfi_startproc what compiler do you use to build mariadb? thanks rich prohaska

4 9

[Maria-developers] Cost analysis: optimizer statistics vs real dataset properties
by Sergei Petrunia 08 Aug '12

08 Aug '12

So, we have done: alter table name add index(n_name); and now we have: explain extended select sql_calc_found_rows s_name, s_address from nation straight_join supplier where s_suppkey in (select ps_suppkey from partsupp where ps_partkey in (select p_partkey from part where p_name like 'forest%') and ps_availqty > (select 0.5 * sum(l_quantity) from lineitem where l_partkey = ps_partkey and l_suppkey = ps_suppkey and l_shipdate >= date('1994-01-01') and l_shipdate < date('1994-01-01') + interval '1' year )) and s_nationkey = n_nationkey and n_name = 'canada' order by s_name limit 10; +--+------------------+-----------+------++-------------------+-------+---------------------------------------+----+---------------------------------------------------------+ |id|select_type |table |type ||key |key_len|ref |rows|Extra | +--+------------------+-----------+------++-------------------+-------+---------------------------------------+----+---------------------------------------------------------+ | 1|PRIMARY |nation |ref ||n_name |26 |const | 1|Using where; Using index; Using temporary; Using filesort| | 1|PRIMARY |supplier |ref ||i_s_nationkey |5 |nation.n_nationkey | 251| | | 1|PRIMARY |<subquery2>|eq_ref||distinct_key |4 |func | 1| | | 2|MATERIALIZED |part |range ||p_name |58 |NULL |2387|Using where; Using index | | 2|MATERIALIZED |partsupp |ref ||PRIMARY |4 |part.p_partkey | 1|Using where | | 4|DEPENDENT SUBQUERY|lineitem |ref ||i_l_suppkey_partkey|10 |partsupp.ps_partkey,partsupp.ps_suppkey| 3|Using where | +--+------------------+-----------+------++-------------------+-------+---------------------------------------+----+---------------------------------------------------------+ === nation: === 1 row scanned, for real OK. - "using where" doesn't filter anything out. === supplier === - dataset has 412 rows for CANADA - rec_per_key: 10L suppliers / 25 countries=400 === <subquery2> == - eq_ref, we know that there's always match - "using where" won't filter anything out. === part === - dataset/where match 2378 rows for (p_name like 'forest%') OK. - "using where" won't filter anything out. === partsupp === - rec_per_key: partsupp has 800K rows and 200K distinct ps_partkey values, which gives rec_per_key=4 (while EXPLAIN shows 1!) - select count(*) from partsupp where ps_partkey in (select p_partkey from part where p_name like 'forest%'); gives 9552 rows. (rec_per_key=4) * 2378 = 9512 , close to 9552. - "using where" won't filter anything out (NOT TAKING scalar-subquery into account) +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ |id|select_type |table |type ||key |key_len|ref |rows|filtered|Extra | +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ | 1|PRIMARY |nation |ref ||n_name |26 |const | 1| 100.00|Using where; Using index; Using temporary; Using filesort| | 1|PRIMARY |supplier|ref ||i_s_nationkey |5 |nation.n_nationkey | 251| 100.00| | | 1|PRIMARY |partsupp|ref ||i_ps_suppkey |4 |supplier.s_suppkey | 34| 100.00|Using where | | 1|PRIMARY |part |eq_ref||PRIMARY |4 |partsupp.ps_partkey | 1| 100.00|Using where; FirstMatch(supplier) | | 4|DEP. SUBQUERY|lineitem|ref ||i_l_suppkey_partkey|10 |partsupp.ps_partkey,partsupp.ps_suppkey| 3| 100.00|Using where | +--+-------------+--------+------++-------------------+-------+---------------------------------------+----+--------+---------------------------------------------------------+ 5 rows in set, 3 warnings (0.01 sec) === nation: === like above, 1 row. OK. - "using where" won't filter anything out. === supplier: === see above, 251 rows, ~ok 412 real... === partsupp: === - 800K rows and 10K distinct ps_suppkey, which gives rec_per_key=80 (EXPLAIN shows 34) - as for data that we will hit: select count(*) from nation, supplier, partsupp where s_suppkey=ps_suppkey and s_nationkey=n_nationkey and n_name='canada'; gives 32690 rows. 32690 / 412 = 80, matches rec_per_key. - "using where" won't filter anything out (NOT TAKING scalar-subquery into account) === part: === - eq_ref, so exactly 1 match. We know DBT-3 dataset is such that it always has one. - "using where" is "p_name like forest"! it will filter stuff out! ^^^^^^^^ but does it matter.. subquery is attached to ps? ********* Some benchmarking ************* Execution times: - FirstMatch is ~15 sec - Materialization: ~5 sec. With scalar-subquery predicate replaced with TRUE: - Materialization: ~0.20 sec. - FirstMatch: 0.19 sec. Number of times the scalar subquery was executed: - DEP. SUBQUERY: ... 24055 times - MATERIALIZATION: 9555 times If we take the cost of scalar-subquery predicate evaluation into account, the difference should be around: (24055 - 9555) * 5 = 72K which is sufficient. The problem is, Materialization executes the scalar-subquery fewer times than FirstMatch. And you expect that taking into account scalar-subquery execution cost will cause Materialization not to be used?? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] Cost analysis: comparison of the two approaches
by Sergei Petrunia 08 Aug '12

08 Aug '12

So, we're comparing two plans that were mentioned earlier. FirstMatch plan has: (gdb) p *current_read_time $426 = 360931 (gdb) p *current_record_count $427 = 4500 which is much more greater than the SJ-Materialization plan, which has: (gdb) print current_record_count $634 = 4500 (gdb) print current_read_time $635 = 8663.1088587744125 Let's check whether the difference can be overcame by adding the cost of the scalar-subquery predicate. Cost of one execution of scalar subquery predicate is 4.5 ~= 5. SJ-Materialization plan will execute the subquery predicate 2378 times: (gdb) p join->best_read $575 = 6229.5578587744112 (gdb) p join->record_count $576 = 2387 which gives added cost of 2378 * 5 = 11890. 8663 + 11890 is still much less than 360931. Hence, just taking scalar-subquery cost into account is not sufficient for query plan to be changed. ps. psergey's opinion: too much unaccounted-for WHERE clauses. to be elaborated further... BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] Review request: SHOW EXPLAIN
by Sergei Petrunia 08 Aug '12

08 Aug '12

Hi Sergei, Please find attached the combined SHOW EXPLAIN patch. It is ready for review of the concurrency part. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

3 12

[Maria-developers] Streamlining Buildbot builds
by Daniel Bartholomew 08 Aug '12

08 Aug '12

Serg brought this to my attention and asked me to look into it. Basically, over time we keep adding more and more builders to buildbot and we build an ever increasing number of packages. One way to keep turnaround times for builds reasonable is to add more hardware. Which we are doing by adding a third KVM build host. However, building every push on every KVM virtual machine is overkill. For example, if one Ubuntu build succeeds, most other Ubuntu builds will also succeed. We only really need to build "everything" when we are making a release. For other builds we can get away with building on a sub-set of builders. The main benefit of doing this is to speed up the turn-around time for builds. One potential issue is the possibility of not catching an error which, for some reason, only manifests on a certain OS version. There may be other issues. Let me know if you think of any. Assuming we want to move ahead with building most builds on a sub-set of our builders instead of the full set, the easiest thing to do is to configure buildbot to only build on every builder for pushes into the "-release" and top-level (5.3, 5.5, etc...) branches. All other branches would have an abridged set of builders assigned to them. First things first, we need to decide which builders will build every push in every tree (aka "build everything") and which builders will only build releases. Builders on dedicated machines don't need to be changed, only KVM builders. Here are the KVM builders I propose to be in the "build everything" set: "kvm-bintar-precise-amd64", "kvm-bintar-precise-x86", "kvm-deb-debian6-amd64", "kvm-deb-debian6-x86", "kvm-deb-precise-amd64", "kvm-deb-precise-x86", "kvm-rpm-fedora17-amd64", "kvm-rpm-fedora17-x86", "kvm-rpm-centos6-amd64", "kvm-rpm-centos6-x86", The reasoning for choosing the above is to have an x86 and amd64 builder for Ubuntu, CentOS, Fedora, and Debian. And also to always build binary tarballs. Thoughts? Thanks. -- Daniel Bartholomew Google+ - http://gplus.to/dbart Twitter - http://twitter.com/daniel_bart MariaDB: An Enhanced Drop-in Replacement for MySQL Website - http://mariadb.org Twitter - http://twitter.com/mariadb Google+ - http://gplus.to/mariadb Facebook - http://fb.com/MariaDB.dbms Knowledgebase - http://kb.askmonty.org Monty Program - http://montyprogram.com

3 2

[Maria-developers] Mysql crashes with signal 11 after a select (or even explain) on a view with a subselect.
by kirchy＠gmx.de 03 Aug '12

03 Aug '12

>Description: we see some problems with views that include other views (I think that is the problem...) and selects on views with subselects. The log says after a crash: Aug 3 17:29:22 db7 mysqld: 120803 17:29:22 [ERROR] mysqld got signal 11 ; Aug 3 17:29:22 db7 mysqld: This could be because you hit a bug. It is also possible that this binary Aug 3 17:29:22 db7 mysqld: or one of the libraries it was linked against is corrupt, improperly built, Aug 3 17:29:22 db7 mysqld: or misconfigured. This error can also be caused by malfunctioning hardware. Aug 3 17:29:22 db7 mysqld: We will try our best to scrape up some info that will hopefully help diagnose Aug 3 17:29:22 db7 mysqld: the problem, but since we have already crashed, something is definitely wrong Aug 3 17:29:22 db7 mysqld: and this may fail. Aug 3 17:29:22 db7 mysqld: Aug 3 17:29:22 db7 mysqld: key_buffer_size=0 Aug 3 17:29:22 db7 mysqld: read_buffer_size=2097152 Aug 3 17:29:22 db7 mysqld: max_used_connections=100 Aug 3 17:29:22 db7 mysqld: max_threads=302 Aug 3 17:29:22 db7 mysqld: threads_connected=35 Aug 3 17:29:22 db7 mysqld: It is possible that mysqld could use up to Aug 3 17:29:22 db7 mysqld: key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 5570184 K Aug 3 17:29:22 db7 mysqld: bytes of memory Aug 3 17:29:22 db7 mysqld: Hope that's ok; if not, decrease some variables in the equation. Aug 3 17:29:22 db7 mysqld: Aug 3 17:29:22 db7 mysqld: Thread pointer: 0x7f3521cba000 Aug 3 17:29:22 db7 mysqld: Attempting backtrace. You can use the following information to find out Aug 3 17:29:22 db7 mysqld: where mysqld died. If you see no messages after this, something went Aug 3 17:29:22 db7 mysqld: terribly wrong... Aug 3 17:29:22 db7 mysqld: stack_bottom = 0x7f1f899b4e98 thread_stack 0x3c000 Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(my_print_stacktrace+0x2e) [0x7f341e1f39de] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(handle_segfault+0x3e5) [0x7f341de306f5] Aug 3 17:29:22 db7 mysqld: /lib/libpthread.so.0(+0xef60) [0x7f341c1def60] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(Field_dependency_recorder::visit_field(Field*)+0x1) [0x7f341dfb48c1] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(Item_field::enumerate_field_refs_processor(unsigned char*)+0x16) [0x7f341dd779d6] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(Item_subselect::enumerate_field_refs_processor(unsigned char*)+0x4c) [0x7f341ddeb6dc] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(Dep_analysis_context::setup_equality_modules_deps(List<Dep_module>*)+0x13e) [0x7f341dfb406e] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(+0x4e140d) [0x7f341dfb440d] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(+0x4e15a0) [0x7f341dfb45a0] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(+0x3d2eae) [0x7f341dea5eae] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(JOIN::optimize()+0x465) [0x7f341dea7a55] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(mysql_select(THD*, Item***, TABLE_LIST*, unsigned int, List<Item>&, Item*, unsigned int, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*)+0xda) [0x7f341deb04da] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(handle_select(THD*, st_lex*, select_result*, unsigned long)+0x15d) [0x7f341deb0f5d] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(+0x368baa) [0x7f341de3bbaa] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(mysql_execute_command(THD*)+0x40e) [0x7f341de3e05e] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(mysql_parse(THD*, char*, unsigned int, char const**)+0x2c1) [0x7f341de43711] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(dispatch_command(enum_server_command, THD*, char*, unsigned int)+0xceb) [0x7f341de4485b] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(do_command(THD*)+0xf4) [0x7f341de45364] Aug 3 17:29:22 db7 mysqld: /usr/sbin/mysqld(handle_one_connection+0x13e) [0x7f341de3762e] Aug 3 17:29:22 db7 mysqld: /lib/libpthread.so.0(+0x68ba) [0x7f341c1d68ba] Aug 3 17:29:22 db7 mysqld: /lib/libc.so.6(clone+0x6d) [0x7f341b7a002d] Aug 3 17:29:22 db7 mysqld: This happens on all servers over here. At least all I tested it on... even if these queries are too complex, the mysqld should never die with signal 11 but give some kind of sql error! >How-To-Repeat: I have some queries over here that crash the server. Is there something like a drop-in binary for mysqld I could use to gather more information for you? or shall I run mysqld with strace or so? the database is far to big and complex to copy it (besides legal reasons, I can't send it to anybody), but I have replication slaves I can use for testing. >Fix: just don't run the query... no idea. >Submitter-Id: <submitter ID> >Originator: root >Organization: kirchy >MySQL support: none >Synopsis: Mysql crashes with signal 11 after a select (or even explain) on a view with a subselect. >Severity: serious >Priority: medium >Category: mysql >Class: sw-bug >Release: mysql-5.2.7-MariaDB-mariadb101~squeeze ((MariaDB - http://mariadb.com/)) >Server: /usr/bin/mysqladmin Ver 9.0 Distrib 5.2.7-MariaDB, for debian-linux-gnu on x86_64 Copyright 2000-2008 MySQL AB, 2008 Sun Microsystems, Inc, 2009 Monty Program Ab This software comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to modify and redistribute it under the GPL license Server version 5.2.7-MariaDB-mariadb101~squeeze-log Protocol version 10 Connection Localhost via UNIX socket UNIX socket /var/run/mysqld/mysqld.sock Uptime: 57 min 5 sec Threads: 23 Questions: 491608 Slow queries: 175 Opens: 3750 Flush tables: 1 Open tables: 460 Queries per second avg: 143.535 >C compiler: gcc-4.4.real (Debian 4.4.4-8) 4.4.5 20100728 (prerelease) >C++ compiler: g++-4.4.real (Debian 4.4.4-8) 4.4.5 20100728 (prerelease) >Environment: System: Linux db7 2.6.39-bpo.2-amd64 #1 SMP Tue Jul 26 10:35:23 UTC 2011 x86_64 GNU/Linux Some paths: /usr/bin/perl /usr/bin/make /usr/bin/gcc /usr/bin/cc GCC: Using built-in specs. Target: x86_64-linux-gnu Configured with: ../src/configure -v --with-pkgversion='Debian 4.4.5-8' --with-bugurl=file:///usr/share/doc/gcc-4.4/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --program-suffix=-4.4 --enable-shared --enable-multiarch --enable-linker-build-id --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.4 --libdir=/usr/lib --enable-nls --enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc --with-arch-32=i586 --with-tune=generic --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu Thread model: posix gcc version 4.4.5 (Debian 4.4.5-8) Compilation info (call): CC='gcc -static-libgcc' CFLAGS='-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized' CXX='g++ -static-libgcc' CXXFLAGS='-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized' LDFLAGS='' ASFLAGS='' Compilation info (used): CC='gcc -static-libgcc' CFLAGS=' -O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized -DUNIV_LINUX' CXX='g++ -static-libgcc' CXXFLAGS=' -O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized -fno-implicit-templates -fno-exceptions -fno-rtti' LDFLAGS=' -rdynamic ' ASFLAGS='' LIBC: lrwxrwxrwx 1 root root 14 3. Mär 2011 /lib/libc.so.6 -> libc-2.11.2.so -rwxr-xr-x 1 root root 1432968 23. Jan 2011 /lib/libc-2.11.2.so -rw-r--r-- 1 root root 4433334 23. Jan 2011 /usr/lib/libc.a -rw-r--r-- 1 root root 247 23. Jan 2011 /usr/lib/libc.so Configure command: ../configure '--build=x86_64-linux-gnu' '--host=x86_64-linux-gnu' '--prefix=/usr' '--exec-prefix=/usr' '--libexecdir=/usr/sbin' '--datadir=/usr/share' '--localstatedir=/var/lib/mysql' '--includedir=/usr/include' '--infodir=/usr/share/info' '--mandir=/usr/share/man' '--with-server-suffix=-mariadb101~squeeze' '--with-comment=(MariaDB - http://mariadb.com/)' '--with-system-type=debian-linux-gnu' '--enable-shared' '--enable-static' '--enable-thread-safe-client' '--enable-assembler' '--enable-local-infile' '--with-pstack' '--with-fast-mutexes' '--with-unix-socket-path=/var/run/mysqld/mysqld.sock' '--with-mysqld-user=mysql' '--with-big-tables' '--without-docs' '--with-extra-charsets=all' '--with-libwrap' '--with-ssl=/usr' '--with-libevent' '--without-readline' '--with-zlib-dir=' '--with-embedded-server' '--with-embedded-privilege-control' '--with-plugins=max-no-ndb' '--without-plugin-innodb_plugin' 'build_alias=x86_64-linux-gnu' 'host_alias=x86_64-linux-gnu' 'CC =gcc -static-libgcc' 'CFLAGS=-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized' 'LDFLAGS=' 'CPPFLAGS=' 'CXX=g++ -static-libgcc' 'CXXFLAGS=-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized'

2 1

Re: [Maria-developers] [Commits] Rev 3353: Performed re-factoring and re-structuring of the code for mwl#248: in file:///home/igor/maria/maria-5.5-mwl248-changes/
by Kristian Nielsen 03 Aug '12

03 Aug '12

Davi Arnaut <davi(a)twitter.com> writes: > On Thu, Jul 26, 2012 at 5:50 PM, Igor Babaev <igor(a)askmonty.org> wrote: >> + bzero(column_stats, sizeof(Column_statistics) * cnt); > > That's not allowed for non-POD types. Can you elaborate on exactly when bzero() (or memset() or similar) is allowed or not allowed? I searched in my C++ reference, and did not really find anything ... in fact I did not find anything that said memset(p, 0, sizeof(*p)) would _ever_ be allowed for a class. It is my understanding that even for C, memset(..., 0, ...) is only valid for integer types. Initialising with memset() floating-point and even pointers values is not well-defined! Or did I miss something? I think this is an important issue, as memset() of a C++ class is certainly wrong in many cases, and can lead to subtle and platform-dependent failures. So let's point out clearly the rules that apply, or define such rules if necessary: 1. Does the C and C++ standards define suitable rules? If so, we should use them. However, I suspect that memset() is invalid in standard C and C++ except for arrays of integers, which is excessively restrictive. 2. If C/C++ standard is not sufficient, we need to agree on a coding style (and stay within what works on all reasonable compilers/platforms). I would much prefer to find an existing standards rule or convention. But failing that I would suggest: - Only memset() and memcpy() structs. - Virtual functions are definitely out (in struct as well as members). - Constructors and destructors also must not be defined for such structs (or its members). - Derived classes also should never be memset()'ed. Class inheritance is likely to need extra info on some (present or future) platforms and compilers. - Do not have private: methods in such memset() structs. I do not think it matters to the code generated, but private: is meaningless if arbitrary code is manipulating the memory of the struct directly. Usually I would have no methods at all in structs that I plan to do memset() on. But I suppose we also need to look at what existing code does. Comments? - Kristian.

2 1

Re: [Maria-developers] [Commits] Rev 3353: Performed re-factoring and re-structuring of the code for mwl#248: in file:///home/igor/maria/maria-5.5-mwl248-changes/
by Michael Widenius 31 Jul '12

31 Jul '12

Hi! >>>>> "Davi" == Davi Arnaut <davi(a)twitter.com> writes: Davi> On Mon, Jul 30, 2012 at 7:34 AM, Oleksandr Byelkin Davi> <sanja(a)montyprogram.com> wrote: >> 27.07.2012 20:57, Davi Arnaut пишет: >>> >>> Also, bzero is deprecated. >> AFAIK bzero is macro in our sources. Davi> It is a macro if not available. Which then becomes a problem because Davi> the bzero interface is not well defined (first argument begin char Davi> pointer versus void pointer). It should be pretty simple, let's not Davi> use deprecated and broken interfaces. Actually, in mariadb code (and original mysql code) we use bzero a lot and it's a stable, well tested, well known and well defined function. Better to always use the same code to reset memory than doing it many different ways. In MySQL code bzero() has been the offical and recommended way to reset memory. Regards, Monty

1 0

Re: [Maria-developers] [Commits] Rev 3453: MWL#182: Explain running statements
by Sergei Golubchik 28 Jul '12

28 Jul '12

Hi, Sergey! On Jul 17, Sergey Petrunya wrote: > ------------------------------------------------------------ > revno: 3453 > revision-id: psergey(a)askmonty.org-20120717192703-bgt3xgkbs8qvz3tc > parent: psergey(a)askmonty.org-20120717175208-ruovz91shg0ph875 > committer: Sergey Petrunya <psergey(a)askmonty.org> > branch nick: 5.5-show-explain-r21 > timestamp: Tue 2012-07-17 23:27:03 +0400 > message: > MWL#182: Explain running statements > Address feedback from the second code review: > - Make SHOW EXPLAIN code convert the query string before pushing it as a warning text. > @@ -2135,8 +2135,35 @@ int fill_show_explain(THD *thd, TABLE_LI > } > else > { > + /* > + Push the query string as a warning. The query may be in a different > + charset than the charset that's used for error messages, so, convert it > + if needed. > + */ there's no test case for this. was it too difficult to create? Regards, Sergei

1 0

[Maria-developers] question about MTR
by Rich Prohaska 26 Jul '12

26 Jul '12

Hello, It appears that MTR looks at a test and figures out the appropriate mysqld configuration needed to run the test. For example, if I have lines like "source include/have_XYZ.inc" in my test, MTR will run the test with the right storage engines loaded (and only those storage engines). Is this accurate? Thanks Rich Prohaska

2 1

[Maria-developers] Please test: BUG#992942 & MDEV-325 patch
by Sergei Petrunia 19 Jul '12

19 Jul '12

Hi Elena, I've made a patch that fixes both BUG#992942 and MDEV-325. It is rather risky, as it touches the part of the code whose interaction with other parts of the code is not very well defined. This is why I request RQG testing. Could you please run some test that would use ORDER BY ... LIMIT, GROUP BY, subqueries, tables that may or may not be read with range/index_merge access, index condition pushdown, etc? (Yes, you may really need all of the above to trigger the problem. Check out analysis comments in the bug). The code is at: bzr+ssh://bazaar.launchpad.net/~maria-captains/maria/5.5-show-explain-test1 it is based on 5.5-show-explain, so SHOW EXPLAIN part can be tested, too. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 1

[Maria-developers] Statistics tables show up in userstat - intentional?
by Sergei Petrunia 12 Jul '12

12 Jul '12

Hello Igor, I've noticed that engine-independent statistics show up in userstats: +--------------+-------------+-----------+--------------+-------------------------+ | Table_schema | Table_name | Rows_read | Rows_changed | Rows_changed_x_#indexes | +--------------+-------------+-----------+--------------+-------------------------+ | bug929732 | c2 | 249 | 0 | 0 | | mysql | index_stat | 84 | 0 | 0 | | bug929732 | v | 276 | 0 | 0 | | mysql | column_stat | 44 | 0 | 0 | | bug929732 | c | 249 | 0 | 0 | | bug929732 | v3 | 837280 | 0 | 0 | | mysql | table_stat | 12 | 0 | 0 | | bug929732 | v4 | 3994 | 0 | 0 | | bug929732 | v2 | 53506 | 0 | 0 | +--------------+-------------+-----------+--------------+-------------------------+ Is this intentional? I've also started collecting some minimal docs at http://kb.askmonty.org/en/engine-independent-table-statistics/ BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] 10.0-base tree created and is in buildbot
by Sergei Petrunia 11 Jul '12

11 Jul '12

Hello, As discussed before, I have created a 10.0-base tree at lp:~maria-captains/maria/10.0-base I've pushed there contents of 10.0-monty tree and have added the new tree to buildbot. lp:~maria-captains/maria/10.0 tree was deleted (it was created in error) BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] problem with savepoints in mariadb 5.5.25
by Rich Prohaska 11 Jul '12

11 Jul '12

Hello, Thanks. Per Elena's request, i created bug #1023404. Rich Prohaska On Wed, Jul 11, 2012 at 7:27 AM, Michael Widenius <michael.widenius(a)gmail.com> wrote: > > Hi! > >>>>>> "Rich" == Rich Prohaska <prohaska(a)tokutek.com> writes: > > Rich> Hello, > Rich> I posted our fix at lp:~prohaska7/maria/5.5-savepoint-offset-fix. > Rich> Does it make sense? > Rich> Thanks > Rich> Rich Prohaska > > Sergei, who wrote the handlerton interface, is just now on his way > home from his conference trip to China. He should be able to help you > with this issue as soon as he is back. > > Regards, > Monty

1 0

[Maria-developers] problem with savepoints in mariadb 5.5.25
by Rich Prohaska 10 Jul '12

10 Jul '12

Hello, We observed a problem with savepoints in Mariadb 5.5.25. The problem at the SQL layer is that a "rollback to savepoint" statement would get an error after the savepoint was created. The problem was caused by the savepoint name in the savepoint stack getting overwritten by the handlerton savepoint_set method. Here is what we think is going on: 1. space is reserved in the savepoint memory for each handlerton in the ha_initialize_handlerton function. 2. then space is reserved for the SAVEPOINT object in ha_init. (Yes, ha_init gets called AFTER the ha_initialize_handlerton calls. 3. ha_savepoint computes the address of the handlerton's object as (uchar*)(sv+1) + ht->savepoint_offset. this assumes that the SAVEPOINT object is laid out first, followed by the handlerton's space. unfortunately, step (1) occurs before step (2) so this is not the case. Here is a possible fix: Move the base initialization of the savepoint_alloc_size from ha_init to the global intiializer. With this change, the handlerton's offsets are computed AFTER the SAVEPOINT space is reserved. Is this correct? Thanks Rich Prohaska

2 3

[Maria-developers] Please test: Rev 3458: MDEV-376: Wrong result (missing rows) with index_merge+index_merge_intersection, join
by Sergei Petrunia 05 Jul '12

05 Jul '12

Hi Elena, Could you please re-run the tests with this fix: ----- Forwarded message from Sergey Petrunya <psergey(a)askmonty.org> ----- From: Sergey Petrunya <psergey(a)askmonty.org> To: commits(a)mariadb.org X-Mailer: mail (GNU Mailutils 1.2) Date: Wed, 4 Jul 2012 13:34:48 +0300 (EEST) Subject: [Commits] Rev 3458: MDEV-376: Wrong result (missing rows) with index_merge+index_merge_intersection, join in file:///home/psergey/dev2/5.5-look6/ At file:///home/psergey/dev2/5.5-look6/ ------------------------------------------------------------ revno: 3458 revision-id: psergey(a)askmonty.org-20120704103445-ro4sk3cwbbzn3t4p parent: psergey(a)askmonty.org-20120623161254-ido3pf16l31psyyv committer: Sergey Petrunya <psergey(a)askmonty.org> branch nick: 5.5-look6 timestamp: Wed 2012-07-04 14:34:45 +0400 message: MDEV-376: Wrong result (missing rows) with index_merge+index_merge_intersection, join - Let QUICK_RANGE_SELECT::init_ror_merged_scan() call quick->reset() only after we've set the column read bitmaps. === modified file 'mysql-test/r/index_merge_innodb.result' --- a/mysql-test/r/index_merge_innodb.result 2012-06-06 19:26:40 +0000 +++ b/mysql-test/r/index_merge_innodb.result 2012-07-04 10:34:45 +0000 @@ -777,4 +777,19 @@ commit; select * from t1 where t1.zone_id=830 AND modified=9; pk zone_id modified drop table t0, t1; +# +# MDEV-376: Wrong result (missing rows) with index_merge+index_merge_intersection, join +# +CREATE TABLE t1 ( +a INT, b CHAR(1), c CHAR(1), KEY(a), KEY(b) +) ENGINE=InnoDB; +INSERT INTO t1 VALUES (8,'v','v'),(8,'m','m'),(9,'d','d'); +SELECT ta.* FROM t1 AS ta, t1 AS tb +WHERE ( tb.b != ta.b OR tb.a = ta.a ) +AND ( tb.b = ta.c OR tb.b = ta.b ); +a b c +8 v v +8 m m +9 d d +DROP TABLE t1; set optimizer_switch= @optimizer_switch_save; === modified file 'mysql-test/t/index_merge_innodb.test' --- a/mysql-test/t/index_merge_innodb.test 2012-06-04 15:26:11 +0000 +++ b/mysql-test/t/index_merge_innodb.test 2012-07-04 10:34:45 +0000 @@ -156,6 +156,21 @@ select * from t1 where t1.zone_id=830 AN drop table t0, t1; +--echo # +--echo # MDEV-376: Wrong result (missing rows) with index_merge+index_merge_intersection, join +--echo # +CREATE TABLE t1 ( + a INT, b CHAR(1), c CHAR(1), KEY(a), KEY(b) +) ENGINE=InnoDB; + +INSERT INTO t1 VALUES (8,'v','v'),(8,'m','m'),(9,'d','d'); + + +SELECT ta.* FROM t1 AS ta, t1 AS tb +WHERE ( tb.b != ta.b OR tb.a = ta.a ) + AND ( tb.b = ta.c OR tb.b = ta.b ); + +DROP TABLE t1; set optimizer_switch= @optimizer_switch_save; === modified file 'sql/opt_range.cc' --- a/sql/opt_range.cc 2012-06-04 15:26:11 +0000 +++ b/sql/opt_range.cc 2012-07-04 10:34:45 +0000 @@ -2008,7 +2008,7 @@ int QUICK_RANGE_SELECT::init_ror_merged_ if (reuse_handler) { DBUG_PRINT("info", ("Reusing handler 0x%lx", (long) file)); - if (init() || reset()) + if (init()) { DBUG_RETURN(1); } @@ -2043,7 +2043,7 @@ int QUICK_RANGE_SELECT::init_ror_merged_ if (file->ha_external_lock(thd, F_RDLCK)) goto failure; - if (init() || reset()) + if (init()) { file->ha_external_lock(thd, F_UNLCK); file->ha_close(); @@ -2090,7 +2090,18 @@ int QUICK_RANGE_SELECT::init_ror_merged_ head->key_read= org_key_read; bitmap_copy(&column_bitmap, head->read_set); head->column_bitmaps_set(&column_bitmap, &column_bitmap); - + + if (reset()) + { + if (!reuse_handler) + { + file->ha_external_lock(thd, F_UNLCK); + file->ha_close(); + goto failure; + } + else + DBUG_RETURN(1); + } DBUG_RETURN(0); failure: _______________________________________________ commits mailing list commits(a)mariadb.org https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits ----- End forwarded message ----- -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] Fixing 10.0 buildbot: OS-dependent error messages
by Sergei Petrunia 04 Jul '12

04 Jul '12

Hi! As some have noticed, 10.0 tree is red for all non-Linux builds. The errors look like this (on Windows) http://buildbot.askmonty.org/buildbot/builders/win64-release2/builds/1558/s… -ERROR HY000: Got error 1478 "Unknown error 1478" from storage engine +ERROR HY000: Got error 1478 "Unknown error" from storage engine ... -Error 1005 Can't create table 'test.t1' (errno: 1478 "Unknown error 1478") +Error 1005 Can't create table 'test.t1' (errno: 1478 "Unknown error") or like this (on Mac) http://buildbot.askmonty.org/buildbot/builders/labrador/builds/95/steps/tes… -ERROR HY000: Got error 1478 "Unknown error 1478" from storage engine +ERROR HY000: Got error 1478 "Unknown error: 1478" from storage engine ... -ERROR HY000: Out of resources when opening file '<partition file>' (Errcode: 24 - Too many open files) +ERROR HY000: Out of resources when opening file '<partition file>' (Errcode: 24 "Too many open files") I've narrowed it down. It is caused by revision 3414 monty(a)askmonty.org-20120529213755-876ptdhhaj0t7l8r. The revision adds %M modifier to error error messages, which prints "error number - system-error-text" into the error message (see strings/my_vsnprintf.c, my_strerror()). The problem is that system-error-text depends on the OS one is running. grep '%M' sql/share/errmsg-utf8.txt | grep 'eng ' | wc -l 27 There are currently 27 errors that use %M modifier in the error message The testsuite's .result files have only 5 of the errors, but there are lots of occurrences. Buildbot can get at least 3 variants of error message (Windows, Mac, Linux). I suppose, --replace_result will be too complex and fragile. We need to solve it somewhere else. My first idea was to: make mysql-test-run recognize error messages that have the %M pattern and remove the custom text, leaving the error code. To keep things simple/loosely coupled, I'd base recognition of error messages on list of patterns. I'm not sure where to keep the patterns and where to do the matching: should it be done inside mysql-test-run, or inside mysqltest.cc, or I should introduce some commands into mysqltest.cc that will make it apply rewrite rules to all error messages. Any ideas? -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] replication tests crash mysqld in mariadb 5.5.25
by Rich Prohaska 28 Jun '12

28 Jun '12

with mariadb-5.5.25 on a centos 5.8 machine, when i run this: ./mysql-test-run.pl --suite=rpl --force --max-test-fail=0 --nowarnings --parallel=4 mysqld crashes in the xtradb code trying to deref a NULL pointer. core files are available. Core was generated by `/home/prohaska/distros/mariadb-5.5.25-linux-x86_64/bin/mysqld --defaults-group-'. Program terminated with signal 11, Segmentation fault. #0 0x000000000095aa15 in os_aio_get_array_and_local_segment (global_seg=4, message1=0x43b910d8, message2=0x43b910d0, type=0x43b910c8, space_id=0x43b910c0) at /home/buildbot/buildbot/build/storage/xtradb/os/os0file.c:3787 3787 /home/buildbot/buildbot/build/storage/xtradb/os/os0file.c: No such file or directory. in /home/buildbot/buildbot/build/storage/xtradb/os/os0file.c (gdb) bt #0 0x000000000095aa15 in os_aio_get_array_and_local_segment (global_seg=4, message1=0x43b910d8, message2=0x43b910d0, type=0x43b910c8, space_id=0x43b910c0) at /home/buildbot/buildbot/build/storage/xtradb/os/os0file.c:3787 #1 os_aio_linux_handle (global_seg=4, message1=0x43b910d8, message2=0x43b910d0, type=0x43b910c8, space_id=0x43b910c0) at /home/buildbot/buildbot/build/storage/xtradb/os/os0file.c:4671 #2 0x00000000009114d7 in fil_aio_wait (segment=4) at /home/buildbot/buildbot/build/storage/xtradb/fil/fil0fil.c:5530 #3 0x000000000088d1ee in io_handler_thread (arg=<value optimized out>) at /home/buildbot/buildbot/build/storage/xtradb/srv/srv0start.c:483 #4 0x000000310e60677d in start_thread () from /lib64/libpthread.so.0 #5 0x000000310dad325d in clone () from /lib64/libc.so.6 (gdb) p os_aio_read_array $1 = (os_aio_array_t *) 0x0

2 1

[Maria-developers] mysql engines tests fail on mariadb 5.5.25
by Rich Prohaska 26 Jun '12

26 Jun '12

hello, when i run "./mysql-test-run.pl --suite=engines/funcs,engines/iuds --force --mysqld=--partition --nowarnings --max-test-fail=0", i see lots of test failures on mariadb-5.5.25 (and other distributions). either (1) i am not running the tests correctly, or (2) the test and/or test result files are broken, or (3) there are real problems being detected. i suspect most of the failures are caused by (2) above. anyway, here are the failures: Completed: Failed 37/323 tests, 88.54% were successful. Failing test(s): engines/funcs.an_calendar engines/funcs.datetime_function engines/funcs.db_alter_collate_ascii engines/funcs.db_alter_collate_utf8 engines/funcs.db_use_error engines/funcs.de_calendar_range engines/funcs.in_calendar_2_unique_constraints_duplicate_update engines/funcs.in_calendar_pk_constraint_duplicate_update engines/funcs.in_calendar_pk_constraint_error engines/funcs.in_calendar_pk_constraint_ignore engines/funcs.in_calendar_unique_constraint_duplicate_update engines/funcs.in_calendar_unique_constraint_error engines/funcs.in_calendar_unique_constraint_ignore engines/funcs.in_multicolumn_calendar_pk_constraint_duplicate_update engines/funcs.in_multicolumn_calendar_pk_constraint_error engines/funcs.in_multicolumn_calendar_pk_constraint_ignore engines/funcs.in_multicolumn_calendar_unique_constraint_duplicate_update engines/funcs.in_multicolumn_calendar_unique_constraint_error engines/funcs.in_multicolumn_calendar_unique_constraint_ignore engines/funcs.in_multicolumn_number_pk_constraint_duplicate_update engines/funcs.in_number_boundary_error engines/funcs.in_number_decimal_boundary_error engines/funcs.ld_all_number_string_calendar_types engines/funcs.se_join_left engines/funcs.se_join_left_outer engines/funcs.se_join_natural_left engines/funcs.se_join_natural_left_outer engines/funcs.sq_all engines/funcs.sq_error engines/funcs.up_calendar_range engines/iuds.delete_time engines/iuds.insert_decimal engines/iuds.insert_number engines/iuds.insert_time engines/iuds.type_bit_iuds engines/iuds.update_delete_number engines/iuds.update_time thanks rich prohaska

3 2

Re: [Maria-developers] Add back testcase for lp:817966 (was lost in the merge)
by Sergei Golubchik 23 Jun '12

23 Jun '12

Hi, Sergey! On Jun 23, Sergey Petrunya wrote: > At file:///home/psergey/dev2/5.5-look6/ > > ------------------------------------------------------------ > revno: 3457 > revision-id: psergey(a)askmonty.org-20120623161254-ido3pf16l31psyyv > parent: knielsen(a)knielsen-hq.org-20120622084255-u53u21xv3w7sqxms > committer: Sergey Petrunya <psergey(a)askmonty.org> > branch nick: 5.5-look6 > timestamp: Sat 2012-06-23 20:12:54 +0400 > message: > Add back testcase for lp:817966 (was lost in the merge) > > === modified file 'mysql-test/t/func_in.test' > --- a/mysql-test/t/func_in.test 2012-06-20 11:01:28 +0000 > +++ b/mysql-test/t/func_in.test 2012-06-23 16:12:54 +0000 > @@ -577,6 +577,24 @@ > --echo # > --echo End of 5.1 tests > > +# > +# lp:817966 int_column IN (string_constant) > +# > +# rather illogically, when BIGINT field is compared to a string, > +# the string is converted to an integer, not to a double. > +# When some other integer field (not BIGINT) is compared to a string, > +# or when the BIGINT is not a field, but an expression, both > +# operands are compared as doubles. The latter behavior is correct, > +# according to the manual. I don't think this comment holds anymore, perhaps it'd better to remove it. > +# > +create table t1 (a bigint, b int); > +insert t1 values (1,1),(2,2),(3,3); > +select * from t1 where a in ('2.1'); > +select * from t1 where b in ('2.1'); > +select * from t1 where a='2.1'; > +select * from t1 where b='2.1'; > +select * from t1 where IF(1,a,a)='2.1'; > +drop table t1; > --echo # > --echo # LP bug#992380 Crash when creating PS for a query with > --echo # subquery in WHERE (see also mysql bug#13012483) > Regards, Sergei

1 0

[Maria-developers] lp:817966 causes wrong results when added into 5.5
by Sergei Petrunia 22 Jun '12

22 Jun '12

Hi Sergei, Remember I've pointed out that the testcase for lp:817966 is present in 5.3 but is missing in 5.5, and we figured it was lost in the merge and decided to add it? Well, if I add the testcase into 5.5 it fails like this: main.func_in [ fail ] Test ended at 2012-06-20 15:11:27 CURRENT_TEST: main.func_in --- /home/psergey/dev2/5.5-look6/mysql-test/r/func_in.result 2012-06-20 15:11:12.000000000 +0400 +++ /home/psergey/dev2/5.5-look6/mysql-test/r/func_in.reject 2012-06-20 15:11:27.000000000 +0400 @@ -793,12 +793,10 @@ insert t1 values (1,1),(2,2),(3,3); select * from t1 where a in ('2.1'); a b -2 2 select * from t1 where b in ('2.1'); a b select * from t1 where a='2.1'; a b -2 2 select * from t1 where b='2.1'; a b select * from t1 where IF(1,a,a)='2.1'; I cannot readily tell whether this is a bug or not (could be, the typecasting rules are changed in 5.5?) Do you think this should be filed as a jira issue? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] MariaDB 5.5.25 release prep
by Daniel Bartholomew 21 Jun '12

21 Jun '12

All, I'm working on getting the bits and pieces ready for this release. Preliminary versions of the Release Notes and Changelog are available here: http://kb.askmonty.org/en/mariadb-5525-release-notes/ http://kb.askmonty.org/en/mariadb-5525-changelog/ As always, if there is anything which should be mentioned in the Release Notes, please let me know ASAP or jump in and edit the page directly. Thanks! -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

Re: [Maria-developers] MDEV-342: Do not mark old binlog file as cleanly closed during rotate until
by Sergei Golubchik 21 Jun '12

21 Jun '12

Hi, Kristian! Look good. Just one comment, see below, and ok to push On Jun 20, Kristian Nielsen wrote: > > ------------------------------------------------------------ > revno: 3449 > revision-id: knielsen(a)knielsen-hq.org-20120619130327-7vvemizq0akoysvn > parent: knielsen(a)knielsen-hq.org-20120620122323-bimn76haijiwpo7u > committer: knielsen(a)knielsen-hq.org > branch nick: mariadb-5.5 > timestamp: Tue 2012-06-19 15:03:27 +0200 > message: > MDEV-342: Do not mark old binlog file as cleanly closed during rotate until > the new file is fully synced to disk and binlog index. This fixes a window > where a crash would leave next server restart unable to detect that a crash > occured, causing recovery to fail. > === modified file 'sql/log.cc' > --- a/sql/log.cc 2012-05-21 18:54:41 +0000 > +++ b/sql/log.cc 2012-06-19 13:03:27 +0000 > @@ -6344,6 +6373,21 @@ void MYSQL_BIN_LOG::close(uint exiting) > } > > > +/* > + For binlog rotation, this function is used to ensure that the old binlog > + file is not marked as cleanly closed until the new binlog file is synced to > + disk and index. This is needed to ensure that we will always detect a prior > + crash at next server restart. > +*/ > +void MYSQL_BIN_LOG::delayed_close(File file) > +{ > + my_off_t offset= BIN_LOG_HEADER_SIZE + FLAGS_OFFSET; > + uchar flags= 0; // clearing LOG_EVENT_BINLOG_IN_USE_F > + mysql_file_pwrite(file, &flags, 1, offset, MYF(0)); I'd rather reuse this piece of code that overwrites the flag. If you move mysql_file_close() to the caller, you can call this function set_binlog_flag_when_closing() and use it both for normal and delayed close. > + mysql_file_close(file, MYF(MY_WME)); > +} > + Regards, Sergei

1 0

[Maria-developers] Some input re MDEV-83
by Sergei Petrunia 21 Jun '12

21 Jun '12

Hi Timour, Please find below some input re MDEV-83: * I expect that MDEV-83 will have an @@optimizer_switch flag. Is that so? * MDEV-83 will change optimizer behavior, however the change will not be visible in EXPLAIN. We have a history of such attempts and we've always ended up having to fix the EXPLAIN so that the change is visible. Do you have any ideas how to do it? My proposal is to show "subquery#n" in the Extra column of the table to which the subquery predicate is attached. * You need to take join buffering into account. For example, for this query: explain select * from ten A, ten B where cond(B.col) and cond(A.col, B.col); +--+-----------+-----+----+-------------+----+-------+----+----+------------------------------------------------+ |id|select_type|table|type|possible_keys|key |key_len|ref |rows|Extra | +--+-----------+-----+----+-------------+----+-------+----+----+------------------------------------------------+ | 1|SIMPLE |A |ALL |NULL |NULL|NULL |NULL| 10| | | 1|SIMPLE |B |ALL |NULL |NULL|NULL |NULL| 10|Using where; Using join buffer (flat, BNL join) | +--+-----------+-----+----+-------------+----+-------+----+----+------------------------------------------------+ both cond(B.col) and cond(A.col, B.col) will be attached to table B, however they will have different cardinalities. * The same goes for semi-join strategies. It is possible to calculate output cardinality for any condition, however it is much less obvious. * Suppose you have a query: select * from ot1, ot2 where expr in (select it1.col1 from it1, it2 where subquery_cond(it2.col2)) suppose the join order is: SJ-Materialize-scan(it1, it2), ot1, ot2 Here, subquery_cond may only be attached inside the SJ-Materialization nest. This is because subquery cond relies on the value of it2.col2, which is not available outside the SJ-Materialization nest. * LEFT JOINs: when the query has outer joins, predicates that are attached to tables that are inner w.r.t some LEFT JOIN, are wrapped in Item_func_trig_cond. I suppose you'll reuse the wrapping code, so that repositioned conditions are wrapped into appropriate Item_func_trig_conds. I'm not sure whether we should support predicate repositioning for parts of ON expressions. I suppose not implementing this allows to deliver sooner. If this is the case, we should not support repositioning inside ON expressions in the first milestone. * Section 1.2 says: " An estimate is made of the partial result cardinality, such that it takes into account actual pushed predicate selectivity" As far as I understand, you only have ability to estimate selectivity of predicates in form "tbl1.col CMP const" For such conditions, it is known in advance which table they will be attached to. There is no need to traverse the "cheap" pushed-down conditions. Then, what exactly do you mean by "take into account actual pushed predicate selectivity"? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] Please review: [Commits] Rev 3538: Fix bug lp:1008773 in file:///home/tsk/mprog/src/5.3-lpb1008773/
by Sergei Petrunia 18 Jun '12

18 Jun '12

Hi Timour, Ok to push. On Tue, Jun 05, 2012 at 04:20:35PM +0300, Timour Katchaounov wrote: > Sergey, > > Could you please review yet another rather trivial fix. > The only thing here is whether the semantics is correct. > As the bug report says, all MariaDB/MySQL versions (except > mariadb 5.3/5.5), and Postgresql return the value of the > constant as part of the empty row produced when we have > implicit grouping. I didn't verify this in the SQL standard, > because AFAIR this is non-standard. > > Timour > > ------------------------------------------------------------ > revno: 3538 > revision-id: timour(a)askmonty.org-20120605124124-akdmgdmr01h10s2w > parent: psergey(a)askmonty.org-20120602121305-mhw0u2zppjk0erhk > fixes bug(s): https://launchpad.net/bugs/1008773 > committer: timour(a)askmonty.org > branch nick: 5.3-lpb1008773 > timestamp: Tue 2012-06-05 15:41:24 +0300 > message: > Fix bug lp:1008773 > > Analysis: > Queries with implicit grouping (there is aggregate, but no group by) > follow some non-obvious semantics in the case of empty result set. > Aggregate functions produce some special "natural" value depending on > the function. For instance MIN/MAX return NULL, COUNT returns 0. > > The complexity comes from non-aggregate expressions in the select list. > If the non-aggregate expression is a constant, it can be computed, so > we should return its value, however if the expression is non-constant, > and depends on columns from the empty result set, then the only meaningful > value is NULL. > > The cause of the wrong result was that for subqueries the optimizer didn't > make a difference between constant and non-constant ones in the case of > empty result for implicit grouping. > > Solution: > In all implementations of Item_subselect::no_rows_in_result() check if the > subquery predicate is constant. If it is constant, do not set it to the > default value for implicit grouping, instead let it be evaluated. > === modified file 'mysql-test/r/subselect4.result' > --- a/mysql-test/r/subselect4.result 2012-02-28 14:41:55 +0000 > +++ b/mysql-test/r/subselect4.result 2012-06-05 12:41:24 +0000 > @@ -2124,6 +2124,61 @@ ERROR 21000: Subquery returns more than > DROP TABLE t2; > ERROR 42S02: Unknown table 't2' > DROP TABLE t1; > +# > +# LP BUG#1008773 Wrong result (NULL instead of a value) with no matching rows, subquery in FROM and HAVING > +# > +CREATE TABLE t1 (a INT) ENGINE=MyISAM; > +CREATE TABLE t2 (b INT) ENGINE=MyISAM; > +INSERT INTO t2 VALUES (1); > +EXPLAIN > +SELECT MAX(a), ( SELECT 1 FROM t2 ) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 1 > +SELECT MAX(a), ( SELECT 1 FROM t2 ) AS bb FROM t1; > +MAX(a) bb > +NULL 1 > +EXPLAIN > +SELECT MAX(a), 1 in ( SELECT b FROM t2 ) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 1 > +SELECT MAX(a), 1 in ( SELECT b FROM t2 ) AS bb FROM t1; > +MAX(a) bb > +NULL 1 > +EXPLAIN > +SELECT MAX(a), 1 >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +SELECT MAX(a), 1 >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +MAX(a) bb > +NULL 1 > +EXPLAIN > +SELECT MAX(a), ( SELECT 1 FROM t2 where b = a) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +SELECT MAX(a), ( SELECT 1 FROM t2 where b = a) AS bb FROM t1; > +MAX(a) bb > +NULL NULL > +EXPLAIN > +SELECT MAX(a), a in ( SELECT b FROM t2 ) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 1 > +SELECT MAX(a), a in ( SELECT b FROM t2 ) AS bb FROM t1; > +MAX(a) bb > +NULL NULL > +EXPLAIN > +SELECT MAX(a), a >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 system NULL NULL NULL NULL 0 const row not found > +2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 1 > +SELECT MAX(a), a >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +MAX(a) bb > +NULL NULL > +drop table t1, t2; > set optimizer_switch=@subselect4_tmp; > SET optimizer_switch= @@global.optimizer_switch; > set @@tmp_table_size= @@global.tmp_table_size; > > === modified file 'mysql-test/t/subselect4.test' > --- a/mysql-test/t/subselect4.test 2012-02-28 14:41:55 +0000 > +++ b/mysql-test/t/subselect4.test 2012-06-05 12:41:24 +0000 > @@ -1761,6 +1761,40 @@ CREATE TABLE t2 AS > DROP TABLE t2; > DROP TABLE t1; > > +--echo # > +--echo # LP BUG#1008773 Wrong result (NULL instead of a value) with no matching rows, subquery in FROM and HAVING > +--echo # > + > +CREATE TABLE t1 (a INT) ENGINE=MyISAM; > +CREATE TABLE t2 (b INT) ENGINE=MyISAM; > +INSERT INTO t2 VALUES (1); > + > +EXPLAIN > +SELECT MAX(a), ( SELECT 1 FROM t2 ) AS bb FROM t1; > +SELECT MAX(a), ( SELECT 1 FROM t2 ) AS bb FROM t1; > + > +EXPLAIN > +SELECT MAX(a), 1 in ( SELECT b FROM t2 ) AS bb FROM t1; > +SELECT MAX(a), 1 in ( SELECT b FROM t2 ) AS bb FROM t1; > + > +EXPLAIN > +SELECT MAX(a), 1 >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +SELECT MAX(a), 1 >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > + > + > +EXPLAIN > +SELECT MAX(a), ( SELECT 1 FROM t2 where b = a) AS bb FROM t1; > +SELECT MAX(a), ( SELECT 1 FROM t2 where b = a) AS bb FROM t1; > + > +EXPLAIN > +SELECT MAX(a), a in ( SELECT b FROM t2 ) AS bb FROM t1; > +SELECT MAX(a), a in ( SELECT b FROM t2 ) AS bb FROM t1; > + > +EXPLAIN > +SELECT MAX(a), a >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > +SELECT MAX(a), a >= ALL ( SELECT b FROM t2 ) AS bb FROM t1; > + > +drop table t1, t2; > > set optimizer_switch=@subselect4_tmp; > > > === modified file 'sql/item_subselect.cc' > --- a/sql/item_subselect.cc 2012-05-22 12:22:55 +0000 > +++ b/sql/item_subselect.cc 2012-06-05 12:41:24 +0000 > @@ -892,6 +892,8 @@ void Item_maxmin_subselect::print(String > > void Item_maxmin_subselect::no_rows_in_result() > { > + if (const_item()) > + return; > value= Item_cache::get_cache(new Item_null()); > null_value= 0; > was_values= 0; > @@ -901,6 +903,8 @@ void Item_maxmin_subselect::no_rows_in_r > > void Item_singlerow_subselect::no_rows_in_result() > { > + if (const_item()) > + return; > value= Item_cache::get_cache(new Item_null()); > reset(); > make_const(); > @@ -1363,6 +1367,8 @@ Item* Item_exists_subselect::expr_cache_ > > void Item_exists_subselect::no_rows_in_result() > { > + if (const_item()) > + return; > value= 0; > null_value= 0; > make_const(); > @@ -2707,6 +2713,8 @@ void Item_allany_subselect::print(String > > void Item_allany_subselect::no_rows_in_result() > { > + if (const_item()) > + return; > value= 0; > null_value= 0; > was_null= 0; > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

3 2

Re: [Maria-developers] [Commits] Rev 3434: MDEV-339, LP1001340 - system_time_zone is wrong on Windows in file:///H:/bzr/5.5/
by Sergei Golubchik 15 Jun '12

15 Jun '12

Hi, Vladislav! On Jun 15, Vladislav Vaintroub wrote: > ------------------------------------------------------------ > revno: 3434 > revision-id: wlad(a)montyprogram.com-20120615152106-p4a50d0yv2qj4fxk > parent: wlad(a)montyprogram.com-20120613093751-tv67fxuh1qcxm0by > fixes bug: https://launchpad.net/bugs/1001340 > committer: Vladislav Vaintroub <wlad(a)montyprogram.com> > branch nick: 5.5 > timestamp: Fri 2012-06-15 17:21:06 +0200 > message: > MDEV-339, LP1001340 - system_time_zone is wrong on Windows > > On localized Windows versions, Windows uses localized time zone > names and contain non-ASCII characters. non-ASCII characters appear > broken when displayed by clients > The fix is to declare system_time_zone variable to have UTF8 > encoding and to convert tzname to UTF8. Looks good, thanks! Did you verify that it works? (I vaguely remember that you don't have a localized Windows version) Ok to push, if yes. Regards, Sergei

1 0

Re: [Maria-developers] MDEV-339 lp:1001340 system_time_zone is wrong on Windows
by Sergei Golubchik 14 Jun '12

14 Jun '12

Hi, Vladislav! On Jun 14, Vladislav Vaintroub wrote: > At file:///H:/bzr/5.5/ > > revno: 3434 > revision-id: wlad(a)montyprogram.com-20120614122958-bq20w9xbqjw032nk > parent: wlad(a)montyprogram.com-20120613093751-tv67fxuh1qcxm0by > fixes bug: https://launchpad.net/bugs/1001340 > committer: Vladislav Vaintroub <wlad(a)montyprogram.com> > branch nick: 5.5 > timestamp: Thu 2012-06-14 14:29:58 +0200 > message: > MDEV-339, LP1001340 - system_time_zone is wrong on Windows > > On localized Windows versions, Windows uses localized time zone > names. Since characters in timezone name can be outside of ASCII > range, thus not every client would be able to read variable without > loosing information, and even server misinterprets the encoding (it > does not try to interpret the encoding). > > The fix is to use the UTC offset format "+/-hh:mm" for > system_time_zone on Windows, compatible with time_zone values > everywhere else. UTC offset format is however not used if TZ > environment variable is defined - MySQL documents to handle TZ > environment variable in OS-dependent way. I am not sure I like this idea. UTC offset format specifies a *different* time zone. Not the system one. Timezone that surely produces different results for the historical dates. Original bug report says that system_time_zone value is not a valid MySQL (or POSIX) time zone name. But is it a problem? When time_zone is set to SYSTEM, MariaDB will use OS functions for date/time conversion, and not the data from timezone tables. So it will use the system time zone *exactly*, not UTC offset approximation. And it won't try to look up system_time_zone value in the timezone tables. So, I think, it's quite ok for system_time_zone to have "Westeuropäische Sommerzeit" value. You only need to fix the charset/truncation issue and that's all. Regards, Sergei

3 2

Re: [Maria-developers] [Commits] Rev 3543: Fix bug lp:1008686 in file:///home/tsk/mprog/src/5.3/
by Sergei Petrunia 14 Jun '12

14 Jun '12

Hi Timour, How does this relate to fix for lp:1008773? In this bug, you've added: > + if (parsing_place != SELECT_LIST) > + return; and in that one: > + if (const_item()) > + return; How will the fixes work together? Generally, I'm wondering if we're taking the right approach here: from an out-of-context view, it seems that approach "equip every Item with ability to evaluate to NULL under some special circumnstances" is very confusing, and one should just ignore the non-group-by items and produce NULLs instead.. Let's discuss it on irc. On Wed, Jun 13, 2012 at 08:52:19AM +0000, timour(a)askmonty.org wrote: > At file:///home/tsk/mprog/src/5.3/ > > ------------------------------------------------------------ > revno: 3543 > revision-id: timour(a)askmonty.org-20120613085100-1a3zlttyiebfvzuy > parent: psergey(a)askmonty.org-20120610100611-0l0qf7ov53h1elr9 > fixes bug(s): https://launchpad.net/bugs/1008686 > committer: timour(a)askmonty.org > branch nick: 5.3 > timestamp: Wed 2012-06-13 11:51:00 +0300 > message: > Fix bug lp:1008686 > > Analysis: > The fix for bug lp:985667 implements the method Item_subselect::no_rows_in_result() > for all main kinds of subqueries. The purpose of this method is to be called from > return_zero_rows() and set Items to some default value in the case when a query > returns no rows. Aggregates and subqueries require special treatment in this case. > > Every implementation of Item_subselect::no_rows_in_result() called > Item_subselect::make_const() to set the subquery predicate to its default value > irrespective of where the predicate was located in the query. Once the predicate > was set to a constant it was never executed. > > At the same time, the JOIN object of the fake select for UNIONs (the one used for > the final result of the UNION), was set after all subqueries in the union were > executed. Since we set the subquery as constant, it was never executed, and the > corresponding JOIN was never created. > > In order to decide whether the result of NOT IN is NULL or FALSE, Item_in_optimizer > needs to check if the subquery result was empty or not. This is where we got the > crash, because subselect_union_engine::no_rows() checks for > unit->fake_select_lex->join->send_records, and the join object was NULL. > > Solution: > If a subquery is in the HAVING clause it must be evaluated in order to know its > result, so that we can properly filter the result records. Once subqueries in the > HAVING clause are executed even in the case of no result rows, this specific > crash will be solved, because the UNION will be executed, and its JOIN will be > constructed. Therefore the fix for this crash is to narrow the fix for lp:985667, > and to apply Item_subselect::no_rows_in_result() only when the subquery predicate > is in the SELECT clause. > === modified file 'mysql-test/r/subselect.result' > --- a/mysql-test/r/subselect.result 2012-06-05 14:25:10 +0000 > +++ b/mysql-test/r/subselect.result 2012-06-13 08:51:00 +0000 > @@ -6112,5 +6112,30 @@ NULL UNION RESULT <union2,3> ALL NULL NU > SELECT SUM(a) AS f1, a AS f2 FROM (t1, t2) HAVING f2 >= ALL (SELECT 4 UNION SELECT 5) AND f1 = 7; > f1 f2 > drop table t1,t2; > +# > +# LP BUG#1008686 Server crashes in subselect_union_engine::no_rows on SELECT with impossible > +# WHERE and UNION in HAVING > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +EXPLAIN > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 0 HAVING a NOT IN ( SELECT 2 UNION SELECT 5 ) OR min_a != 1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL No tables used > +3 DEPENDENT UNION NULL NULL NULL NULL NULL NULL NULL No tables used > +NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 0 HAVING a NOT IN ( SELECT 2 UNION SELECT 5 ) OR min_a != 1; > +min_a a > +EXPLAIN > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 1=2 HAVING a NOT IN ( SELECT a from t1 UNION select a+1 from t1 ) OR min_a != 1; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +3 DEPENDENT UNION t1 ALL NULL NULL NULL NULL 2 Using where > +NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 1=2 HAVING a NOT IN ( SELECT a from t1 UNION select a+1 from t1 ) OR min_a != 1; > +min_a a > +drop table t1; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'mysql-test/t/subselect.test' > --- a/mysql-test/t/subselect.test 2012-05-22 12:22:55 +0000 > +++ b/mysql-test/t/subselect.test 2012-06-13 08:51:00 +0000 > @@ -5190,5 +5190,23 @@ SELECT SUM(a) AS f1, a AS f2 FROM (t1, t > > drop table t1,t2; > > +--echo # > +--echo # LP BUG#1008686 Server crashes in subselect_union_engine::no_rows on SELECT with impossible > +--echo # WHERE and UNION in HAVING > +--echo # > + > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > + > +EXPLAIN > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 0 HAVING a NOT IN ( SELECT 2 UNION SELECT 5 ) OR min_a != 1; > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 0 HAVING a NOT IN ( SELECT 2 UNION SELECT 5 ) OR min_a != 1; > + > +EXPLAIN > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 1=2 HAVING a NOT IN ( SELECT a from t1 UNION select a+1 from t1 ) OR min_a != 1; > +SELECT MIN(a) AS min_a, a FROM t1 WHERE 1=2 HAVING a NOT IN ( SELECT a from t1 UNION select a+1 from t1 ) OR min_a != 1; > + > +drop table t1; > + > --echo # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'sql/item_subselect.cc' > --- a/sql/item_subselect.cc 2012-05-22 12:22:55 +0000 > +++ b/sql/item_subselect.cc 2012-06-13 08:51:00 +0000 > @@ -892,6 +892,8 @@ void Item_maxmin_subselect::print(String > > void Item_maxmin_subselect::no_rows_in_result() > { > + if (parsing_place != SELECT_LIST) > + return; > value= Item_cache::get_cache(new Item_null()); > null_value= 0; > was_values= 0; > @@ -901,6 +903,8 @@ void Item_maxmin_subselect::no_rows_in_r > > void Item_singlerow_subselect::no_rows_in_result() > { > + if (parsing_place != SELECT_LIST) > + return; > value= Item_cache::get_cache(new Item_null()); > reset(); > make_const(); > @@ -1363,6 +1367,8 @@ Item* Item_exists_subselect::expr_cache_ > > void Item_exists_subselect::no_rows_in_result() > { > + if (parsing_place != SELECT_LIST) > + return; > value= 0; > null_value= 0; > make_const(); > @@ -2707,6 +2713,8 @@ void Item_allany_subselect::print(String > > void Item_allany_subselect::no_rows_in_result() > { > + if (parsing_place != SELECT_LIST) > + return; > value= 0; > null_value= 0; > was_null= 0; > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] Please review: [Commits] Rev 3406: Patch for mdev-287: CHEAP SQ: A query with subquery in SELECT list, EXISTS, inner joins takes hundreds times longer in file:///home/tsk/mprog/src/5.5-lpb944706/
by Sergei Petrunia 14 Jun '12

14 Jun '12

Hi Timour, Ok to push. I was wondering what are your plans with regards to pushing 5.5-timour to main? Is this the last issue? On Wed, May 30, 2012 at 12:22:06AM +0300, Timour Katchaounov wrote: > Sergey, > > Please review the following bug fix that we discussed today. > > > Timour > > > ------------------------------------------------------------ > revno: 3406 > revision-id: timour(a)askmonty.org-20120529211853-hww47vl7d4u4ae23 > parent: timour(a)askmonty.org-20120524110828-r0mm8sm1vn8a095e > committer: timour(a)askmonty.org > branch nick: 5.5-lpb944706 > timestamp: Wed 2012-05-30 00:18:53 +0300 > message: > Patch for mdev-287: CHEAP SQ: A query with subquery in SELECT list, EXISTS, inner joins takes hundreds times longer > > Analysis: > > The fix for lp:944706 introduces early subquery optimization. > While a subquery is being optimized some of its predicates may be > removed. In the test case, the EXISTS subquery is constant, and is > evaluated to TRUE. As a result the whole OR is TRUE, and thus the > correlated condition "b = alias1.b" is optimized away. The subquery > becomes non-correlated. > > The subquery cache is designed to work only for correlated subqueries. > If constant subquery optimization is disallowed, then the constant > subquery is not evaluated, the subquery remains correlated, and its > execution is cached. As a result execution is fast. > > However, when the constant subquery was optimized away, it was neither > cached by the subquery cache, nor it was cached by the internal subquery > caching. The latter was due to the fact that the subquery still appeared > as correlated to the subselect_XYZ_engine::exec methods, and they > re-executed the subquery on each call to Item_subselect::exec. > > Solution: > > The solution is to update the correlated status of the subquery after it has > been optimized. This status consists of: > - st_select_lex::is_correlated > - Item_subselect::is_correlated > - SELECT_LEX::uncacheable > - SELECT_LEX_UNIT::uncacheable > The status is updated by st_select_lex::update_correlated_cache(), and its > caller st_select_lex::optimize_unflattened_subqueries. The solution relies > on the fact that the optimizer already called > st_select_lex::update_used_tables() for each subquery. This allows to > efficiently update the correlated status of each subquery without walking > the whole subquery tree. > > Notice that his patch is an improvement over MySQL 5.6 and older, where > subqueries are not pre-optimized, and the above analysis is not possible. > === modified file 'mysql-test/r/derived_opt.result' > --- a/mysql-test/r/derived_opt.result 2012-02-14 14:52:56 +0000 > +++ b/mysql-test/r/derived_opt.result 2012-05-29 21:18:53 +0000 > @@ -161,24 +161,24 @@ prepare stmt1 from @stmt ; > execute stmt1 ; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > execute stmt1 ; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > explain SELECT (SELECT SUM(c1 + c12 + 0.0) FROM t2 where (t1.c2 - 0e-3) = t2.c2 GROUP BY t1.c15 LIMIT 1) as scalar_s, exists (select 1.0e+0 from t2 where t2.c3 * 9.0000000000 = t1.c4) as exists_s, c5 * 4 in (select c6 + 0.3e+1 from t2) as in_s, (c7 - 4, c8 - 4) in (select c9 + 4.0, c10 + 40e-1 from t2) as in_row_s FROM t1, (select c25 x, c32 y from t2) tt WHERE x * 1 = c25; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > deallocate prepare stmt1; > drop tables t1,t2; > set @@optimizer_switch=@save_optimizer_switch; > > === modified file 'mysql-test/r/explain.result' > --- a/mysql-test/r/explain.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/explain.result 2012-05-29 21:18:53 +0000 > @@ -225,10 +225,10 @@ INSERT INTO t2 VALUES (NULL), (0); > EXPLAIN EXTENDED SELECT (SELECT 1 FROM t2 WHERE d = c) FROM t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > Note 1276 Field or reference 'test.t1.c' of SELECT #2 was resolved in SELECT #1 > -Note 1003 select <expr_cache><NULL>((select 1 from `test`.`t2` where (`test`.`t2`.`d` = NULL))) AS `(SELECT 1 FROM t2 WHERE d = c)` from dual > +Note 1003 select (select 1 from `test`.`t2` where (`test`.`t2`.`d` = NULL)) AS `(SELECT 1 FROM t2 WHERE d = c)` from dual > DROP TABLE t1, t2; > # > # Bug#30302: Tables that were optimized away are printed in the > > === modified file 'mysql-test/r/index_merge_innodb.result' > --- a/mysql-test/r/index_merge_innodb.result 2012-02-21 19:51:56 +0000 > +++ b/mysql-test/r/index_merge_innodb.result 2012-05-29 21:18:53 +0000 > @@ -654,7 +654,7 @@ EXPLAIN SELECT t1.f1 FROM t1 > WHERE (SELECT COUNT(*) FROM t2 WHERE t2.f3 = 'h' AND t2.f2 = t1.f1) = 0 AND t1.f1 = 2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 const PRIMARY PRIMARY 4 const 1 Using index > -2 DEPENDENT SUBQUERY t2 index_merge f2,f3 f3,f2 2,5 NULL 1 Using intersect(f3,f2); Using where; Using index > +2 SUBQUERY t2 index_merge f2,f3 f3,f2 2,5 NULL 1 Using intersect(f3,f2); Using where; Using index > DROP TABLE t1,t2; > # > # BUG#56862/640419: Wrong result with sort_union index merge when one > > === modified file 'mysql-test/r/index_merge_myisam.result' > --- a/mysql-test/r/index_merge_myisam.result 2012-02-15 17:08:08 +0000 > +++ b/mysql-test/r/index_merge_myisam.result 2012-05-29 21:18:53 +0000 > @@ -1487,7 +1487,7 @@ EXPLAIN SELECT t1.f1 FROM t1 > WHERE (SELECT COUNT(*) FROM t2 WHERE t2.f3 = 'h' AND t2.f2 = t1.f1) = 0 AND t1.f1 = 2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t2 ref f2,f3 f2 5 const 1 Using where > +2 SUBQUERY t2 ref f2,f3 f2 5 const 1 Using where > DROP TABLE t1,t2; > create table t0 (a int); > insert into t0 values (0),(1),(2),(3),(4),(5),(6),(7),(8),(9); > > === modified file 'mysql-test/r/join_outer.result' > --- a/mysql-test/r/join_outer.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/join_outer.result 2012-05-29 21:18:53 +0000 > @@ -1872,7 +1872,7 @@ SELECT t2.a FROM t1 LEFT JOIN t2 ON (6) > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > 1 PRIMARY t2 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > +2 SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > Note 1003 select NULL AS `a` from `test`.`t2` where 1 > DROP TABLE t1,t2,t3; > > === modified file 'mysql-test/r/join_outer_jcl6.result' > --- a/mysql-test/r/join_outer_jcl6.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/join_outer_jcl6.result 2012-05-29 21:18:53 +0000 > @@ -1883,7 +1883,7 @@ SELECT t2.a FROM t1 LEFT JOIN t2 ON (6) > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > 1 PRIMARY t2 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > +2 SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > Note 1003 select NULL AS `a` from `test`.`t2` where 1 > DROP TABLE t1,t2,t3; > > === modified file 'mysql-test/r/ps.result' > --- a/mysql-test/r/ps.result 2012-04-10 06:28:13 +0000 > +++ b/mysql-test/r/ps.result 2012-05-29 21:18:53 +0000 > @@ -160,26 +160,26 @@ execute stmt1 ; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 6 DERIVED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > execute stmt1 ; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 6 DERIVED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > explain SELECT (SELECT SUM(c1 + c12 + 0.0) FROM t2 where (t1.c2 - 0e-3) = t2.c2 GROUP BY t1.c15 LIMIT 1) as scalar_s, exists (select 1.0e+0 from t2 where t2.c3 * 9.0000000000 = t1.c4) as exists_s, c5 * 4 in (select c6 + 0.3e+1 from t2) as in_s, (c7 - 4, c8 - 4) in (select c9 + 4.0, c10 + 40e-1 from t2) as in_row_s FROM t1, (select c25 x, c32 y from t2) tt WHERE x * 1 = c25; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 6 DERIVED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > -5 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -4 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +5 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +4 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > deallocate prepare stmt1; > set optimizer_switch=@tmp_optimizer_switch; > drop tables t1,t2; > > === modified file 'mysql-test/r/subselect.result' > --- a/mysql-test/r/subselect.result 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/r/subselect.result 2012-05-29 21:18:53 +0000 > @@ -51,12 +51,12 @@ ERROR 42S22: Reference 'a' not supported > EXPLAIN EXTENDED SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY <derived2> system NULL NULL NULL NULL 1 100.00 > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > 2 DERIVED NULL NULL NULL NULL NULL NULL NULL NULL No tables used > Warnings: > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > -Note 1003 select 1 AS `1` from dual having (<expr_cache><1>((select 1)) = 1) > +Note 1003 select 1 AS `1` from dual having ((select 1) = 1) > SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > 1 > 1 > @@ -904,7 +904,7 @@ a t1.a in (select t2.a from t2) > explain extended SELECT t1.a, t1.a in (select t2.a from t2) FROM t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL PRIMARY 4 NULL 4 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > +2 SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > Warnings: > Note 1003 select `test`.`t1`.`a` AS `a`,<expr_cache><`test`.`t1`.`a`>(<in_optimizer>(`test`.`t1`.`a`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`a`) in t2 on a checking NULL having <is_not_null_test>(`test`.`t2`.`a`))))) AS `t1.a in (select t2.a from t2)` from `test`.`t1` > CREATE TABLE t3 (a int(11) default '0'); > @@ -1317,7 +1317,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from dual where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > INSERT INTO t1 (pseudo) VALUES ('test1'); > @@ -1327,7 +1327,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from `test`.`t1` `a` where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > drop table t1; > @@ -1605,25 +1605,25 @@ a3 1 > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 NOT IN (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 = ANY (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))) AS `s1 = ANY (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 <> ALL (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 <> ALL (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2') from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL where (`test`.`t2`.`s1` < 'a2') having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2')` from `test`.`t1` > drop table t1,t2; > @@ -3097,7 +3097,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > @@ -3109,7 +3109,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > @@ -3160,7 +3160,7 @@ INSERT INTO t2 VALUES (1),(2),(3); > EXPLAIN SELECT a, a IN (SELECT a FROM t1) FROM t2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 3 > -2 DEPENDENT SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > +2 SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > SELECT a, a IN (SELECT a FROM t1) FROM t2; > a a IN (SELECT a FROM t1) > 1 1 > @@ -3696,7 +3696,7 @@ ORDER BY t1.t DESC LIMIT 1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > 1 PRIMARY t1 index NULL PRIMARY 16 NULL 11 Using where; Using index > -2 DEPENDENT SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > +2 SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > SELECT * FROM t1,t2 > WHERE t1.t = (SELECT t1.t FROM t1 > WHERE t1.t < t2.t AND t1.i2=1 AND t2.i1=t1.i1 > @@ -4577,7 +4577,7 @@ FROM t1 > WHERE a = 230; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > SELECT MAX(b), (SELECT COUNT(*) FROM st1,st2 WHERE st2.b <= t1.b) > FROM t1 > WHERE a = 230; > @@ -5713,7 +5713,7 @@ WHERE ('0') IN ( SELECT a FROM t1 GROUP > GROUP BY b; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > SELECT b FROM t1 > WHERE ('0') IN ( SELECT a FROM t1 GROUP BY a ) > GROUP BY b; > @@ -6530,7 +6530,7 @@ EXPLAIN > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 NULL > @@ -6538,7 +6538,7 @@ EXPLAIN > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 0 > @@ -6573,7 +6573,7 @@ WHERE alias1.a = alias2.a OR alias1.a = > HAVING field>'B' AND ( 'Moscow' ) IN ( SELECT a FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible HAVING > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) AS field > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR alias1.a = 'y' > @@ -6587,7 +6587,7 @@ id select_type table type possible_keys > 1 PRIMARY alias1 index a a 19 NULL 11 Using where; Using index > 1 PRIMARY alias2 ref a a 19 test.alias1.a 2 Using index > 1 PRIMARY alias3 index NULL a 19 NULL 11 Using index; Using join buffer (flat, BNL join) > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR ('Moscow') IN ( SELECT a FROM t1 ); > @@ -6615,5 +6615,90 @@ SELECT * FROM t1 > WHERE a1 = (SELECT COUNT(*) FROM t1 WHERE a1 IN (SELECT a1 FROM t1, t2)); > a1 > drop table t1, t2; > +# > +# MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +# inner joins takes hundreds times longer > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > +set @@expensive_subquery_limit= 0; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 6 > +Subquery_cache_miss 2 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 8 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 22 > +set @@expensive_subquery_limit= default; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 16 > +drop table t1, t2, t3; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'mysql-test/r/subselect4.result' > --- a/mysql-test/r/subselect4.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/subselect4.result 2012-05-29 21:18:53 +0000 > @@ -226,10 +226,10 @@ NULL > EXPLAIN EXTENDED SELECT (SELECT 1 FROM t2 WHERE d = c) AS RESULT FROM t1 ; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1276 Field or reference 'test.t1.c' of SELECT #2 was resolved in SELECT #1 > -Note 1003 select <expr_cache><>((select 1 from `test`.`t2` where 0)) AS `RESULT` from dual > +Note 1003 select (select 1 from `test`.`t2` where 0) AS `RESULT` from dual > first equivalent variant > SELECT (SELECT 1 FROM t2 WHERE d = IFNULL(c,NULL)) AS RESULT FROM t1 GROUP BY c ; > RESULT > @@ -237,10 +237,10 @@ NULL > EXPLAIN EXTENDED SELECT (SELECT 1 FROM t2 WHERE d = IFNULL(c,NULL)) AS RESULT FROM t1 GROUP BY c; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1276 Field or reference 'test.t1.c' of SELECT #2 was resolved in SELECT #1 > -Note 1003 select <expr_cache><>((select 1 from `test`.`t2` where 0)) AS `RESULT` from dual group by NULL > +Note 1003 select (select 1 from `test`.`t2` where 0) AS `RESULT` from dual group by NULL > second equivalent variant > SELECT (SELECT 1 FROM t2 WHERE d = c) AS RESULT FROM t1 GROUP BY c ; > RESULT > @@ -248,10 +248,10 @@ NULL > EXPLAIN EXTENDED SELECT (SELECT 1 FROM t2 WHERE d = c) AS RESULT FROM t1 GROUP BY c ; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1276 Field or reference 'test.t1.c' of SELECT #2 was resolved in SELECT #1 > -Note 1003 select <expr_cache><>((select 1 from `test`.`t2` where 0)) AS `RESULT` from dual group by NULL > +Note 1003 select (select 1 from `test`.`t2` where 0) AS `RESULT` from dual group by NULL > DROP TABLE t1,t2; > # > # BUG#45928 "Differing query results depending on MRR and > @@ -649,7 +649,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2) as not_in; > not_in > NULL > @@ -671,7 +671,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2) as not_in; > not_in > NULL > @@ -679,7 +679,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) > 7) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) > 7) as not_in; > not_in > 1 > @@ -687,7 +687,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) is null) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) is null) as not_in; > not_in > NULL > @@ -695,7 +695,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2) as not_in; > not_in > NULL > @@ -703,7 +703,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2) as not_in; > not_in > NULL > @@ -719,28 +719,28 @@ EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4) FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4) FROM t2); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3+f4, min(f4) FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3+f4, min(f4) FROM t2); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4)+max(f4) FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4)+max(f4) FROM t2); > f1 f2 > EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2) as not_in; > not_in > NULL > @@ -748,21 +748,21 @@ EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, count(f4) FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, count(f4) FROM t2); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, f3 + count(f4) FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, f3 + count(f4) FROM t2); > f1 f2 > EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2) as not_in; > not_in > NULL > @@ -770,7 +770,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) > 7) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) > 7) as not_in; > not_in > 1 > @@ -778,7 +778,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) is null) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 HAVING max(f4) is null) as not_in; > not_in > NULL > @@ -786,7 +786,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2) as not_in; > not_in > NULL > @@ -794,7 +794,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2) as not_in; > not_in > NULL > @@ -835,7 +835,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -857,7 +857,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -865,7 +865,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) > 7) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) > 7) as not_in; > not_in > 1 > @@ -873,7 +873,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) is null) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) is null) as not_in; > not_in > NULL > @@ -881,7 +881,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -889,7 +889,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -905,28 +905,28 @@ EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3+f4, min(f4) FROM t2 WHERE f3 > 10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3+f4, min(f4) FROM t2 WHERE f3 > 10); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4)+max(f4) FROM t2 WHERE f3 > 10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, min(f4)+max(f4) FROM t2 WHERE f3 > 10); > f1 f2 > EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, min(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -934,21 +934,21 @@ EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10); > f1 f2 > EXPLAIN > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, f3 + count(f4) FROM t2 WHERE f3 > 10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT * FROM t1 WHERE (2, 0) NOT IN (SELECT f3, f3 + count(f4) FROM t2 WHERE f3 > 10); > f1 f2 > EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -956,7 +956,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) > 7) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) > 7) as not_in; > not_in > 1 > @@ -964,7 +964,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) is null) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT f3, count(f4) FROM t2 WHERE f3 > 10 HAVING max(f4) is null) as not_in; > not_in > NULL > @@ -972,7 +972,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4) FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -980,7 +980,7 @@ EXPLAIN > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2 WHERE f3 > 10) as not_in; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > +2 SUBQUERY t2 range PRIMARY PRIMARY 4 NULL 1 Using index condition; Rowid-ordered scan > SELECT (2, 0) NOT IN (SELECT max(f3+f3), count(f4)+f3 FROM t2 WHERE f3 > 10) as not_in; > not_in > NULL > @@ -1240,7 +1240,7 @@ EXPLAIN > SELECT i FROM t1 WHERE (1) NOT IN (SELECT i FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t2 index_subquery k k 5 const 2 Using index > +2 SUBQUERY t2 index_subquery k k 5 const 2 Using index > DROP TABLE t2; > DROP TABLE t1; > # > @@ -1258,8 +1258,8 @@ WHERE ('v') IN (SELECT f4 FROM t2) > GROUP BY f9; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > (SELECT COUNT(f3) FROM t1 WHERE t2.f1) AS f9 > FROM t2 JOIN t1 ON t1.f3 > @@ -1274,8 +1274,8 @@ WHERE ('v') IN (SELECT f4 FROM t2) > ORDER BY f9; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > (SELECT COUNT(f3) FROM t1 WHERE t2.f1) AS f9 > FROM t2 JOIN t1 ON t1.f3 > @@ -1291,8 +1291,8 @@ WHERE ('v') IN (SELECT f4 FROM t2) > GROUP BY f9; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > (SELECT t2.f1 FROM t1 limit 1) AS f9 > FROM t2 JOIN t1 > @@ -1307,8 +1307,8 @@ WHERE ('v') IN (SELECT f4 FROM t2) > ORDER BY f9; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > (SELECT t2.f1 FROM t1 limit 1) AS f9 > FROM t2 JOIN t1 > @@ -1390,7 +1390,7 @@ EXPLAIN > SELECT 'bug' FROM DUAL WHERE ( 5 ) IN ( SELECT * FROM v1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED NULL NULL NULL NULL NULL NULL NULL No tables used > 4 UNION NULL NULL NULL NULL NULL NULL NULL No tables used > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1400,7 +1400,7 @@ EXPLAIN > SELECT ( 5 ) IN ( SELECT * FROM v1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED NULL NULL NULL NULL NULL NULL NULL No tables used > 4 UNION NULL NULL NULL NULL NULL NULL NULL No tables used > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1411,7 +1411,7 @@ EXPLAIN > SELECT 'bug' FROM DUAL WHERE ( 5 ) IN (SELECT * FROM v2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED t1 system NULL NULL NULL NULL 1 > 4 UNION t2 system NULL NULL NULL NULL 1 > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1431,7 +1431,7 @@ EXPLAIN > SELECT ( 5 ) IN ( SELECT * FROM v2 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED t1 system NULL NULL NULL NULL 1 > 4 UNION t2 system NULL NULL NULL NULL 1 > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1443,7 +1443,7 @@ EXPLAIN > SELECT 'bug' FROM DUAL WHERE ( 5 ) IN ( SELECT * FROM v1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED NULL NULL NULL NULL NULL NULL NULL No tables used > 4 UNION NULL NULL NULL NULL NULL NULL NULL No tables used > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1453,7 +1453,7 @@ EXPLAIN > SELECT ( 5 ) IN ( SELECT * FROM v1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED NULL NULL NULL NULL NULL NULL NULL No tables used > 4 UNION NULL NULL NULL NULL NULL NULL NULL No tables used > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1464,7 +1464,7 @@ EXPLAIN > SELECT 'bug' FROM DUAL WHERE ( 5 ) IN (SELECT * FROM v2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED t1 system NULL NULL NULL NULL 1 > 4 UNION t2 system NULL NULL NULL NULL 1 > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1474,7 +1474,7 @@ EXPLAIN > SELECT 'bug' FROM t3 WHERE ( 5 ) IN (SELECT * FROM v2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t3 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED t1 system NULL NULL NULL NULL 1 > 4 UNION t2 system NULL NULL NULL NULL 1 > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1484,7 +1484,7 @@ EXPLAIN > SELECT ( 5 ) IN ( SELECT * FROM v2 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY <derived3> ALL NULL NULL NULL NULL 2 Using where > 3 DERIVED t1 system NULL NULL NULL NULL 1 > 4 UNION t2 system NULL NULL NULL NULL 1 > NULL UNION RESULT <union3,4> ALL NULL NULL NULL NULL NULL > @@ -1635,8 +1635,8 @@ EXPLAIN > SELECT * FROM t1 WHERE ( 6 ) NOT IN ( SELECT t2.f3 FROM t2 JOIN t3 ON t3.f10 = t2.f10); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -2 DEPENDENT SUBQUERY t3 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t2 ref_or_null f10 f10 10 const,const 2 Using where; Using index > +2 SUBQUERY t3 system NULL NULL NULL NULL 1 > +2 SUBQUERY t2 ref_or_null f10 f10 10 const,const 2 Using where; Using index > SELECT * FROM t1 WHERE ( 6 ) NOT IN ( SELECT t2.f3 FROM t2 JOIN t3 ON t3.f10 = t2.f10); > f4 > drop table t1,t2,t3; > > === modified file 'mysql-test/r/subselect_extra_no_semijoin.result' > --- a/mysql-test/r/subselect_extra_no_semijoin.result 2012-01-13 14:50:02 +0000 > +++ b/mysql-test/r/subselect_extra_no_semijoin.result 2012-05-29 21:18:53 +0000 > @@ -46,7 +46,7 @@ select * from t1 > where id in (select id from t1 as x1 where (t1.cur_date is null)); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > Warnings: > Note 1276 Field or reference 'test.t1.cur_date' of SELECT #2 was resolved in SELECT #1 > Note 1003 select 1 AS `id`,'2007-04-25 18:30:22' AS `cur_date` from dual where <expr_cache><1>(<in_optimizer>(1,<exists>(select `test`.`x1`.`id` from `test`.`t1` `x1` where 0))) > @@ -58,7 +58,7 @@ select * from t2 > where id in (select id from t2 as x1 where (t2.cur_date is null)); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 100.00 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > Warnings: > Note 1276 Field or reference 'test.t2.cur_date' of SELECT #2 was resolved in SELECT #1 > Note 1003 select 1 AS `id`,'2007-04-25' AS `cur_date` from dual where <expr_cache><1>(<in_optimizer>(1,<exists>(select `test`.`x1`.`id` from `test`.`t2` `x1` where 0))) > > === modified file 'mysql-test/r/subselect_mat.result' > --- a/mysql-test/r/subselect_mat.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/subselect_mat.result 2012-05-29 21:18:53 +0000 > @@ -1163,7 +1163,7 @@ set @@optimizer_switch='materialization= > explain select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > min(a1) > NULL > @@ -1182,7 +1182,7 @@ set @@optimizer_switch='materialization= > explain select min(a1) from t1 where 7 in (select b1 from t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2); > min(a1) > NULL > @@ -1190,7 +1190,7 @@ NULL > explain select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > min(a1) > NULL > @@ -2162,7 +2162,7 @@ EXPLAIN > SELECT (SELECT f3a FROM t3) NOT IN (SELECT f1a FROM t1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -3 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +3 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > 2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT (SELECT f3a FROM t3) NOT IN (SELECT f1a FROM t1); > (SELECT f3a FROM t3) NOT IN (SELECT f1a FROM t1) > @@ -2179,7 +2179,7 @@ EXPLAIN > SELECT (SELECT f3a, f3b FROM t3) NOT IN (SELECT f1a, f1b FROM t1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL No tables used > -3 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +3 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > 2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT (SELECT f3a, f3b FROM t3) NOT IN (SELECT f1a, f1b FROM t1); > (SELECT f3a, f3b FROM t3) NOT IN (SELECT f1a, f1b FROM t1) > @@ -2234,7 +2234,7 @@ EXPLAIN EXTENDED > SELECT MAX(t1.b) AS max_res FROM t1 WHERE (9) IN (SELECT a FROM t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > Note 1003 select max(`test`.`t1`.`b`) AS `max_res` from `test`.`t1` where 0 > DROP TABLE t1,t2; > > === modified file 'mysql-test/r/subselect_mat_cost.result' > --- a/mysql-test/r/subselect_mat_cost.result 2012-03-01 22:22:22 +0000 > +++ b/mysql-test/r/subselect_mat_cost.result 2012-05-29 21:18:53 +0000 > @@ -496,7 +496,7 @@ from City > where City.population > 10000000; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY City range Population Population 5 NULL 4 Using index condition; Rowid-ordered scan > -2 DEPENDENT SUBQUERY Country index_subquery CountryCapital CountryCapital 5 func 2 Using index; Using where > +2 SUBQUERY Country index_subquery CountryCapital CountryCapital 5 func 2 Using index; Using where > select Name, City.id in (select capital from Country where capital is not null) as is_capital > from City > where City.population > 10000000; > > === modified file 'mysql-test/r/subselect_mat_cost_bugs.result' > --- a/mysql-test/r/subselect_mat_cost_bugs.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/subselect_mat_cost_bugs.result 2012-05-29 21:18:53 +0000 > @@ -124,7 +124,7 @@ FROM t3 RIGHT JOIN t1 ON t1.pk = t3.f1 > WHERE t3.f3 OR ( 3 ) IN ( SELECT f2 FROM t2 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT t1.* > FROM t3 RIGHT JOIN t1 ON t1.pk = t3.f1 > WHERE t3.f3 OR ( 3 ) IN ( SELECT f2 FROM t2 ); > > === modified file 'mysql-test/r/subselect_no_mat.result' > --- a/mysql-test/r/subselect_no_mat.result 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/r/subselect_no_mat.result 2012-05-29 21:18:53 +0000 > @@ -58,12 +58,12 @@ ERROR 42S22: Reference 'a' not supported > EXPLAIN EXTENDED SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY <derived2> system NULL NULL NULL NULL 1 100.00 > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > 2 DERIVED NULL NULL NULL NULL NULL NULL NULL NULL No tables used > Warnings: > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > -Note 1003 select 1 AS `1` from dual having (<expr_cache><1>((select 1)) = 1) > +Note 1003 select 1 AS `1` from dual having ((select 1) = 1) > SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > 1 > 1 > @@ -911,7 +911,7 @@ a t1.a in (select t2.a from t2) > explain extended SELECT t1.a, t1.a in (select t2.a from t2) FROM t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL PRIMARY 4 NULL 4 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > +2 SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > Warnings: > Note 1003 select `test`.`t1`.`a` AS `a`,<expr_cache><`test`.`t1`.`a`>(<in_optimizer>(`test`.`t1`.`a`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`a`) in t2 on a checking NULL having <is_not_null_test>(`test`.`t2`.`a`))))) AS `t1.a in (select t2.a from t2)` from `test`.`t1` > CREATE TABLE t3 (a int(11) default '0'); > @@ -1324,7 +1324,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from dual where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > INSERT INTO t1 (pseudo) VALUES ('test1'); > @@ -1334,7 +1334,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from `test`.`t1` `a` where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > drop table t1; > @@ -1612,25 +1612,25 @@ a3 1 > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 NOT IN (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 = ANY (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))) AS `s1 = ANY (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 <> ALL (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 <> ALL (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2') from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<expr_cache><`test`.`t1`.`s1`>(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL where (`test`.`t2`.`s1` < 'a2') having trigcond(<is_not_null_test>(`test`.`t2`.`s1`)))))))) AS `s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2')` from `test`.`t1` > drop table t1,t2; > @@ -3103,7 +3103,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > @@ -3115,7 +3115,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > @@ -3166,7 +3166,7 @@ INSERT INTO t2 VALUES (1),(2),(3); > EXPLAIN SELECT a, a IN (SELECT a FROM t1) FROM t2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 3 > -2 DEPENDENT SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > +2 SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > SELECT a, a IN (SELECT a FROM t1) FROM t2; > a a IN (SELECT a FROM t1) > 1 1 > @@ -3700,7 +3700,7 @@ ORDER BY t1.t DESC LIMIT 1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > 1 PRIMARY t1 index NULL PRIMARY 16 NULL 11 Using where; Using index > -2 DEPENDENT SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > +2 SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > SELECT * FROM t1,t2 > WHERE t1.t = (SELECT t1.t FROM t1 > WHERE t1.t < t2.t AND t1.i2=1 AND t2.i1=t1.i1 > @@ -4526,13 +4526,13 @@ SET join_cache_level=0; > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > @@ -4579,7 +4579,7 @@ FROM t1 > WHERE a = 230; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > SELECT MAX(b), (SELECT COUNT(*) FROM st1,st2 WHERE st2.b <= t1.b) > FROM t1 > WHERE a = 230; > @@ -5714,7 +5714,7 @@ WHERE ('0') IN ( SELECT a FROM t1 GROUP > GROUP BY b; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > SELECT b FROM t1 > WHERE ('0') IN ( SELECT a FROM t1 GROUP BY a ) > GROUP BY b; > @@ -6529,7 +6529,7 @@ EXPLAIN > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 NULL > @@ -6537,7 +6537,7 @@ EXPLAIN > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 0 > @@ -6572,7 +6572,7 @@ WHERE alias1.a = alias2.a OR alias1.a = > HAVING field>'B' AND ( 'Moscow' ) IN ( SELECT a FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible HAVING > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) AS field > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR alias1.a = 'y' > @@ -6586,7 +6586,7 @@ id select_type table type possible_keys > 1 PRIMARY alias1 index a a 19 NULL 11 Using where; Using index > 1 PRIMARY alias2 ref a a 19 test.alias1.a 2 Using index > 1 PRIMARY alias3 index NULL a 19 NULL 11 Using index; Using join buffer (flat, BNL join) > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR ('Moscow') IN ( SELECT a FROM t1 ); > @@ -6613,6 +6613,91 @@ SELECT * FROM t1 > WHERE a1 = (SELECT COUNT(*) FROM t1 WHERE a1 IN (SELECT a1 FROM t1, t2)); > a1 > drop table t1, t2; > +# > +# MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +# inner joins takes hundreds times longer > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > +set @@expensive_subquery_limit= 0; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 6 > +Subquery_cache_miss 2 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 8 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 22 > +set @@expensive_subquery_limit= default; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 16 > +drop table t1, t2, t3; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set optimizer_switch=default; > > === modified file 'mysql-test/r/subselect_no_opts.result' > --- a/mysql-test/r/subselect_no_opts.result 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/r/subselect_no_opts.result 2012-05-29 21:18:53 +0000 > @@ -54,7 +54,7 @@ ERROR 42S22: Reference 'a' not supported > EXPLAIN EXTENDED SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY <derived2> system NULL NULL NULL NULL 1 100.00 > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > 2 DERIVED NULL NULL NULL NULL NULL NULL NULL NULL No tables used > Warnings: > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > @@ -907,7 +907,7 @@ a t1.a in (select t2.a from t2) > explain extended SELECT t1.a, t1.a in (select t2.a from t2) FROM t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL PRIMARY 4 NULL 4 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > +2 SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > Warnings: > Note 1003 select `test`.`t1`.`a` AS `a`,<in_optimizer>(`test`.`t1`.`a`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`a`) in t2 on a checking NULL having <is_not_null_test>(`test`.`t2`.`a`)))) AS `t1.a in (select t2.a from t2)` from `test`.`t1` > CREATE TABLE t3 (a int(11) default '0'); > @@ -1320,7 +1320,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from dual where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > INSERT INTO t1 (pseudo) VALUES ('test1'); > @@ -1330,7 +1330,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from `test`.`t1` `a` where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > drop table t1; > @@ -1608,25 +1608,25 @@ a3 1 > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 NOT IN (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 = ANY (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))) AS `s1 = ANY (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 <> ALL (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 <> ALL (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2') from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL where (`test`.`t2`.`s1` < 'a2') having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2')` from `test`.`t1` > drop table t1,t2; > @@ -3099,7 +3099,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > @@ -3111,7 +3111,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > @@ -3162,7 +3162,7 @@ INSERT INTO t2 VALUES (1),(2),(3); > EXPLAIN SELECT a, a IN (SELECT a FROM t1) FROM t2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 3 > -2 DEPENDENT SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > +2 SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > SELECT a, a IN (SELECT a FROM t1) FROM t2; > a a IN (SELECT a FROM t1) > 1 1 > @@ -3696,7 +3696,7 @@ ORDER BY t1.t DESC LIMIT 1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > 1 PRIMARY t1 index NULL PRIMARY 16 NULL 11 Using where; Using index > -2 DEPENDENT SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > +2 SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > SELECT * FROM t1,t2 > WHERE t1.t = (SELECT t1.t FROM t1 > WHERE t1.t < t2.t AND t1.i2=1 AND t2.i1=t1.i1 > @@ -4522,13 +4522,13 @@ SET join_cache_level=0; > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > @@ -4575,7 +4575,7 @@ FROM t1 > WHERE a = 230; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > SELECT MAX(b), (SELECT COUNT(*) FROM st1,st2 WHERE st2.b <= t1.b) > FROM t1 > WHERE a = 230; > @@ -5710,7 +5710,7 @@ WHERE ('0') IN ( SELECT a FROM t1 GROUP > GROUP BY b; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > SELECT b FROM t1 > WHERE ('0') IN ( SELECT a FROM t1 GROUP BY a ) > GROUP BY b; > @@ -6525,7 +6525,7 @@ EXPLAIN > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 NULL > @@ -6533,7 +6533,7 @@ EXPLAIN > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 0 > @@ -6568,7 +6568,7 @@ WHERE alias1.a = alias2.a OR alias1.a = > HAVING field>'B' AND ( 'Moscow' ) IN ( SELECT a FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible HAVING > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) AS field > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR alias1.a = 'y' > @@ -6582,7 +6582,7 @@ id select_type table type possible_keys > 1 PRIMARY alias1 index a a 19 NULL 11 Using where; Using index > 1 PRIMARY alias2 ref a a 19 test.alias1.a 2 Using index > 1 PRIMARY alias3 index NULL a 19 NULL 11 Using index; Using join buffer (flat, BNL join) > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR ('Moscow') IN ( SELECT a FROM t1 ); > @@ -6610,6 +6610,91 @@ SELECT * FROM t1 > WHERE a1 = (SELECT COUNT(*) FROM t1 WHERE a1 IN (SELECT a1 FROM t1, t2)); > a1 > drop table t1, t2; > +# > +# MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +# inner joins takes hundreds times longer > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > +set @@expensive_subquery_limit= 0; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 6 > +Subquery_cache_miss 2 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 8 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 22 > +set @@expensive_subquery_limit= default; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 16 > +drop table t1, t2, t3; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set @optimizer_switch_for_subselect_test=null; > > === modified file 'mysql-test/r/subselect_no_scache.result' > --- a/mysql-test/r/subselect_no_scache.result 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/r/subselect_no_scache.result 2012-05-29 21:18:53 +0000 > @@ -57,7 +57,7 @@ ERROR 42S22: Reference 'a' not supported > EXPLAIN EXTENDED SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY <derived2> system NULL NULL NULL NULL 1 100.00 > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > 2 DERIVED NULL NULL NULL NULL NULL NULL NULL NULL No tables used > Warnings: > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > @@ -910,7 +910,7 @@ a t1.a in (select t2.a from t2) > explain extended SELECT t1.a, t1.a in (select t2.a from t2) FROM t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL PRIMARY 4 NULL 4 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > +2 SUBQUERY t2 index_subquery a a 5 func 2 100.00 Using index > Warnings: > Note 1003 select `test`.`t1`.`a` AS `a`,<in_optimizer>(`test`.`t1`.`a`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`a`) in t2 on a checking NULL having <is_not_null_test>(`test`.`t2`.`a`)))) AS `t1.a in (select t2.a from t2)` from `test`.`t1` > CREATE TABLE t3 (a int(11) default '0'); > @@ -1323,7 +1323,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from dual where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > INSERT INTO t1 (pseudo) VALUES ('test1'); > @@ -1333,7 +1333,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from `test`.`t1` `a` where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > drop table t1; > @@ -1611,25 +1611,25 @@ a3 1 > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 NOT IN (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 = ANY (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))) AS `s1 = ANY (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 <> ALL (SELECT s1 FROM t2) from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 <> ALL (SELECT s1 FROM t2)` from `test`.`t1` > explain extended select s1, s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2') from t1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 index NULL s1 6 NULL 3 100.00 Using index > -2 DEPENDENT SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > +2 SUBQUERY t2 index_subquery s1 s1 6 func 2 100.00 Using index; Using where; Full scan on NULL key > Warnings: > Note 1003 select `test`.`t1`.`s1` AS `s1`,(not(<in_optimizer>(`test`.`t1`.`s1`,<exists>(<index_lookup>(<cache>(`test`.`t1`.`s1`) in t2 on s1 checking NULL where (`test`.`t2`.`s1` < 'a2') having trigcond(<is_not_null_test>(`test`.`t2`.`s1`))))))) AS `s1 NOT IN (SELECT s1 FROM t2 WHERE s1 < 'a2')` from `test`.`t1` > drop table t1,t2; > @@ -3103,7 +3103,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > @@ -3115,7 +3115,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > @@ -3166,7 +3166,7 @@ INSERT INTO t2 VALUES (1),(2),(3); > EXPLAIN SELECT a, a IN (SELECT a FROM t1) FROM t2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 3 > -2 DEPENDENT SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > +2 SUBQUERY t1 index_subquery a a 5 func 2 Using index; Full scan on NULL key > SELECT a, a IN (SELECT a FROM t1) FROM t2; > a a IN (SELECT a FROM t1) > 1 1 > @@ -3702,7 +3702,7 @@ ORDER BY t1.t DESC LIMIT 1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > 1 PRIMARY t1 index NULL PRIMARY 16 NULL 11 Using where; Using index > -2 DEPENDENT SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > +2 SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > SELECT * FROM t1,t2 > WHERE t1.t = (SELECT t1.t FROM t1 > WHERE t1.t < t2.t AND t1.i2=1 AND t2.i1=t1.i1 > @@ -4583,7 +4583,7 @@ FROM t1 > WHERE a = 230; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > SELECT MAX(b), (SELECT COUNT(*) FROM st1,st2 WHERE st2.b <= t1.b) > FROM t1 > WHERE a = 230; > @@ -5719,7 +5719,7 @@ WHERE ('0') IN ( SELECT a FROM t1 GROUP > GROUP BY b; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > SELECT b FROM t1 > WHERE ('0') IN ( SELECT a FROM t1 GROUP BY a ) > GROUP BY b; > @@ -6536,7 +6536,7 @@ EXPLAIN > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 NULL > @@ -6544,7 +6544,7 @@ EXPLAIN > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 0 > @@ -6579,7 +6579,7 @@ WHERE alias1.a = alias2.a OR alias1.a = > HAVING field>'B' AND ( 'Moscow' ) IN ( SELECT a FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible HAVING > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) AS field > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR alias1.a = 'y' > @@ -6593,7 +6593,7 @@ id select_type table type possible_keys > 1 PRIMARY alias1 index a a 19 NULL 11 Using where; Using index > 1 PRIMARY alias2 ref a a 19 test.alias1.a 2 Using index > 1 PRIMARY alias3 index NULL a 19 NULL 11 Using index; Using join buffer (flat, BNL join) > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR ('Moscow') IN ( SELECT a FROM t1 ); > @@ -6621,6 +6621,91 @@ SELECT * FROM t1 > WHERE a1 = (SELECT COUNT(*) FROM t1 WHERE a1 IN (SELECT a1 FROM t1, t2)); > a1 > drop table t1, t2; > +# > +# MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +# inner joins takes hundreds times longer > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > +set @@expensive_subquery_limit= 0; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 58 > +set @@expensive_subquery_limit= default; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 16 > +drop table t1, t2, t3; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set optimizer_switch=default; > > === modified file 'mysql-test/r/subselect_no_semijoin.result' > --- a/mysql-test/r/subselect_no_semijoin.result 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/r/subselect_no_semijoin.result 2012-05-29 21:18:53 +0000 > @@ -54,12 +54,12 @@ ERROR 42S22: Reference 'a' not supported > EXPLAIN EXTENDED SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY <derived2> system NULL NULL NULL NULL 1 100.00 > -3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > +3 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > 2 DERIVED NULL NULL NULL NULL NULL NULL NULL NULL No tables used > Warnings: > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > Note 1276 Field or reference 'b.a' of SELECT #3 was resolved in SELECT #1 > -Note 1003 select 1 AS `1` from dual having (<expr_cache><1>((select 1)) = 1) > +Note 1003 select 1 AS `1` from dual having ((select 1) = 1) > SELECT 1 FROM (SELECT 1 as a) as b HAVING (SELECT a)=1; > 1 > 1 > @@ -1320,7 +1320,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from dual where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > INSERT INTO t1 (pseudo) VALUES ('test1'); > @@ -1330,7 +1330,7 @@ SELECT 0 IN (SELECT 1 FROM t1 a); > EXPLAIN EXTENDED SELECT 0 IN (SELECT 1 FROM t1 a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL No tables used > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > Warnings: > Note 1003 select <in_optimizer>(0,<exists>(select 1 from `test`.`t1` `a` where (0 = 1))) AS `0 IN (SELECT 1 FROM t1 a)` > drop table t1; > @@ -3099,7 +3099,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > @@ -3111,7 +3111,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > 1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > -2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > +2 SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > @@ -3696,7 +3696,7 @@ ORDER BY t1.t DESC LIMIT 1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > 1 PRIMARY t1 index NULL PRIMARY 16 NULL 11 Using where; Using index > -2 DEPENDENT SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > +2 SUBQUERY t1 range PRIMARY PRIMARY 16 NULL 5 Using where; Using index > SELECT * FROM t1,t2 > WHERE t1.t = (SELECT t1.t FROM t1 > WHERE t1.t < t2.t AND t1.i2=1 AND t2.i1=t1.i1 > @@ -4522,13 +4522,13 @@ SET join_cache_level=0; > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > @@ -4575,7 +4575,7 @@ FROM t1 > WHERE a = 230; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > SELECT MAX(b), (SELECT COUNT(*) FROM st1,st2 WHERE st2.b <= t1.b) > FROM t1 > WHERE a = 230; > @@ -5710,7 +5710,7 @@ WHERE ('0') IN ( SELECT a FROM t1 GROUP > GROUP BY b; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > SELECT b FROM t1 > WHERE ('0') IN ( SELECT a FROM t1 GROUP BY a ) > GROUP BY b; > @@ -6525,7 +6525,7 @@ EXPLAIN > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), (SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 NULL > @@ -6533,7 +6533,7 @@ EXPLAIN > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(f1), exists(SELECT f1 FROM t1 WHERE f2 > 0 limit 1) AS f4 FROM t2, t1 WHERE 'v'= f3; > COUNT(f1) f4 > 0 0 > @@ -6568,7 +6568,7 @@ WHERE alias1.a = alias2.a OR alias1.a = > HAVING field>'B' AND ( 'Moscow' ) IN ( SELECT a FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible HAVING > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) AS field > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR alias1.a = 'y' > @@ -6582,7 +6582,7 @@ id select_type table type possible_keys > 1 PRIMARY alias1 index a a 19 NULL 11 Using where; Using index > 1 PRIMARY alias2 ref a a 19 test.alias1.a 2 Using index > 1 PRIMARY alias3 index NULL a 19 NULL 11 Using index; Using join buffer (flat, BNL join) > -2 DEPENDENT SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > +2 SUBQUERY t1 index_subquery a a 19 const 1 Using index; Using where > SELECT MAX( alias2.a ) > FROM t1 AS alias1, t1 AS alias2, t1 AS alias3 > WHERE alias1.a = alias2.a OR ('Moscow') IN ( SELECT a FROM t1 ); > @@ -6610,6 +6610,91 @@ SELECT * FROM t1 > WHERE a1 = (SELECT COUNT(*) FROM t1 WHERE a1 IN (SELECT a1 FROM t1, t2)); > a1 > drop table t1, t2; > +# > +# MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +# inner joins takes hundreds times longer > +# > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > +set @@expensive_subquery_limit= 0; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using where > +2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 6 > +Subquery_cache_miss 2 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 8 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 22 > +set @@expensive_subquery_limit= default; > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY alias1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY alias2 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +1 PRIMARY alias3 ALL NULL NULL NULL NULL 2 Using join buffer (flat, BNL join) > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where; Using join buffer (flat, BNL join) > +3 SUBQUERY t3 ALL NULL NULL NULL NULL 2 > +flush status; > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > +(SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +NULL > +show status like "subquery_cache%"; > +Variable_name Value > +Subquery_cache_hit 0 > +Subquery_cache_miss 0 > +show status like '%Handler_read%'; > +Variable_name Value > +Handler_read_first 0 > +Handler_read_key 0 > +Handler_read_last 0 > +Handler_read_next 0 > +Handler_read_prev 0 > +Handler_read_rnd 0 > +Handler_read_rnd_deleted 0 > +Handler_read_rnd_next 16 > +drop table t1, t2, t3; > # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > set @optimizer_switch_for_subselect_test=null; > > === modified file 'mysql-test/r/subselect_partial_match.result' > --- a/mysql-test/r/subselect_partial_match.result 2011-12-04 21:31:42 +0000 > +++ b/mysql-test/r/subselect_partial_match.result 2012-05-29 21:18:53 +0000 > @@ -879,7 +879,7 @@ EXPLAIN > SELECT * FROM t1 WHERE (6, 4 ) NOT IN (SELECT b, a FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where > SELECT * FROM t1 WHERE (6, 4 ) NOT IN (SELECT b, a FROM t2); > c > 0 > @@ -888,7 +888,7 @@ EXPLAIN > SELECT * FROM t1 WHERE (6, 4 ) NOT IN (SELECT a, b FROM t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > -2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where > SELECT * FROM t1 WHERE (6, 4 ) NOT IN (SELECT a, b FROM t2); > c > 0 > @@ -913,7 +913,7 @@ set @@optimizer_switch='in_to_exists=on, > EXPLAIN SELECT * FROM t2 WHERE ( 3 , 1 ) NOT IN ( SELECT f1 , f2 FROM t1 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 3 Using where > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 3 Using where > SELECT * FROM t2 WHERE ( 3 , 1 ) NOT IN ( SELECT f1 , f2 FROM t1 ); > f3 > 5 > > === modified file 'mysql-test/r/subselect_sj_mat.result' > --- a/mysql-test/r/subselect_sj_mat.result 2012-05-17 10:46:05 +0000 > +++ b/mysql-test/r/subselect_sj_mat.result 2012-05-29 21:18:53 +0000 > @@ -1198,7 +1198,7 @@ set @@optimizer_switch='materialization= > explain select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > min(a1) > NULL > @@ -1207,7 +1207,7 @@ set @@optimizer_switch='semijoin=off'; > explain select min(a1) from t1 where 7 in (select b1 from t2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2); > min(a1) > NULL > @@ -1224,7 +1224,7 @@ NULL > explain select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > min(a1) > NULL > > === modified file 'mysql-test/t/subselect.test' > --- a/mysql-test/t/subselect.test 2012-05-18 11:52:01 +0000 > +++ b/mysql-test/t/subselect.test 2012-05-29 21:18:53 +0000 > @@ -5580,5 +5580,49 @@ WHERE a1 = (SELECT COUNT(*) FROM t1 WHER > > drop table t1, t2; > > +--echo # > +--echo # MDEV-287 CHEAP SQ: A query with subquery in SELECT list, EXISTS, > +--echo # inner joins takes hundreds times longer > +--echo # > + > +CREATE TABLE t1 (a INT); > +INSERT INTO t1 VALUES (1),(7); > + > +CREATE TABLE t2 (b INT); > +INSERT INTO t2 VALUES (4),(5); > + > +CREATE TABLE t3 (c INT); > +INSERT INTO t3 VALUES (8),(3); > + > +set @@expensive_subquery_limit= 0; > + > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > + > +flush status; > + > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > + > +show status like "subquery_cache%"; > +show status like '%Handler_read%'; > + > +set @@expensive_subquery_limit= default; > + > +EXPLAIN > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > + > +flush status; > + > +SELECT (SELECT MIN(b) FROM t1, t2 WHERE b = a AND (b = alias1.b OR EXISTS (SELECT * FROM t3))) > +FROM t2 alias1, t1 alias2, t1 alias3; > + > +show status like "subquery_cache%"; > +show status like '%Handler_read%'; > + > +drop table t1, t2, t3; > + > --echo # return optimizer switch changed in the beginning of this test > set optimizer_switch=@subselect_tmp; > > === modified file 'sql/sql_lex.cc' > --- a/sql/sql_lex.cc 2012-05-17 10:46:05 +0000 > +++ b/sql/sql_lex.cc 2012-05-29 21:18:53 +0000 > @@ -3443,6 +3443,7 @@ bool st_select_lex::optimize_unflattened > } > > bool empty_union_result= true; > + bool is_correlated_unit= false; > /* > If the subquery is a UNION, optimize all the subqueries in the UNION. If > there is no UNION, then the loop will execute once for the subquery. > @@ -3467,6 +3468,8 @@ bool st_select_lex::optimize_unflattened > inner_join->select_options|= SELECT_DESCRIBE; > } > res= inner_join->optimize(); > + sl->update_correlated_cache(); > + is_correlated_unit|= sl->is_correlated; > inner_join->select_options= save_options; > un->thd->lex->current_select= save_select; > if (empty_union_result) > @@ -3482,6 +3485,9 @@ bool st_select_lex::optimize_unflattened > } > if (empty_union_result) > subquery_predicate->no_rows_in_result(); > + if (!is_correlated_unit) > + un->uncacheable&= ~UNCACHEABLE_DEPENDENT; > + subquery_predicate->is_correlated= is_correlated_unit; > } > } > return FALSE; > @@ -3850,6 +3856,61 @@ void SELECT_LEX::update_used_tables() > } > > > +/** > + @brief > + Update is_correlated cache for this select > + > + @details > +*/ > + > +void st_select_lex::update_correlated_cache() > +{ > + TABLE_LIST *tl; > + List_iterator<TABLE_LIST> ti(leaf_tables); > + > + is_correlated= false; > + > + while ((tl= ti++)) > + { > + if (tl->on_expr) > + is_correlated|= test(tl->on_expr->used_tables() & OUTER_REF_TABLE_BIT); > + for (TABLE_LIST *embedding= tl->embedding ; embedding ; > + embedding= embedding->embedding) > + { > + if (embedding->on_expr) > + is_correlated|= test(embedding->on_expr->used_tables() & > + OUTER_REF_TABLE_BIT); > + } > + } > + > + if (join->conds) > + is_correlated|= test(join->conds->used_tables() & OUTER_REF_TABLE_BIT); > + > + if (join->having) > + is_correlated|= test(join->having->used_tables() & OUTER_REF_TABLE_BIT); > + > + if (join->tmp_having) > + is_correlated|= test(join->tmp_having->used_tables() & OUTER_REF_TABLE_BIT); > + > + Item *item; > + List_iterator_fast<Item> it(join->fields_list); > + while ((item= it++)) > + is_correlated|= test(item->used_tables() & OUTER_REF_TABLE_BIT); > + > + for (ORDER *order= group_list.first; order; order= order->next) > + is_correlated|= test((*order->item)->used_tables() & OUTER_REF_TABLE_BIT); > + > + if (!master_unit()->is_union()) > + { > + for (ORDER *order= order_list.first; order; order= order->next) > + is_correlated|= test((*order->item)->used_tables() & OUTER_REF_TABLE_BIT); > + } > + > + if (!is_correlated) > + uncacheable&= ~UNCACHEABLE_DEPENDENT; > +} > + > + > /** > Set the EXPLAIN type for this subquery. > */ > > === modified file 'sql/sql_lex.h' > --- a/sql/sql_lex.h 2012-05-17 10:46:05 +0000 > +++ b/sql/sql_lex.h 2012-05-29 21:18:53 +0000 > @@ -1018,6 +1018,7 @@ class st_select_lex: public st_select_le > void mark_as_belong_to_derived(TABLE_LIST *derived); > void increase_derived_records(ha_rows records); > void update_used_tables(); > + void update_correlated_cache(); > void mark_const_derived(bool empty); > > bool save_leaf_tables(THD *thd); > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-181: XID crash recovery across binlog boundaries
by Sergei Golubchik 12 Jun '12

12 Jun '12

Hi, Kristian! Here's my review. Only few comments and questions, see below. On May 14, Kristian Nielsen wrote: > At http://bazaar.launchpad.net/~maria-captains/maria/5.5 > > ------------------------------------------------------------ > revno: 3372 > revision-id: knielsen(a)knielsen-hq.org-20120514095140-n685kn0n4vtronn5 > timestamp: Mon 2012-05-14 11:51:40 +0200 > message: > MDEV-181: XID crash recovery across binlog boundaries > === added file 'mysql-test/suite/binlog/t/binlog_xa_recover.test' > --- mysql-test/suite/binlog/t/binlog_xa_recover.test 1970-01-01 00:00:00 +0000 > +++ mysql-test/suite/binlog/t/binlog_xa_recover.test 2012-05-14 09:51:40 +0000 > @@ -0,0 +1,172 @@ > +--source include/have_innodb.inc > +--source include/have_debug.inc > +--source include/have_debug_sync.inc > +--source include/have_binlog_format_row.inc > + > +SET GLOBAL max_binlog_size= 4096; > + > +CREATE TABLE t1 (a INT PRIMARY KEY, b MEDIUMTEXT) ENGINE=Innodb; > +CREATE TABLE t2 (a INT PRIMARY KEY, b MEDIUMTEXT) ENGINE=Myisam; > + > +# Transactions are not guaranteed stored durably on disk in the engine until > +# they are fsync()ed, which normally happens during commit(). But there is no > +# guarantee that they will _not_ be durable, in particular loosing results > +# of a write(2) system call normally requires a kernel crash (as opposed to > +# just mysqld crash), which is inconvenient to do in a test suite. > +# So instead we do an error insert to prevent commit_ordered() from being > +# called in the engine - so nothing will be written to disk at all, and crash > +# recovery is sure to be needed. Are these assumptions reasonable? You seem to assume that without commit_ordered the engine will still work normally. but simply won't write changes to disk. Is it always the case? > +SET @@global.debug_dbug='+d,skip_commit_ordered'; I don't see where you restore the old value of debug_dbug > + > +INSERT INTO t1 VALUES (0, REPEAT("x", 4100)); how comes this insert is not affected by skip_commit_ordered ? > + > +# Now start a bunch of transactions that span multiple binlog > +# files. Leave then in the state prepared-but-not-committed in the engine > +# and crash the server. Check that crash recovery is able to recover all > +# of them. > + > +connect(con1,localhost,root,,); > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con1_ready WAIT_FOR _ever"; > +send INSERT INTO t1 VALUES (1, REPEAT("x", 4100)); > + > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con1_ready"; > +INSERT INTO t2 VALUES (1, "force binlog rotation"); > + > +connect(con2,localhost,root,,); > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con2_ready WAIT_FOR _ever"; > +send INSERT INTO t1 VALUES (2, NULL); > + > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con2_ready"; > + > +connect(con3,localhost,root,,); > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con3_ready WAIT_FOR _ever"; > +send INSERT INTO t1 VALUES (3, REPEAT("x", 4100)); > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con3_ready"; > +INSERT INTO t2 VALUES (2, "force binlog rotation"); > +# So we won't get warnings about t2 being crashed. > +FLUSH TABLES t2; > + > +# Check that everything is committed in binary log. > +--source include/show_binary_logs.inc > +--let $binlog_file= master-bin.000001 > +--let $binlog_start= 4 > +--source include/show_binlog_events.inc > +--let $binlog_file= master-bin.000002 > +--source include/show_binlog_events.inc > +--let $binlog_file= master-bin.000003 > +--source include/show_binlog_events.inc > +--let $binlog_file= master-bin.000004 > +--source include/show_binlog_events.inc > + > +# Check that transactions really are not yet committed in engine. > +# (This works because of debug_dbug='+d,skip_commit_ordered'). > +--echo We should see only one entry here, a=0: > +SELECT a FROM t1 ORDER BY a; > + > + > +# Check that server will not purge too much. > +PURGE BINARY LOGS TO "master-bin.000004"; > +--source include/show_binary_logs.inc > + > +# Now crash the server with one more transaction in prepared state. > +system echo wait-binlog_xa_recover.test >> $MYSQLTEST_VARDIR/tmp/mysqld.1.expect; better use --write_file directive > +SET SESSION debug_dbug="+d,crash_commit_after_log"; > +--error 2006,2013 > +INSERT INTO t1 VALUES (4, NULL); > + > +system echo restart-group_commit_binlog_pos.test >> $MYSQLTEST_VARDIR/tmp/mysqld.1.expect; same > + > +connection default; > +--enable_reconnect > +--source include/wait_until_connected_again.inc > + > +# Check that all transactions are recovered. > +SELECT a FROM t1 ORDER BY a; > + > + > +--echo *** Test that RESET MASTER waits for pending XIDs to be unlogged. Ah, good point > + > +SET @old_max_binlog_size= @@global.max_binlog_size; > +SET GLOBAL max_binlog_size= 4096; > +# con10 will hang with a pending XID, blocking RESET MASTER. > +connect(con10,localhost,root,,); > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con10_ready WAIT_FOR con10_go"; > +send INSERT INTO t1 VALUES (10, NULL); > + > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con10_ready"; > +# Let's add a few binlog rotations just for good measure. > +INSERT INTO t2 VALUES (10, REPEAT("x", 4100)); > +INSERT INTO t2 VALUES (11, REPEAT("x", 4100)); > +--source include/show_binary_logs.inc > +SET DEBUG_SYNC= "execute_command_after_close_tables SIGNAL reset_master_done"; > +send RESET MASTER; > + > +connect(con11,localhost,root,,); > +--echo This will timeout, as RESET MASTER is blocked > +SET DEBUG_SYNC= "now WAIT_FOR reset_master_done TIMEOUT 1"; > +# Wake up transaction to allow RESET MASTER to complete. > +SET DEBUG_SYNC= "now SIGNAL con10_go"; > + > +connection con10; > +reap; > + > +connection default; > +reap; > +--source include/show_binary_logs.inc > + > + > +--echo *** Test that binlog N is active, and last pending trx in (N-1) is > +--echo unlogged while there is still a pending trx in (N-2). > + > +connection con10; > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con10_ready WAIT_FOR con10_continue"; > +send INSERT INTO t1 VALUES (20, REPEAT("x", 4100)); > + > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con10_ready"; > +INSERT INTO t2 VALUES (3, "force binlog rotation"); > + > +connection con11; > +SET DEBUG_SYNC= "ha_commit_trans_after_log_and_order SIGNAL con11_ready WAIT_FOR con11_continue"; > +send INSERT INTO t1 VALUES (21, REPEAT("x", 4100)); > + > +connection default; > +SET DEBUG_SYNC= "now WAIT_FOR con11_ready"; > +INSERT INTO t2 VALUES (4, "force binlog rotation"); > +--source include/show_binary_logs.inc > +--let $binlog_file= master-bin.000001 > +--source include/show_binlog_events.inc > +--let $binlog_file= master-bin.000002 > +--source include/show_binlog_events.inc > +--let $binlog_file= master-bin.000003 > +--source include/show_binlog_events.inc > + > +SET DEBUG_SYNC= "now SIGNAL con11_continue"; > + > +connection con11; > +reap; > + > +connection default; > +--echo con10 is still pending, no new binlog checkpoint should have been logged. > +--let $binlog_file= master-bin.000003 > +--source include/show_binlog_events.inc > + > +SET DEBUG_SYNC= "now SIGNAL con10_continue"; > + > +connection con10; > +reap; > + > +connection default; > +--echo No XIDs are pending, a new binlog checkpoint should have been logged. Could you also test recovery with multiple checkpoints in a binlog? E.g. binlog, like the above, then crash, and see that it uses the *last* checkpoint in the binlog. > +--let $binlog_file= master-bin.000003 > +--source include/show_binlog_events.inc > + > + > +# Cleanup > +connection default; > +DROP TABLE t1, t2; > +SET GLOBAL max_binlog_size= @old_max_binlog_size; > > === modified file 'sql/log.h' > --- sql/log.h 2012-01-16 19:16:35 +0000 > +++ sql/log.h 2012-05-14 09:51:40 +0000 > @@ -348,6 +348,15 @@ > time_t last_time; > }; > > +/* > + We assign each binlog file an internal ID, used to identify them for unlog(). > + Ids start from BINLOG_COOKIE_START; the value BINLOG_COOKIE_DUMMY is special > + meaning "no binlog" (we cannot use zero as that is reserved for error return > + from log_and_order). > +*/ > +#define BINLOG_COOKIE_DUMMY 1 > +#define BINLOG_COOKIE_START 2 Are these values completely internal and can be changed in any minor release without any compatibility concerns? Or they're written to binlog or something? > + > class binlog_cache_mngr; > class MYSQL_BIN_LOG: public TC_LOG, private MYSQL_LOG > { > @@ -388,10 +397,40 @@ > bool all; > }; > > + /* > + A list of struct xid_count_per_binlog is used to keep track of how many > + XIDs are in preared, but not committed, state in each binlog. "prepared" > + > + When count drops to zero in a binlog after rotation, it means that there > + are no more XIDs in prepared state, so that binlog is no longer needed > + for XA crash recovery, and we can log a new binlog checkpoint event. > + > + The list is protected against simultaneous access from multiple > + threads by LOCK_xid_list. > + */ > + struct xid_count_per_binlog : public ilink { > + char *binlog_name; > + uint binlog_name_len; > + ulong binlog_id; > + long xid_count; > + xid_count_per_binlog(); /* Give link error if constructor used. */ > + }; > + ulong current_binlog_id; > + I_List<xid_count_per_binlog> binlog_xid_count_list; Why I_List? > + /* > + When this is set, a RESET MASTER is in progress. > + > + Then we should not write any binlog checkpoints into the binlog (that > + could result in deadlock on LOCK_log, and we will delete all binlog files > + anyway). Instead we should signal COND_prep_xids whenever a new binlog COND_xid_list? > + checkpoint arrives - when all have arrived, RESET MASTER will complete. > + */ > + bool reset_master_pending; > + > /* LOCK_log and LOCK_index are inited by init_pthread_objects() */ > mysql_mutex_t LOCK_index; > - mysql_mutex_t LOCK_prep_xids; > - mysql_cond_t COND_prep_xids; > + mysql_mutex_t LOCK_xid_list; > + mysql_cond_t COND_xid_list; > mysql_cond_t update_cond; > ulonglong bytes_written; > IO_CACHE index_file; > @@ -415,7 +454,6 @@ It'd be great, if your diff would include function names (diff -p). An easy way of doing it is to use my diff_p plugin, see the last section in http://kb.askmonty.org/en/how-to-get-more-out-of-bzr-when-working-on-mariadb it's the most flexible, robust and universal solution that I have, and it puts function names in all bzr-generated diffs, not only in "bzr diff" command. > fix_max_relay_log_size). > */ > ulong max_size; > - long prepared_xids; /* for tc log - number of xids to remember */ > // current file sequence number for load data infile binary logging > uint file_id; > uint open_count; // For replication > === modified file 'sql/log_event.cc' > --- sql/log_event.cc 2012-05-09 07:35:46 +0000 > +++ sql/log_event.cc 2012-05-14 09:51:40 +0000 > @@ -5842,6 +5848,84 @@ > > > /************************************************************************** > + Binlog_checkpoint_log_event methods > +**************************************************************************/ > + > +#if defined(HAVE_REPLICATION) && !defined(MYSQL_CLIENT) > +void Binlog_checkpoint_log_event::pack_info(Protocol *protocol) > +{ > + protocol->store(binlog_file_name, binlog_file_len, &my_charset_bin); > +} > +#endif > + > + > +#ifdef MYSQL_CLIENT > +void Binlog_checkpoint_log_event::print(FILE *file, > + PRINT_EVENT_INFO *print_event_info) > +{ > + Write_on_release_cache cache(&print_event_info->head_cache, file, > + Write_on_release_cache::FLUSH_F); > + > + if (print_event_info->short_form) > + return; > + print_header(&cache, print_event_info, FALSE); > + my_b_printf(&cache, "\tBinlog checkpoint "); > + my_b_write(&cache, (uchar*)binlog_file_name, binlog_file_len); > + my_b_printf(&cache, "\n"); > +} > +#endif /* MYSQL_CLIENT */ > + > + > +#ifdef MYSQL_SERVER > +Binlog_checkpoint_log_event::Binlog_checkpoint_log_event( > + const char *binlog_file_name_arg, > + uint binlog_file_len_arg) > + :Log_event(), > + binlog_file_name(my_strndup(binlog_file_name_arg, binlog_file_len_arg, > + MYF(MY_WME))), > + binlog_file_len(binlog_file_len_arg) > +{ > + cache_type= EVENT_NO_CACHE; > +} > +#endif /* MYSQL_SERVER */ > + > + > +Binlog_checkpoint_log_event::Binlog_checkpoint_log_event( > + const char *buf, uint event_len, > + const Format_description_log_event *description_event) > + :Log_event(buf, description_event), binlog_file_name(0) > +{ > + uint8 header_size= description_event->common_header_len; > + uint8 post_header_len= > + description_event->post_header_len[BINLOG_CHECKPOINT_EVENT-1]; > + if (event_len < header_size + post_header_len || > + post_header_len < 4) I'd replace this 4 with BINLOG_CHECKPOINT_HEADER_LEN > + return; > + buf+= header_size; and here I'd add compile_time_assert(BINLOG_CHECKPOINT_HEADER_LEN == 4); // see uint4korr and int4store below > + binlog_file_len= uint4korr(buf); > + if (event_len - (header_size + post_header_len) < binlog_file_len) > + return; > + binlog_file_name= my_strndup(buf + post_header_len, binlog_file_len, > + MYF(MY_WME)); > + return; > +} > + > + > +#ifndef MYSQL_CLIENT > +bool Binlog_checkpoint_log_event::write(IO_CACHE *file) > +{ > + uchar buf[BINLOG_CHECKPOINT_HEADER_LEN]; > + int4store(buf, binlog_file_len); > + return write_header(file, BINLOG_CHECKPOINT_HEADER_LEN + binlog_file_len) || > + wrapper_my_b_safe_write(file, buf, BINLOG_CHECKPOINT_HEADER_LEN) || > + wrapper_my_b_safe_write(file, (const uchar *)binlog_file_name, > + binlog_file_len) || > + write_footer(file); > +} > +#endif /* MYSQL_CLIENT */ > + > + > +/************************************************************************** > Intvar_log_event methods > **************************************************************************/ > > === modified file 'sql/log.cc' > --- sql/log.cc 2012-04-10 06:28:13 +0000 > +++ sql/log.cc 2012-05-14 09:51:40 +0000 > @@ -2912,13 +2929,30 @@ > DBUG_ENTER("cleanup"); > if (inited) > { > + xid_count_per_binlog *b; > + > inited= 0; > close(LOG_CLOSE_INDEX|LOG_CLOSE_STOP_EVENT); > delete description_event_for_queue; > delete description_event_for_exec; > + > + while ((b= binlog_xid_count_list.get())) 'get' - what a bad name. I checked the implementation and now I know what it's doing, but absoletely not what one would expect :( It's not your fault, of course, just ranting... > + { > + /* > + There should be no pending XIDs at shutdown, and only one entry (for > + the active binlog file) in the list. > + */ > + DBUG_ASSERT(b->xid_count == 0); > + DBUG_ASSERT(!binlog_xid_count_list.head()); > + my_free(b); > + } > + > mysql_mutex_destroy(&LOCK_log); > mysql_mutex_destroy(&LOCK_index); > + mysql_mutex_destroy(&LOCK_xid_list); > mysql_cond_destroy(&update_cond); > + mysql_cond_destroy(&COND_queue_busy); > + mysql_cond_destroy(&COND_xid_list); > } > DBUG_VOID_RETURN; > } > @@ -3506,6 +3587,42 @@ > mysql_mutex_lock(&LOCK_log); > mysql_mutex_lock(&LOCK_index); > > + if (!is_relay_log) > + { > + /* > + We are going to nuke all binary log files. > + So first wait until all pending binlog checkpoints have completed. > + */ > + mysql_mutex_lock(&LOCK_xid_list); > + xid_count_per_binlog *b; > + reset_master_pending= true; > + for (;;) > + { > + I_List_iterator<xid_count_per_binlog> it(binlog_xid_count_list); > + while ((b= it++)) > + { > + if (b->xid_count > 0) > + break; > + } > + if (!b) > + break; /* No more pending XIDs */ > + /* > + Wait until signalled that one more binlog dropped to zero, then check > + again. > + */ > + mysql_cond_wait(&COND_xid_list, &LOCK_xid_list); > + } I would've done it simpler, like while (b= binlog_xid_count_list.head()) { while (b->xid_count > 0) mysql_cond_wait(&COND_xid_list, &LOCK_xid_list); my_free(binlog_xid_count_list.get()); } > + > + /* > + Now all XIDs are fully flushed to disk, and we are holding LOCK_log so > + no new ones will be written. So we can proceed to delete the logs. > + */ > + while ((b= binlog_xid_count_list.get())) > + my_free(b); > + reset_master_pending= false; > + mysql_mutex_unlock(&LOCK_xid_list); > + } > + > /* > The following mutex is needed to ensure that no threads call > 'delete thd' as we would then risk missing a 'rollback' from this > @@ -5761,6 +5877,37 @@ > DBUG_RETURN(error); > } > > +void > +MYSQL_BIN_LOG::write_binlog_checkpoint_event_already_locked(const char *name, > + uint len) > +{ > + Binlog_checkpoint_log_event ev(name, len); > + /* > + Note that we must sync the binlog checkpoint to disk. > + Otherwise a subsequent log purge could delete binlogs that XA recovery > + thinks are needed (even though they are not really). > + */ > + if (!ev.write(&log_file) && !flush_and_sync(0)) > + { > + bool check_purge= false; > + signal_update(); > + rotate(false, &check_purge); > + if (check_purge) > + purge(); why "write binlog checkpoint event" implies signal_update(), rotate(), and purge() ? > + return; > + } > + > + /* > + If we fail to write the checkpoint event, something is probably really > + bad with the binlog. We complain in the error log. > + Note that failure to write binlog checkpoint does not compromise the > + ability to do crash recovery - crash recovery will just have to scan a > + bit more of the binlog than strictly necessary. > + */ > + sql_print_error("Failed to write binlog checkpoint event to binary log\n"); > +} > + > + > /** > Write a cached log entry to the binary log. > - To support transaction over replication, we wrap the transaction > @@ -6061,11 +6207,11 @@ > { > /* > If we fail to rotate, which thread should get the error? > - We give the error to the *last* transaction thread; that seems to > - make the most sense, as it was the last to write to the log. > + We give the error to the leader, as any my_error() thrown inside > + rotate() will have been registered for the leader THD. > */ > - last_in_queue->error= ER_ERROR_ON_WRITE; > - last_in_queue->commit_errno= errno; > + leader->error= ER_ERROR_ON_WRITE; > + leader->commit_errno= errno; Where did that change happen? Is it related to MDEV-181 at all, or it's an unrelated bugfix? > check_purge= false; > } > } > @@ -6082,9 +6228,6 @@ > */ > mysql_mutex_unlock(&LOCK_log); > > - if (check_purge) > - purge(); Why was it moved down? > - > DEBUG_SYNC(leader->thd, "commit_after_release_LOCK_log"); > ++num_group_commits; > > @@ -7427,11 +7560,23 @@ > > cache_mngr->using_xa= TRUE; > cache_mngr->xa_xid= xid; > +#ifndef DBUG_OFF > + cache_mngr->cookie= 0; > +#endif > err= binlog_commit_flush_xid_caches(thd, cache_mngr, all, xid); > > DEBUG_SYNC(thd, "binlog_after_log_and_order"); > > - DBUG_RETURN(!err); > + if (err) > + DBUG_RETURN(0); > + /* > + If using explicit user XA, we will not have XID. We must still return a > + non-zero cookie (as zero cookie signals error). > + */ > + if (!xid) > + DBUG_RETURN(1); s/1/BINLOG_COOKIE_DUMMY/ > + DBUG_ASSERT(cache_mngr->cookie != 0); > + DBUG_RETURN(cache_mngr->cookie); > } > > /* > @@ -7446,40 +7591,135 @@ > binary log. > */ > void > -TC_LOG_BINLOG::mark_xids_active(uint xid_count) > +TC_LOG_BINLOG::mark_xids_active(ulong cookie, uint xid_count) > { > + xid_count_per_binlog *b; > + > DBUG_ENTER("TC_LOG_BINLOG::mark_xids_active"); > - DBUG_PRINT("info", ("xid_count=%u", xid_count)); > - mysql_mutex_lock(&LOCK_prep_xids); > - prepared_xids+= xid_count; > - mysql_mutex_unlock(&LOCK_prep_xids); > + DBUG_PRINT("info", ("cookie=%lu xid_count=%u", cookie, xid_count)); > + DBUG_ASSERT(cookie != 0 && cookie != BINLOG_COOKIE_DUMMY); > + > + mysql_mutex_lock(&LOCK_xid_list); > + I_List_iterator<xid_count_per_binlog> it(binlog_xid_count_list); > + while ((b= it++)) > + { > + if (b->binlog_id == cookie) > + { > + b->xid_count += xid_count; > + break; > + } > + } > + /* > + As we do not delete elements until count reach zero, elements should always > + be found. > + */ > + DBUG_ASSERT(b); > + mysql_mutex_unlock(&LOCK_xid_list); > DBUG_VOID_RETURN; > } > > /* > - Once an XID is committed, it is safe to rotate the binary log, as it can no > - longer be needed during crash recovery. > + Once an XID is committed, it can no longer be needed during crash recovery, > + as it has been durably recorded on disk as "committed". > > This function is called to mark an XID this way. It needs to decrease the > - count of pending XIDs, and signal the log rotator thread when it reaches zero. > + count of pending XIDs in the corresponding binlog. When the count reaches > + zero (for an "old" binlog that is not the active one), that binlog file no > + longer need to be scanned during crash recovery, so we can log a new binlog > + checkpoint. > */ > void > -TC_LOG_BINLOG::mark_xid_done() > +TC_LOG_BINLOG::mark_xid_done(ulong cookie) > { > - my_bool send_signal; > + xid_count_per_binlog *b; > + bool first; > + ulong current; > > DBUG_ENTER("TC_LOG_BINLOG::mark_xid_done"); > - mysql_mutex_lock(&LOCK_prep_xids); > - // prepared_xids can be 0 if the transaction had ignorable errors. > - DBUG_ASSERT(prepared_xids >= 0); > - if (prepared_xids > 0) > - prepared_xids--; > - send_signal= (prepared_xids == 0); > - mysql_mutex_unlock(&LOCK_prep_xids); > - if (send_signal) { > - DBUG_PRINT("info", ("prepared_xids=%lu", prepared_xids)); > - mysql_cond_signal(&COND_prep_xids); > - } > + if (cookie == BINLOG_COOKIE_DUMMY) > + DBUG_VOID_RETURN; /* Nothing to do. */ > + > + mysql_mutex_lock(&LOCK_xid_list); > + current= current_binlog_id; > + I_List_iterator<xid_count_per_binlog> it(binlog_xid_count_list); > + first= true; > + while ((b= it++)) > + { > + if (b->binlog_id == cookie) > + { > + --b->xid_count; > + break; > + } > + first= false; > + } > + /* Binlog is always found, as we do not remove until count reaches 0 */ > + DBUG_ASSERT(b); > + if (likely(cookie == current && !reset_master_pending) || > + b->xid_count != 0 || !first) > + { > + /* No new binlog checkpoint reached yet. */ > + mysql_mutex_unlock(&LOCK_xid_list); > + DBUG_VOID_RETURN; > + } > + > + /* > + Now log a binlog checkpoint for the first binlog file with a non-zero count. > + > + Note that it is possible (though perhaps unlikely) that when count of > + binlog (N-2) drops to zero, binlog (N-1) is already at zero. So we may > + need to skip several entries before we find the one to log in the binlog > + checkpoint event. > + > + We chain the locking of LOCK_xid_list and LOCK_log, so that we ensure that > + Binlog_checkpoint_events are logged in order. This simplifies recovery a > + bit, as it can just take the last binlog checkpoint in the log, rather > + than compare all found against each other to find the one pointing to the > + most recent binlog. > + > + Note also that we need to first release LOCK_xid_list, then aquire > + LOCK_log, then re-aquire LOCK_xid_list. If we were to take LOCK_log while > + holding LOCK_xid_list, we might deadlock with other threads that take the > + locks in the opposite order. > + > + If a RESET MASTER is pending, we are about to remove all log files, and > + the RESET MASTER thread is waiting for all pending unlog() calls to > + complete while holding LOCK_log. In this case we should not log a binlog > + checkpoint event (it would be deleted immediately anywat and we would > + deadlock on LOCK_log) but just signal the thread. > + */ > + if (!reset_master_pending) > + { > + mysql_mutex_unlock(&LOCK_xid_list); > + mysql_mutex_lock(&LOCK_log); > + mysql_mutex_lock(&LOCK_xid_list); > + } > + for (;;) > + { > + /* Remove initial element(s) with zero count. */ > + b= binlog_xid_count_list.get(); Eh. You need to check whether binlog_xid_count_list.head()->xid_count == 0 before removing it like that. When you released LOCK_xid_list another thread could've taken over and deleted all zero count elements. So now you could find that binlog_xid_count_list.head() has non-zero count. > + my_free(b); > + b= binlog_xid_count_list.head(); > + /* > + Normally, we must not remove all elements in the list. > + Only if a RESET MASTER is in progress may we delete everything - RESET > + MASTER has LOCK_log held, and will create a new initial element before > + releasing the lock. > + */ > + DBUG_ASSERT(b || reset_master_pending); > + if (unlikely(!b) || likely(b->binlog_id == current) || b->xid_count > 0) > + break; > + } > + if (reset_master_pending) > + { > + mysql_cond_signal(&COND_xid_list); > + mysql_mutex_unlock(&LOCK_xid_list); > + DBUG_VOID_RETURN; > + } > + > + mysql_mutex_unlock(&LOCK_xid_list); > + write_binlog_checkpoint_event_already_locked(b->binlog_name, > + b->binlog_name_len); > + mysql_mutex_unlock(&LOCK_log); > DBUG_VOID_RETURN; > } > > @@ -7507,19 +7755,104 @@ > > fdle->flags&= ~LOG_EVENT_BINLOG_IN_USE_F; // abort on the first error > > - while ((ev= Log_event::read_log_event(log, 0, fdle, > - opt_master_verify_checksum)) > - && ev->is_valid()) > + /* > + Scan the binlog for XIDs that need to be committed if still in the > + prepared stage. > + > + Start with the latest binlog file, then continue with any other binlog > + files if the last found binlog checkpoint indicates it is needed. > + */ > + > + first_round= true; > + for (;;) > { > - if (ev->get_type_code() == XID_EVENT) > - { > - Xid_log_event *xev=(Xid_log_event *)ev; > - uchar *x= (uchar *) memdup_root(&mem_root, (uchar*) &xev->xid, > - sizeof(xev->xid)); > - if (!x || my_hash_insert(&xids, x)) > + while ((ev= Log_event::read_log_event(first_round ? first_log : &log, > + 0, fdle, opt_master_verify_checksum)) > + && ev->is_valid()) > + { > + switch (ev->get_type_code()) > + { > + case XID_EVENT: > + { > + Xid_log_event *xev=(Xid_log_event *)ev; > + uchar *x= (uchar *) memdup_root(&mem_root, (uchar*) &xev->xid, > + sizeof(xev->xid)); > + if (!x || my_hash_insert(&xids, x)) > + { > + delete ev; > + goto err2; > + } > + break; > + } > + case BINLOG_CHECKPOINT_EVENT: > + if (first_round) > + { > + Binlog_checkpoint_log_event *cev= (Binlog_checkpoint_log_event *)ev; > + if (cev->binlog_file_len >= FN_REFLEN) > + { > + sql_print_error("Incorrect binlog checkpoint event with too long " > + "file name found. Aborting."); This is a recovery. After a crash. You can be more tolerant here. In this particular case there is no need to abort, you can safely ignore incorrect checkpoints. This only means that the recovery will start from the previous checkpoint and it'll take longer. > + delete ev; > + goto err2; > + } > + binlog_checkpoint_len= cev->binlog_file_len; > + memcpy(binlog_checkpoint_name, cev->binlog_file_name, > + binlog_checkpoint_len); > + binlog_checkpoint_name[binlog_checkpoint_len]= '\0'; > + break; > + } > + default: > + /* Nothing. */ > + break; > + } > + delete ev; > + } > + > + /* > + If the last binlog checkpoint event points to an older log, we have to > + scan all logs from there also, to get all possible XIDs to recover. > + > + If there was no binlog checkpoint event at all, this means the log was > + written by an older version of MariaDB (or MySQL) - these always have an > + (implicit) binlog checkpoint event at the start of the last binlog file. > + */ > + if (first_round) > + { > + if (!binlog_checkpoint_len) > + break; > + first_round= false; > + if (find_log_pos(linfo, binlog_checkpoint_name, 1)) > + { > + sql_print_error("Binlog file '%s' not found in binlog index, needed " > + "for recovery. Aborting.", binlog_checkpoint_name); > goto err2; > - } > - delete ev; > + } > + } > + else > + { > + end_io_cache(&log); > + mysql_file_close(file, MYF(MY_WME)); > + file= -1; > + } > + > + if (0 == strcmp(linfo->log_file_name, last_log_name)) > + break; // No more files to do > + if ((file= open_binlog(&log, linfo->log_file_name, &errmsg)) < 0) > + { > + sql_print_error("%s", errmsg); > + goto err2; > + } > + /* > + We do not need to read the Format_description_log_event of other binlog > + files. It is not possible for a binlog checkpoint to span multiple > + binlog files written by different versions of the server. So we can use > + the first one read for reading from all binlog files. > + */ > + if (find_next_log(linfo, 1)) > + { > + sql_print_error("Error reading binlog files during recovery. Aborting."); > + goto err2; > + } > } > > if (ha_recover(&xids)) > === modified file 'sql/sql_repl.cc' > --- sql/sql_repl.cc 2012-04-27 08:20:38 +0000 > +++ sql/sql_repl.cc 2012-05-14 09:51:40 +0000 > @@ -624,6 +624,30 @@ > } > > /* > + Do not send binlog checkpoint events to a slave that does not understand it. > + */ > + if (unlikely(event_type == BINLOG_CHECKPOINT_EVENT) && > + mariadb_slave_capability < MARIA_SLAVE_CAPABILITY_BINLOG_CHECKPOINT) > + { > + if (mariadb_slave_capability >= MARIA_SLAVE_CAPABILITY_TOLERATE_HOLES) > + { > + /* This slave can tolerate events omitted from the binlog stream. */ > + return NULL; > + } > + else > + { > + /* > + The slave does not understand BINLOG_CHECKPOINT_EVENT. Send a dummy > + event instead, with same length so slave does not get confused about > + binlog positions. > + */ Hm. Above and below the code that you've added events are simply skipped, without checks and dummy events. Why do you bother to replace with dummies? > + if (Query_log_event::dummy_event(packet, ev_offset, current_checksum_alg)) > + return "Failed to replace binlog checkpoint event with dummy: " > + "too small event."; > + } > + } > + > + /* > Skip events with the @@skip_replication flag set, if slave requested > skipping of such events. > */ Regards, Sergei

2 1

[Maria-developers] Fwd: crash in slave replication with XA transaction
by Zardosht Kasheff 11 Jun '12

11 Jun '12

Hello all, We see this on Maria 5.2 and Maria 5.5 as well. Does the following fix work? -Zardosht ---------- Forwarded message ---------- From: Rich Prohaska <prohaska(a)tokutek.com> Date: Mon, Jun 4, 2012 at 1:23 PM Subject: crash in slave replication with XA transaction To: internals(a)lists.mysql.com Hello, We are running MySQL 5.5.21 with both InnoDB and TokuDB as XA storage engines. When one runs a transaction that inserts into an InnoDB table and a TokuDB table, the commit crashes the replication slave at log.cc line 6602 because cache_mngr == NULL. We found a pattern in various log functions in log.cc that calls binlog_setup_trx_data to initialize the txn state. This change fixes our immediate problem. $ diff log.cc.orig log.cc 6595a6596,6597 > if (thd_get_ha_data(thd, binlog_hton) == NULL) > thd->binlog_setup_trx_data(); Is this a correct fix? If so, are there other places that need a similar change? Thanks Rich Prohaska The transaction is: create table t1 (a int primary key) engine=innodb; create table t2 (a int primary key) engine=tokudb; set autocommit=0; insert into t1 values (1); insert into t2 values (2); commit; the slave crashes while handling the commit binlog entry. -- MySQL Internals Mailing List For list archives: http://lists.mysql.com/internals To unsubscribe: http://lists.mysql.com/internals

3 5

Re: [Maria-developers] [Commits] Rev 3426: LP1008334 : Speedup specific datetime queries that got slower with introduction of microseconds in file:///H:/bzr/5.5/
by Sergei Golubchik 08 Jun '12

08 Jun '12

Hi, Vladislav! On Jun 08, Vladislav Vaintroub wrote: > ------------------------------------------------------------ > revno: 3426 > revision-id: wlad(a)montyprogram.com-20120608000841-ieo4vumio11pc8p9 > parent: wlad(a)montyprogram.com-20120530182054-g7sll9u4gsitvgqe > fixes bug(s): https://launchpad.net/bugs/1008334 > committer: Vladislav Vaintroub <wlad(a)montyprogram.com> > branch nick: 5.5 > timestamp: Fri 2012-06-08 02:08:41 +0200 > message: > LP1008334 : Speedup specific datetime queries that got slower with introduction of microseconds > > - Item::get_seconds() now skips decimal arithmetic, if decimals is 0. This significantly speeds up from_unixtime() if no fractional part is passed. > - replace sprintfs used to format temporal values by hand-coded formatting Looks ok. Would be good to mention the ~25% speedup in the comment above fmt_number. Like "Gives 25% speedup when converting temporal values to strings, as compared to sprintf". So that one wouldn't need to dig up bzr history every time. > === modified file 'sql/item.cc' > --- a/sql/item.cc 2012-05-21 18:54:41 +0000 > +++ b/sql/item.cc 2012-06-08 00:08:41 +0000 > @@ -1265,7 +1265,7 @@ > > bool Item::get_seconds(ulonglong *sec, ulong *sec_part) > { > - if (result_type() == INT_RESULT) > + if (decimals == 0 || result_type() == INT_RESULT) I'd expect that checking for decimals==0 alone should be enough. Could you try to run the test suite with the DBUG_ASSERT(result_type() != INT_RESULT || decimals == 0); please? > { // optimize for an important special case > longlong val= val_int(); > bool neg= val < 0 && !unsigned_flag; Regards, Sergei

1 0

Re: [Maria-developers] [Commits] Rev 3429: MDEV-329: MariaDB 5.5 does not use fdatasync(). in http://bazaar.launchpad.net/~maria-captains/maria/5.5
by Kristian Nielsen 08 Jun '12

08 Jun '12

Hi wlad, Can I get you to review the following one-liner? (Just the Cmake part). The problem is that the code in 5.5 currently does not use fdatasync(), falling back to fsync(). The fdatasync() is detected correctly by cmake, but the HAVE_DECL_FDATASYNC symbol does not get defined so code does not use it. Is adding HAVE_DECL_FDATASYNC to config.h.cmake the correct fix? - Kristian. knielsen(a)knielsen-hq.org writes: > At http://bazaar.launchpad.net/~maria-captains/maria/5.5 > > ------------------------------------------------------------ > revno: 3429 > revision-id: knielsen(a)knielsen-hq.org-20120608091856-jmxb22ehrdp8xjdo > parent: timour(a)askmonty.org-20120606131948-0k4ih4get2fttge0 > committer: knielsen(a)knielsen-hq.org > branch nick: mariadb-5.5 > timestamp: Fri 2012-06-08 11:18:56 +0200 > message: > MDEV-329: MariaDB 5.5 does not use fdatasync(). > > The --debug-no-sync incorrectly defaulted to ON, disabling sync calls > by default which can loose data or cause corruption. Also, the code > used fsync() instead of the sometimes more efficient fdatasync(). > === modified file 'config.h.cmake' > --- a/config.h.cmake 2012-04-19 02:02:28 +0000 > +++ b/config.h.cmake 2012-06-08 09:18:56 +0000 > @@ -150,6 +150,7 @@ > #cmakedefine HAVE_FCNTL 1 > #cmakedefine HAVE_FCONVERT 1 > #cmakedefine HAVE_FDATASYNC 1 > +#cmakedefine HAVE_DECL_FDATASYNC 1 > #cmakedefine HAVE_FESETROUND 1 > #cmakedefine HAVE_FINITE 1 > #cmakedefine HAVE_FP_EXCEPT 1 > > === modified file 'sql/mysqld.cc' > --- a/sql/mysqld.cc 2012-05-21 18:54:41 +0000 > +++ b/sql/mysqld.cc 2012-06-08 09:18:56 +0000 > @@ -6161,7 +6161,7 @@ struct my_option my_long_options[]= > #endif > {"debug-no-sync", 0, > "Disables system sync calls. Only for running tests or debugging!", > - &my_disable_sync, &my_disable_sync, 0, GET_BOOL, NO_ARG, 1, 0, 0, 0, 0, 0}, > + &my_disable_sync, &my_disable_sync, 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0}, > #ifdef HAVE_REPLICATION > {"debug-sporadic-binlog-dump-fail", 0, > "Option used by mysql-test for debugging and testing of replication.", > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits

2 2

[Maria-developers] shrink() in sql_string.h
by MARK CALLAGHAN 07 Jun '12

07 Jun '12

Why doesn't the shrink() method in sql_string.h check the value of alloced before calling my_realloc? my_realloc might free Ptr, but when alloced=0 that memory is not owned by the string. 2 bk@work | inline void shrink(uint32 arg_length) // Shrink buffer 2 bk@work | { 2 bk@work | if (arg_length < Alloced_length) 2 bk@work | { 2 bk@work | char *new_ptr; 2 bk@work | if (!(new_ptr=(char*) my_realloc(Ptr,arg_length,MYF(0)))) 2 bk@work | { 228 sasha@m | Alloced_length = 0; 2 bk@work | real_alloc(arg_length); 2 bk@work | } 2 bk@work | else 2 bk@work | { 2 bk@work | Ptr=new_ptr; 2 bk@work | Alloced_length=arg_length; 2 bk@work | } 2 bk@work | } 2 bk@work | } -- Mark Callaghan mdcallag(a)gmail.com

2 2

[Maria-developers] SHOW EXPLAIN: All concurrency problems should be gone, please test
by Sergei Petrunia 07 Jun '12

07 Jun '12

Hi Elena, I've pushed the fix for all known SHOW EXPLAIN problems. I think, another round of testing is needed. You can do it right away, or wait until I resolve the open "different query output" problems. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] Please review: [Commits] Rev 3535: Fix for bug lp:1006231 in file:///home/tsk/mprog/src/5.3/
by Sergei Petrunia 01 Jun '12

01 Jun '12

Hi Timour, As discussed on the phone call: - Let's add a separate MDEV item to add "the assert". - Ok to push tis patch. On Wed, May 30, 2012 at 07:12:33PM +0300, Timour Katchaounov wrote: > Sergey, > > Could you please review yet another bug fix for expensive subqueries > executed during optimiziation. > > Timour > > ------------------------------------------------------------ > revno: 3535 > revision-id: timour(a)askmonty.org-20120530161018-kgsiqesmnohtnbc6 > parent: holyfoot(a)askmonty.org-20120529045925-h35akgdit83ql5yq > fixes bug(s): https://launchpad.net/bugs/1006231 > committer: timour(a)askmonty.org > branch nick: 5.3 > timestamp: Wed 2012-05-30 19:10:18 +0300 > message: > Fix for bug lp:1006231 > > Analysis: > > When a subquery that needs a temp table is executed during > the prepare or optimize phase of the outer query, at the end > of the subquery execution all the JOIN_TABs of the subquery > are replaced by a new JOIN_TAB that selects from the temp table. > However that temp table has no corresponding TABLE_LIST. > Once EXPLAIN execution reaches its last phase, it tries to print > the names of the subquery tables through its TABLE_LISTs, but in > the case of this bug there is no such TABLE_LIST (it is NULL), > hence a crash. > > Solution: > The fix is to block subquery evaluation inside > Item_func_like::fix_fields and Item_func_like::select_optimize() > using the Item::is_expensive() test. > === modified file 'mysql-test/r/subselect_innodb.result' > --- a/mysql-test/r/subselect_innodb.result 2012-05-20 12:57:29 +0000 > +++ b/mysql-test/r/subselect_innodb.result 2012-05-30 16:10:18 +0000 > @@ -365,4 +365,19 @@ GROUP BY 1 > MAX( f1 ) > NULL > drop table t1, t2, t3; > +# > +# LP BUG#1006231 crash in select_describe > +# > +create table t1(a1 int) ENGINE=InnoDB; > +insert into t1 values (1); > +explain > +select 1 from t1 where 1 like (select 1 from t1 where 1 <=> (select 1 from t1 group by a1)); > +id select_type table type possible_keys key key_len ref rows Extra > +1 PRIMARY t1 ALL NULL NULL NULL NULL 1 > +2 SUBQUERY t1 ALL NULL NULL NULL NULL 1 > +3 SUBQUERY t1 ALL NULL NULL NULL NULL 1 Using temporary; Using filesort > +select 1 from t1 where 1 like (select 1 from t1 where 1 <=> (select 1 from t1 group by a1)); > +1 > +1 > +drop table t1; > set optimizer_switch=@subselect_innodb_tmp; > > === modified file 'mysql-test/t/subselect_innodb.test' > --- a/mysql-test/t/subselect_innodb.test 2012-05-20 12:57:29 +0000 > +++ b/mysql-test/t/subselect_innodb.test 2012-05-30 16:10:18 +0000 > @@ -356,4 +356,16 @@ WHERE f2 >= ( > > drop table t1, t2, t3; > > +--echo # > +--echo # LP BUG#1006231 crash in select_describe > +--echo # > + > +create table t1(a1 int) ENGINE=InnoDB; > +insert into t1 values (1); > +explain > +select 1 from t1 where 1 like (select 1 from t1 where 1 <=> (select 1 from t1 group by a1)); > +select 1 from t1 where 1 like (select 1 from t1 where 1 <=> (select 1 from t1 group by a1)); > +drop table t1; > + > + > set optimizer_switch=@subselect_innodb_tmp; > > === modified file 'sql/item_cmpfunc.cc' > --- a/sql/item_cmpfunc.cc 2012-05-20 12:57:29 +0000 > +++ b/sql/item_cmpfunc.cc 2012-05-30 16:10:18 +0000 > @@ -4656,7 +4656,7 @@ longlong Item_func_like::val_int() > > Item_func::optimize_type Item_func_like::select_optimize() const > { > - if (args[1]->const_item()) > + if (args[1]->const_item() && !args[1]->is_expensive()) > { > String* res2= args[1]->val_str((String *)&cmp.value2); > const char *ptr2; > @@ -4743,7 +4743,8 @@ bool Item_func_like::fix_fields(THD *thd > We could also do boyer-more for non-const items, but as we would have to > recompute the tables for each row it's not worth it. > */ > - if (args[1]->const_item() && !use_strnxfrm(collation.collation)) > + if (args[1]->const_item() && !use_strnxfrm(collation.collation) && > + !args[1]->is_expensive()) > { > String* res2 = args[1]->val_str(&cmp.value2); > if (!res2) > > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] WL#252 New (by DonQuichote): IF (NOT) EXIST clauses for ALTER TABLE
by worklog-noreply＠askmonty.org 01 Jun '12

01 Jun '12

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: IF (NOT) EXIST clauses for ALTER TABLE CREATION DATE..: Fri, 01 Jun 2012, 09:35 SUPERVISOR.....: IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 252 (https://askmonty.org/worklog/?tid=252) VERSION........: WorkLog-4.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: MySQL / MariaDB SQL is great as a functional language: you tell what you want and not how to do it. This includes suppressing of errors when the situation you want already exists. So there is a CREATE TABLE IF NOT EXISTS statement, for example (you say that you want the table to exist, not that it has to be built right now). It would be nice if altering a table could be done the same way: ALTER TABLE SomeTable ADD COLUMN IF NOT EXISTS SomeColumn ... ALTER TABLE SomeTable DROP COLUMN IF EXISTS SomeColumn ... ALTER TABLE SomeTable DROP FOREIGN KEY IF EXISTS SomeConstraint ... This would make definition scripts more legible and more friendly to existing databases. It would facilitate what I call "repeatable scripts", that can both create a new database and update an existing one to the latest structure. Mind you, this is already possible using temporary stored procedures (as shown in http://www.howtoforge.com/node/4833) but it would make definition scripts more maintainable and the SQL language more consistent. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v4.0.0)

1 0

Re: [Maria-developers] MariaDB 5.5.24 release prep
by Sergei Golubchik 31 May '12

31 May '12

Hi, Jocelyn! On May 31, Jocelyn Fournier wrote: > > Just to be sure : it means also the debian package will be properly > linked against libaio ? (it wasn't the case in 5.5.23 release). Unfortunately not. I mean that debian packages will be linked against libaio only in 5.5.25, but not in 5.5.24. Currently (in 5.5.24) only RPM packages are linked against libaio. Debian packages and bintar package are not. Regards, Sergei

1 0

Re: [Maria-developers] MariaDB 5.5.24 release prep
by Sergei Golubchik 31 May '12

31 May '12

Hi, md! On May 31, md wrote: > Sergei: > > I guess you can get 5.5.24 to build ok. I had a bunch of trouble with > 5.5.23 and a compiler error that I could not seem to make go away. > > [ 47%] Building CXX object > storage/oqgraph/CMakeFiles/oqgraph.dir/graphcore.cc.o > > In file included from > /usr/include/boost/smart_ptr/detail/sp_counted_base_gcc_x86.hpp:27:0, This is a known issue. An the OQGraph engine that we include is incompatible with the new Boost. It is a third-party engine, not ours. OQGraph developer knows about this problem and he's working on it. When he'll fix OQGraph, we will update it in our sources. Meanwhile we build without. Try to run cmake -DWITHOUT_OQGRAPH=1 in your build tree, and run make again. In 5.5.25 I'll make this the default behavior, until OQGraph is fixed. Regards, Sergei

1 0

Re: [Maria-developers] MariaDB 5.5.24 release prep
by Sergei Golubchik 31 May '12

31 May '12

Hi, Jocelyn! On May 31, Jocelyn Fournier wrote: > > Just wondering if the build issue with binaries and packages which don't > use native AIO is fixed as well ? > > (See https://bugs.launchpad.net/maria/+bug/988204 ) Doesn't seem so :( For some reason it was submitted in the wrong bugtracker, and I missed it when preparing a release. Sorry for this, it'll be in 5.5.25. This time - for sure. Note that 5.5 RPM packages are not affected by this bug - they are all properly linked with AIO. Regards, Sergei

1 0

[Maria-developers] MariaDB 5.5.24 release prep
by Daniel Bartholomew 31 May '12

31 May '12

All, I'm uploading the files for MariaDB 5.5.24 and getting other bits and pieces ready. Preliminary versions of the Release Notes and Changelog are available here: http://kb.askmonty.org/en/mariadb-5524-release-notes/ http://kb.askmonty.org/en/mariadb-5524-changelog/ The big news in this release is the new RPM packages and the YUM repositories. I'm still working on getting all the various bits set up, but it's all coming together and with any luck, I should be able to flip the switch and announce the release tomorrow morning (U.S. Eastern Time). In the mean time, if there is anything which should be mentioned in the Release Notes, please let me know ASAP or jump in and edit the Release Notes directly. Thanks! -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

Re: [Maria-developers] Support of the extended syntax for ANALYZE
by Sergei Golubchik 26 May '12

26 May '12

Hi, Igor! On May 10, Igor Babaev wrote: > Serg, > > Here's the patch we talked about on IRC today. ... > +ANALYZE TABLE t1(COLUMN(), INDEX()); > +ANALYZE TABLE t1(COLUMN(c,e,b), INDEX(idx2,idx4)); > +ANALYZE TABLE t1(COLUMN(*), INDEX(*)); I see what you want to do. I think it'll be more SQL-like, if it'll be more verbose, for example ANALYZE TABLE t1 FOR COLUMNS (c,e,d) INDEXES (idx2, idx4); I would prefer to drop parentheses too, but then the grammar will become ambiguous :( And instead of double (*) it could simply be ANALYZE TABLE t1 FULL; or ALL; it'd be similar to CHECK TABLE t1 EXTENDED; REPAIR TABLE t1 QUICK; etc Alternatively, it could have an explicit "PERSISTENT" keyword: ANALYZE TABLE t1 PERSISTENT; -- meaning all columns and indexes ANALYZE TABLE t1 PERSISTENT FOR COLUMNS (c,e,d) INDEXES (idx2, idx4); That's probably the best one. Reasonably SQL-like. And the most common use case (analyze all) is short, it's clearly the main, default behavior. And the clause to specify columns and indexes extends the default syntax. Regards, Sergei

4 6

[Maria-developers] SHOW EXPLAIN and LOCK_thd_data
by Sergei Petrunia 26 May '12

26 May '12

Hi Sergei, I've got a problem with SHOW EXPLAIN and LOCK_thd_data. SHOW EXPLAIN was modeled after the KILL command: it's actions are: lock LOCK_thread_count; find the thread $target_thr we need explain for; lock $target_thr->LOCK_thread_count; unlock LOCK_thread_count; send SHOW EXPLAIN request to $target_thr. wait until request is executed (or discarded) unlock $target_thr->LOCK_thread_count; This scheme works, as long as the target thread doesn't select from information_schema tables. With queries over I_S tables, it breaks, because I_S table may try to acquire current_thd->LOCK_thread_count during execution. Below I provide a stacktrace of one such case, line numbers are from lp:~maria-captains/maria/5.5-show-explain. My first thought was to change LOCK_thread_count to be a read/write mutex, but I wanted to hear your opinion, too. (gdb) wher #0 inline_mysql_mutex_lock (that=0xa1111d4, src_file=0x894b75c "sql/sql_class.h", src_line=3020) at include/mysql/psi/mysql_thread.h:608 #1 0x081dda85 in THD::set_open_tables (this=0xa1104b0, open_tables_arg=0xa12de10) at sql/sql_class.h:3020 #2 0x081d9024 in open_table (thd=0xa1104b0, table_list=0xa1289b0, mem_root=0x99134000, ot_ctx=0x99134020) at sql/sql_base.cc:3180 #3 0x081d9aee in open_and_process_table (thd=0xa1104b0, lex=0x9913417c, tables=0xa1289b0, counter=0x991340bc, flags=1026, prelocking_strategy=0x991340c0, has_prelocking_list=false, ot_ctx=0x99134020, new_frm_mem=0x99134000) at sql/sql_base.cc:4509 #4 0x081da269 in open_tables (thd=0xa1104b0, start=0x991340d4, counter=0x991340bc, flags=1026, prelocking_strategy=0x991340c0) at sql/sql_base.cc:4958 #5 0x081da6de in open_normal_and_derived_tables (thd=0xa1104b0, tables=0xa1289b0, flags=1026, dt_phases=34) at sql/sql_base.cc:5617 #6 0x082a3e23 in fill_schema_table_by_open (thd=0xa1104b0, is_show_fields_or_keys=true, table=0xa1263c0, schema_table=0x8c262b0, orig_db_name=0x99134e04, orig_table_name=0x99134e1c, open_tables_state_backup=0x99134d94, can_deadlock=false) at sql/sql_show.cc:3659 #7 0x082b8638 in get_all_tables (thd=0xa1104b0, tables=0xa11d560, cond=0x0) at sql/sql_show.cc:4214 #8 0x0829f943 in do_fill_table (thd=0xa1104b0, table_list=0xa11d560, join_table=0xa128548) at sql/sql_show.cc:7503 #9 0x0829fce2 in get_schema_tables_result (join=0xa1279a8, executed_place=PROCESSED_BY_JOIN_EXEC) at sql/sql_show.cc:7610 #10 0x0829468d in JOIN::exec_inner (this=0xa1279a8) at sql/sql_select.cc:2354 #11 0x082931c8 in JOIN::exec (this=0xa1279a8) at sql/sql_select.cc:2201 BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

Re: [Maria-developers] Please review: [Commits] Rev 3405: Test case for bug lp:1001117, MySQL BUG#12330344 in file:///home/tsk/mprog/src/5.5-lpb944706/
by Sergei Petrunia 24 May '12

24 May '12

Hi Timour, On Thu, May 24, 2012 at 02:12:12PM +0300, Timour Katchaounov wrote: > Sergey, > > Could you please review the following patch. So far it is only in my > tree for bug lp:944706 (because this tree already fixes the problem as > explained below). > Ok to push. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] MDEV-83: pushdown of expensive predicates
by Sergei Petrunia 21 May '12

21 May '12

Hi Timour, Some thoughts about MDEV-83. The jira entry says: > The main difference in these two plans is that with the first execution plan > the dependent subquery: ... is evaluated before the table part is accessed > while many rows where (p_name like 'forest%') is not true are filtered out > and evaluation of the dependent subquery is not needed for them. The way I read this, the only way one could figure out that the subquery should be evaluated at later phase (after table part) is to figure out that selectivity of table `part`, together with its condition of p_name like 'forest%' is less than one. That way, subquery is evaluated fewer times, and query execution will be cheaper. I don't see any index that covers p_name column. This means that current optimizer has no clue about selectivity of conditions over p_name column, and hence is unable to make the correct decision. If I add an index on p_name, then I get a different query plan (I didn't check if is slower or faster than the original). I think, one of the first things this MDEV needs is to figure out how exactly we will get selectivities for non-indexed columns. If we get them, and start to take them into account in join optimizer(*), the example provided in this MDEV/ bug may just go away :-). The general problem will remain, though. (*) { Currently, the optimizer has "quick_condition_rows", which may cause join fanout to be less than 1 (check out examples with EXPLAIN EXTENDED and filtered != 100% in the testsuite). The problem is that quick_condition_rows is not used consistently for all ways to access the table, for example it is not taken into account when considering ref access. I do not know whether Igor's work in best_access_path() includes making it consistently make use of quick_condition_rows estimates. } == Subquery predicate attachment considerations == Some thoughts about the problem. There are three factors we may want/need to take into account when attaching subquery predicate to some join_tab: 1. Fanout of join tabs 2. Cost of the subquery predicate evaluation 3. Selectivity of the subquery predicate. The example used in this MDEV item can be resolved by taking into account #1 and #2. For the first step, we only need to know - accurate join tab fanout (#1), - that subquery predicate evaluation has some cost ( #2 != 0) these two allow us to set/resolve a task of attaching the subquery predicate to the join_tab where total cost of its evaluations is minimized. The exact value of subquery evaluation cost is irrelevant. It will only become relevant when/if we start to make a choice between evaluating subquery predicate, or preforming join with the next join_tab. In order for that choice to be meaningful, we will need to have an idea about subquery predicate's selectivity. That way, we will be able to make a choice between 1. evaluate subquery, then make join with next join tab cost(subquery_pred) + selectivity(subquery_pred) * cost(join_tab) and 2. make join with next join_tab, then evaluate the subquery: cost(join_tab) + selectivity(join_tab) * cost(join_tab) BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-136 Non-blocking "set read_only"
by Sergei Golubchik 20 May '12

20 May '12

Hi, Holyfoot! Looks good to me! See a couple of comments below. Please push after fixing all that. On May 17, holyfoot(a)askmonty.org wrote: > ------------------------------------------------------------ > revno: 3518 > revision-id: holyfoot(a)askmonty.org-20120517131405-y12erfkgta2ym2pn > parent: sanja(a)montyprogram.com-20120507181437-1zl1v1uxje7v6994 > committer: Alexey Botchkov <holyfoot(a)askmonty.org> > branch nick: mdev-136 > timestamp: Thu 2012-05-17 18:14:05 +0500 > message: > MDEV-136 Non-blocking "set read_only". > Handle the 'set read_only=1' in lighter way, than the FLUSH TABLES READ LOCK; > For the transactional engines we don't wait for operations on that tables to finish. > > === modified file 'mysql-test/t/read_only_innodb.test' > --- a/mysql-test/t/read_only_innodb.test 2008-04-08 05:20:58 +0000 > +++ b/mysql-test/t/read_only_innodb.test 2012-05-17 13:14:05 +0000 > @@ -75,7 +75,28 @@ > SELECT * FROM t1; > COMMIT; > > -connection default; > +# > +# Tests that LOCK TABLE doesn't block the SET READ_ONLY=1 for the InnoDB tables > +# > +connection default; > +UNLOCK TABLES; > + > +--echo connection con1; > +connection con1; > +lock table t1 write; See whether you can directly test that the table was not reopened. Perhaps by examining the corresponding SHOW STATUS variables you could do that. > + > +--echo connection default; > +connection default; > +set global read_only=1; > + > +--echo connection con1; > +connection con1; > +unlock tables; > + > +--echo connection default; > +connection default; > +SET GLOBAL read_only=0; > + > UNLOCK TABLES; > DROP TABLE t1; > DROP USER test@localhost; > === modified file 'sql/sql_base.cc' > --- a/sql/sql_base.cc 2012-04-19 06:16:30 +0000 > +++ b/sql/sql_base.cc 2012-05-17 13:14:05 +0000 > @@ -933,7 +938,14 @@ > for (uint idx=0 ; idx < open_cache.records ; idx++) > { > TABLE *table=(TABLE*) hash_element(&open_cache,idx); > - if (table->in_use) > + /* We don't increment the refresh_version when set_readonly_mode, */ > + /* but we still need non-transactional tables to be reopened. */ > + /* So we set their versions as 'refresh_version - 1', which marks */ > + /* them for the 'needs_reopen_or_table_lock()' */ please, use the standard mysql comment style as you can see everywhere in the sources > + if (set_readonly_mode && !table->file->has_transactions()) > + table->s->version= refresh_version-1; better set it to 0, that's what tdc_remove_table() does. And we won't need to think about possible race conditions. > + if (table->in_use && > + (!set_readonly_mode || !table->file->has_transactions())) > table->in_use->some_tables_deleted= 1; > } > } Regards, Sergei

1 0

[Maria-developers] On the DBUG_ASSERT in create_ref_for_key()
by Sergei Petrunia 20 May '12

20 May '12

Hi Sergei, Re the DBUG_ASSERT(length > 0 && keyparts != 0); line in create_ref_for_key() in 5.3: I've looked - the assert can be commented out, because it's valid to have length=0 && !keyparts for a variant of LooseScan algorithm. Some more details: I've tried to come up another assert statement, that would allow length==0 only when LooseScan is used. Unfortunately, it is not easy - the query plan is encoded in join->best_positions, while create_ref_for_key() has access only to the JOIN_TAB. Since SJ-Materialization, the query plan is represented with multiple JOIN_TAB arrays, while there is only one POSITION array, which means that one needs to loop to find the POSITION which corresponds to the JOIN_TAB. EXPLAIN and execution use JOIN_TAB::loosescan_match_tab, but it is set by setup_semijoin_dups_elimination() which is invoked after create_ref_for_key() is. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-225: Replace with dummy events an event that is not understood by a slave to which it should be sent
by Sergei Golubchik 19 May '12

19 May '12

Hi, Kristian! On Apr 27, knielsen(a)knielsen-hq.org wrote: > revno: 3368 > revision-id: knielsen(a)knielsen-hq.org-20120427082038-frqt8ykqy77x3r8e > parent: sergii(a)pisem.net-20120410063020-fm6a5ds0iggihq02 > committer: knielsen(a)knielsen-hq.org > branch nick: work-5.5-mdev225 > timestamp: Fri 2012-04-27 10:20:38 +0200 > message: > MDEV-225: Replace with dummy events an event that is not understood > by a slave to which it should be sent > > Add function to replace arbitrary event with dummy event. > > Add code which uses this to fix the bug that enabling row_annotate > events on the master breaks slaves which do not request such events. > > Add that slaves set a variable @mariadb_slave_capability to inform > the master in a robust way about which events it can, and cannot, > handle. I'm not sure I like the idea of magic user variables. Why wouldn't you create a session server variable @@mariadb_slave_capability? Also: Why would "a later release of a slave change what an earlier release of a master believes about its capabilities"? See comments to the code below. > === added file 'mysql-test/suite/rpl/t/rpl_mariadb_slave_capability.test' > --- a/mysql-test/suite/rpl/t/rpl_mariadb_slave_capability.test 1970-01-01 00:00:00 +0000 > +++ b/mysql-test/suite/rpl/t/rpl_mariadb_slave_capability.test 2012-04-27 08:20:38 +0000 > @@ -0,0 +1,96 @@ > +--source include/master-slave.inc > +--source include/have_debug.inc > +--source include/have_binlog_format_row.inc > + > +connection master; > + > +set @old_master_binlog_checksum= @@global.binlog_checksum; > +CREATE TABLE t1 (a INT PRIMARY KEY); > +INSERT INTO t1 VALUES (0); > + > +sync_slave_with_master; > +connection slave; > + > +--echo # Test slave with no capability gets dummy event, which is ignored. > +--source include/stop_slave.inc > +SET @@global.debug_dbug='d,simulate_slave_capability_none'; Don't do that please, it'll stop any debug trace for ./mtr --debug. Instead, write SET @@global.debug_dbug='+d,simulate_slave_capability_old_53'; ... your tests SET @@global.debug_dbug='-d,simulate_slave_capability_old_53'; or SET @debug_old=@@global.debug_dbug; SET @@global.debug_dbug='+d,simulate_slave_capability_old_53'; ... your tests SET @@global.debug_dbug=@debug_old; > +--source include/start_slave.inc > +connection master; > +sync_slave_with_master; > +connection slave; > +let $relaylog_start= query_get_value(SHOW SLAVE STATUS, Relay_Log_Pos, 1); > + > +connection master; > +SET SESSION binlog_annotate_row_events = ON; > +let $binlog_file= query_get_value(SHOW MASTER STATUS, File, 1); > +let $binlog_start= query_get_value(SHOW MASTER STATUS, Position, 1); > +# A short event, to test when we need to use user_var_event for dummy event. ... > === modified file 'sql/log_event.cc' > --- a/sql/log_event.cc 2012-04-10 06:28:13 +0000 > +++ b/sql/log_event.cc 2012-04-27 08:20:38 +0000 > @@ -3277,6 +3277,114 @@ Query_log_event::Query_log_event(const c > } > > > +/* > + Replace a binlog event read into a packet with a dummy event. Either a > + Query_log_event that has just a comment, or if that will not fit in the > + space used for the event to be replaced, then a NULL user_var event. > + > + This is used when sending binlog data to a slave which does not understand > + this particular event and which is too old to support informational events > + or holes in the event stream. > + > + This allows to write such events into the binlog on the master and still be > + able to replicate against old slaves without them breaking. > + > + Clears the flag LOG_EVENT_THREAD_SPECIFIC_F and set LOG_EVENT_SUPPRESS_USE_F. > + Overwrites the type with QUERY_EVENT (or USER_VAR_EVENT), and replaces the > + body with a minimal query / NULL user var. > + > + Returns zero on success, -1 if error due to too little space in original > + event. A minimum of 25 bytes (19 bytes fixed header + 6 bytes in the body) > + is needed in any event to be replaced with a dummy event. > +*/ > +int > +Query_log_event::dummy_event(String *packet, ulong ev_offset, > + uint8 checksum_alg) > +{ > + uchar *p= (uchar *)packet->ptr() + ev_offset; > + size_t data_len= packet->length() - ev_offset; > + uint16 flags; > + static const size_t min_user_var_event_len= > + LOG_EVENT_HEADER_LEN + UV_NAME_LEN_SIZE + 1 + UV_VAL_IS_NULL; // 25 > + static const size_t min_query_event_len= > + LOG_EVENT_HEADER_LEN + QUERY_HEADER_LEN + 1 + 1; // 34 > + > + if (checksum_alg == BINLOG_CHECKSUM_ALG_CRC32) > + data_len-= BINLOG_CHECKSUM_LEN; > + else > + DBUG_ASSERT(checksum_alg == BINLOG_CHECKSUM_ALG_UNDEF || > + checksum_alg == BINLOG_CHECKSUM_ALG_OFF); > + > + if (data_len < min_user_var_event_len) > + /* Cannot replace with dummy, event too short. */ > + return -1; > + > + flags= uint2korr(p + FLAGS_OFFSET); > + flags&= ~LOG_EVENT_THREAD_SPECIFIC_F; > + flags|= LOG_EVENT_SUPPRESS_USE_F; > + int2store(p + FLAGS_OFFSET, flags); > + > + if (data_len < min_query_event_len) > + { > + /* > + Have to use dummy user_var event for such a short packet. > + > + This works, but the event will be considered part of an event group with > + the following event, so things like START SLAVE UNTIL may behave slightly > + unexpected. Why? > + > + We write a NULL user var with the name @`!dummyvar` (or as much > + as that as will fit within the size of the original event - so > + possibly just @`!`). > + */ > + static const char var_name[]= "!dummyvar"; > + uint name_len= data_len - (min_user_var_event_len - 1); > + > + p[EVENT_TYPE_OFFSET]= USER_VAR_EVENT; > + int4store(p + LOG_EVENT_HEADER_LEN, name_len); > + memcpy(p + LOG_EVENT_HEADER_LEN + UV_NAME_LEN_SIZE, var_name, name_len); > + p[LOG_EVENT_HEADER_LEN + UV_NAME_LEN_SIZE + name_len]= 1; // indicates NULL > + } > + else > + { > + /* > + Use a dummy query event, just a comment. > + */ > + static const char message[]= > + "# Dummy event replacing event type %u that slave cannot handle."; > + char buf[sizeof(message)+1]; /* +1, as %u can expand to 3 digits. */ > + uchar old_type= p[EVENT_TYPE_OFFSET]; > + uchar *q= p + LOG_EVENT_HEADER_LEN; > + size_t comment_len, len; > + > + p[EVENT_TYPE_OFFSET]= QUERY_EVENT; > + int4store(q + Q_THREAD_ID_OFFSET, 0); > + int4store(q + Q_EXEC_TIME_OFFSET, 0); > + q[Q_DB_LEN_OFFSET]= 0; > + int2store(q + Q_ERR_CODE_OFFSET, 0); > + int2store(q + Q_STATUS_VARS_LEN_OFFSET, 0); > + q[Q_DATA_OFFSET]= 0; /* Zero terminator for empty db */ > + q+= Q_DATA_OFFSET + 1; > + len= my_snprintf(buf, sizeof(buf), message, old_type); > + comment_len= data_len - (min_query_event_len - 1); > + if (comment_len <= len) > + memcpy(q, buf, comment_len); > + else > + { > + memcpy(q, buf, len); > + memset(q+len, ' ', comment_len - len); > + } > + } > + > + if (checksum_alg == BINLOG_CHECKSUM_ALG_CRC32) > + { > + ha_checksum crc= my_checksum(0L, p, data_len); > + int4store(p + data_len, crc); > + } > + return 0; > +} > + > + > #ifdef MYSQL_CLIENT > /** > Query_log_event::print(). > === modified file 'sql/log_event.h' > --- a/sql/log_event.h 2012-04-10 06:28:13 +0000 > +++ b/sql/log_event.h 2012-04-27 08:20:38 +0000 > @@ -566,6 +566,43 @@ enum enum_binlog_checksum_alg { > #define BINLOG_CHECKSUM_LEN CHECKSUM_CRC32_SIGNATURE_LEN > #define BINLOG_CHECKSUM_ALG_DESC_LEN 1 /* 1 byte checksum alg descriptor */ > > +/* > + These are cabability numbers for MariaDB slave servers. Typo. capability. > + > + Newer MariaDB slaves set this to inform the master about their capabilities. > + This allows the master to decide which events it can send to the slave > + without breaking replication on old slaves that maybe do not understand > + all events from newer masters. > + > + As new releases are backwards compatible, a given capability implies also > + all capabilities with smaller number. > + > + Older MariaDB slaves and other MySQL slave servers do not set this, so they > + are recorded with capability 0. > +*/ > + > +/* MySQL or old MariaDB slave with no announced capability. */ > +#define MARIA_SLAVE_CAPABILITY_UNKNOWN 0 > +/* MariaDB >= 5.3, which understands ANNOTATE_ROWS_EVENT. */ > +#define MARIA_SLAVE_CAPABILITY_ANNOTATE 1 > +/* > + MariaDB >= 5.5. This version has the capability to tolerate events omitted > + from the binlog stream without breaking replication (MySQL slaves fail > + because they mis-compute the offsets into the master's binlog). > +*/ > +#define MARIA_SLAVE_CAPABILITY_TOLERATE_HOLES 2 > +/* MariaDB > 5.5, which knows about binlog_checkpoint_log_event. */ > +#define MARIA_SLAVE_CAPABILITY_BINLOG_CHECKPOINT 3 > +/* > + MariaDB server which understands MySQL 5.6 ignorable events. This server > + can tolerate receiving any event with the LOG_EVENT_IGNORABLE_F flag set. > +*/ > +#define MARIA_SLAVE_CAPABILITY_IGNORABLE 4 > + > +/* Our capability. */ > +#define MARIA_SLAVE_CAPABILITY_MINE MARIA_SLAVE_CAPABILITY_BINLOG_CHECKPOINT > + > + > /** > @enum Log_event_type > > === modified file 'sql/slave.cc' > --- a/sql/slave.cc 2012-04-10 06:28:13 +0000 > +++ b/sql/slave.cc 2012-04-27 08:20:38 +0000 > @@ -1747,6 +1747,37 @@ past_checksum: > } > } > } > + > + /* Announce MariaDB slave capabilities. */ > + DBUG_EXECUTE_IF("simulate_slave_capability_none", goto after_set_cabability;); > + { > + int rc= DBUG_EVALUATE_IF("simulate_slave_capability_old_53", > + mysql_real_query(mysql, STRING_WITH_LEN("SET @mariadb_slave_capability=" > + STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_ANNOTATE))), > + mysql_real_query(mysql, STRING_WITH_LEN("SET @mariadb_slave_capability=" > + STRINGIFY_ARG(MARIA_SLAVE_CAPABILITY_MINE)))); Better put DBUG_EVALUATE_IF only around the STRINGIFY_ARG values: DBUG_EVALUATE_IF("...", STRINGIFY_ARG(...), STRINGIFY_ARG(...)) so that next time I (or somebody) wouldn't need to vdiff two long function calls, trying to count parentheses and commans :) > + if (rc) > + { > + err_code= mysql_errno(mysql); > + if (is_network_error(err_code)) > + { > + mi->report(ERROR_LEVEL, err_code, > + "Setting @mariadb_slave_capability failed with error: %s", > + mysql_error(mysql)); > + goto network_err; > + } > + else > + { > + /* Fatal error */ > + errmsg= "The slave I/O thread stops because a fatal error is " > + "encountered when it tries to set @mariadb_slave_capability."; > + sprintf(err_buff, "%s Error: %s", errmsg, mysql_error(mysql)); > + goto err; > + } > + } > + } > +after_set_cabability: > + > err: > if (errmsg) > { > === modified file 'sql/sql_repl.cc' > --- a/sql/sql_repl.cc 2012-03-28 17:26:00 +0000 > +++ b/sql/sql_repl.cc 2012-04-27 08:20:38 +0000 > @@ -563,14 +584,44 @@ static int send_heartbeat_event(NET* net > static const char * > send_event_to_slave(THD *thd, NET *net, String* const packet, ushort flags, > Log_event_type event_type, char *log_file_name, > - IO_CACHE *log) > + IO_CACHE *log, int mariadb_slave_capability, > + ulong ev_offset, uint8 current_checksum_alg) current_checksum_alg is part of THD, why do you bother to pass is separately? > { > my_off_t pos; > > /* Do not send annotate_rows events unless slave requested it. */ > - if (event_type == ANNOTATE_ROWS_EVENT && > - !(flags & BINLOG_SEND_ANNOTATE_ROWS_EVENT)) > - return NULL; > + if (event_type == ANNOTATE_ROWS_EVENT && !(flags & BINLOG_SEND_ANNOTATE_ROWS_EVENT)) > + { > + if (mariadb_slave_capability >= MARIA_SLAVE_CAPABILITY_TOLERATE_HOLES) > + { > + /* This slave can tolerate events omitted from the binlog stream. */ > + return NULL; > + } > + else if (mariadb_slave_capability >= MARIA_SLAVE_CAPABILITY_ANNOTATE) > + { > + /* > + The slave did not request ANNOTATE_ROWS_EVENT (it does not need them as > + it will not log them in its own binary log). However, it understands the > + event and will just ignore it, and it would break if we omitted it, > + leaving a hole in the binlog stream. So just send the event as-is. > + */ > + } > + else > + { > + /* > + The slave does not understand ANNOTATE_ROWS_EVENT. > + > + Older MariaDB slaves (and MySQL slaves) will break replication if there > + are holes in the binlog stream (they will miscompute the binlog offset > + and request the wrong position when reconnecting). > + > + So replace the event with a dummy event of the same size that will be > + a no-operation on the slave. > + */ > + if (Query_log_event::dummy_event(packet, ev_offset, current_checksum_alg)) > + return "Failed to replace row annotate event with dummy: too small event."; No ER_xxx error message for that? > + } > + } > > /* > Skip events with the @@skip_replication flag set, if slave requested Regards, Sergei

3 5

[Maria-developers] BUG#1001117: a problem with constants.
by Sergei Petrunia 18 May '12

18 May '12

Hi Timour, In regular maria-5.5, I get: (gdb) wher 1 #0 Item_func_like::select_optimize (this=0xa11c478) at /home/psergey/dev2/5.5-look7/sql/item_cmpfunc.cc:4739 (More stack frames follow...) (gdb) p args[1] $123 = (Item_singlerow_subselect *) 0xa1194a0 (gdb) p args[1]->used_tables() $124 = 0 (gdb) p args[1]->const_item() $125 = true In 5.5-timour tree, I get: #1 0x0841a035 in Item_func_like::select_optimize (this=0xa109820) at /home/psergey/dev2/5.5-timour/sql/item_cmpfunc.cc:4739 (gdb) p args[1] $47 = (Item_singlerow_subselect *) 0xa106848 (gdb) p args[1]->used_tables() $49 = 0 (gdb) p args[1]->const_item() $51 = false I also put a breakpoint in JOIN::get_examined_rows(), it is never hit. Can you explain this effect? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] MDEV-136 Non-blocking "set read_only"
by Sergei Golubchik 17 May '12

17 May '12

Hi, Holyfoot! On May 09, holyfoot(a)askmonty.org wrote: > ------------------------------------------------------------ > revno: 3518 > revision-id: holyfoot(a)askmonty.org-20120509185945-i1yx6d38phgoqzes > parent: sanja(a)montyprogram.com-20120507181437-1zl1v1uxje7v6994 > committer: Alexey Botchkov <holyfoot(a)askmonty.org> > branch nick: mdev-136 > timestamp: Wed 2012-05-09 23:59:45 +0500 > message: > MDEV-136 Non-blocking "set read_only". > Handle the 'set read_only=1' in lighter way, than the FLUSH TABLES READ LOCK; > For the transactional engines we don't wait for operations on that tables to finish. See my comments below. > === modified file 'mysql-test/r/read_only.result' > --- a/mysql-test/r/read_only.result 2009-03-06 14:56:17 +0000 > +++ b/mysql-test/r/read_only.result 2012-05-09 18:59:45 +0000 > @@ -59,7 +59,7 @@ > connection con1; > select @@global.read_only; > @@global.read_only > -0 > +1 This is prone to race conditions. Please, fix the test to remove "send" here and below. (assuming the new result is correct) But is it correct ? Why set read_only is not blocked by a write locked myisam table? > unlock tables ; > select @@global.read_only; > @@global.read_only > @@ -80,7 +80,7 @@ > connection con1; > select @@global.read_only; > @@global.read_only > -0 > +1 > unlock tables ; > select @@global.read_only; > @@global.read_only Where are the tests for the changed functionality? > === modified file 'sql/sql_base.cc' > --- a/sql/sql_base.cc 2012-04-19 06:16:30 +0000 > +++ b/sql/sql_base.cc 2012-05-09 18:59:45 +0000 > @@ -933,7 +938,8 @@ > for (uint idx=0 ; idx < open_cache.records ; idx++) > { > TABLE *table=(TABLE*) hash_element(&open_cache,idx); > - if (table->in_use) > + if (table->in_use && > + (!set_readonly_mode || !table->file->has_transactions())) I wonder how this could work. The line below sets a flag *on a thread*. The task description tells "not wait for transactional tables", while your change means "not set a flag if all tables used in a thread are transactional". That is, if a thread uses both transactional and non-transactional tables, your change does nothing. Back to my first question - where are the tests for this new task? > table->in_use->some_tables_deleted= 1; > } > } Regards, Sergei

3 6

Re: [Maria-developers] Review request for: [Commits] Rev 3403: Fix for bug lp:944706, task MDEV-193 in file:///home/tsk/mprog/src/5.5-lpb944706/
by Sergei Petrunia 17 May '12

17 May '12

On Mon, May 14, 2012 at 11:22:52PM +0300, Timour Katchaounov wrote: > > > /** > > > + Estimate the number of rows that query execution will read. > > > + > > > + @todo This is a very pessimistic upper bound. Use join selectivity > > > + when available to produce a more realistic number. > > > +*/ > > > + > > > +double JOIN::get_examined_rows() > > > +{ > > > + /* Each constant table examines one row, and the result is at most one row. */ > > > + ha_rows examined_rows= const_tables; > > > + uint i= const_tables; > > > + double prev_fanout; > > > + > > > + if (table_count == const_tables) > > > + return examined_rows; > > > + > > > + examined_rows+= join_tab[i++].get_examined_rows(); > > > + for (; i < table_count ; i++) > > > + { > > > + if (join_tab[i].type == JT_EQ_REF) > > > + prev_fanout= 1; > > > + else > > > + prev_fanout= best_positions[i-1].records_read; > > This looks wrong. Declaration of POSITION::records_read has this comment: > > /* > The "fanout": number of output rows that will be produced (after > pushed down selection condition is applied) per each row combination of > previous tables. > */ > > note the "PER EACH ROW COMBINATION .." part. I would expect that this function > would calculate a product of records_read values. > > timour: > > In this function we want to estimate how many rows will be *examined*, not > produced by each JOIN operator. For the estimate I assume a simple nested loops > model, where the JOIN read every row of its right table as many times as many > rows there are in the left operand. The partial join that serves as left > operand, contains records_read rows. This is the multiplier of the number of > rows that will be examined in the right table. > > Of course, for each partial join the join condition will filter a subset of > these rows. > > I agree that blocking algorithms may examine a lot less rows, but it doesn't > make sense to have a very tight bound here. We use an upper bound, because it > is better to miss some constant optimizations, rather than execute very > expensive subqueries here. I was talking about something much more simpler than that. Consider this example: create table ten (a int); insert into ten values (0),(1),(2),(3),(4),(5),(6),(7),(8),(9); create table one_k (a int); insert into one_k select A.a + 10*B.a + 100*C.a from ten A, ten B, ten C; create table ti1 ( a int); insert into ti1 select a from ten limit 4; alter table ti1 add b int; create table ti2 (a int primary key, b int); create table ti3 (a int primary key, b int); insert into ti2 select a, a from one_k; insert into ti3 select a, a from one_k; MariaDB [test]> explain select * from ten where 3 in (select ti2.b + ti3.b from ti1, ti2, ti3 where ti2.a=ti1.a and ti3.a=ti1.a) or a <5; +------+--------------------+-------+--------+---------------+---------+---------+------------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +------+--------------------+-------+--------+---------------+---------+---------+------------+------+-------------+ | 1 | PRIMARY | ten | ALL | NULL | NULL | NULL | NULL | 10 | Using where | | 2 | DEPENDENT SUBQUERY | ti1 | ALL | NULL | NULL | NULL | NULL | 4 | Using where | | 2 | DEPENDENT SUBQUERY | ti3 | eq_ref | PRIMARY | PRIMARY | 4 | test.ti1.a | 1 | | | 2 | DEPENDENT SUBQUERY | ti2 | eq_ref | PRIMARY | PRIMARY | 4 | test.ti1.a | 1 | Using where | +------+--------------------+-------+--------+---------------+---------+---------+------------+------+-------------+ For this example, JOIN::get_examined_rows() produces 6. The number comes from 4 rows examined in table ti1 (correct) 1 row examined in table ti2 (incorrect, should be 4) 1 row examined in table ti3 (incorrect, should be 4) Do you agree that the number of 6 is incorrect and needs to be fixed? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] KB Question: Check constraints support
by Daniel Bartholomew 15 May '12

15 May '12

The following question has been posted to the Knowledgebase: http://kb.askmonty.org/en/check-constraints-support - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Check constraints support Hi, I have been looking in JIRA, opened bugs, etc. But haven't been able to find an answer. Are check constraints going to be supported? Because they give me the opportunity to give more consistency to my Database, and avoid problems of corruption. The mysql bug is this: http://bugs.mysql.com/bug.php?id=3464 Has been 8 years abandoned, and I expect that you don't do the same, Thank you in advance, txomon - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Thanks. -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

Re: [Maria-developers] Review request for: [Commits] Rev 3403: Fix for bug lp:944706, task MDEV-193 in file:///home/tsk/mprog/src/5.5-lpb944706/
by Sergei Petrunia 14 May '12

14 May '12

On Sun, May 13, 2012 at 02:11:15PM +0300, Timour Katchaounov wrote: > Sergey, > > At the optimizer call this Tuesday Igor suggested that you > should review the patch for this bug. The commit message > contains a description of the patch. Let me know if you believe > a more detailed description is needed. > > The patch is also pushed into the following tree: > lp:~maria-captains/maria/5.5-timour > > > Thanks, > Timour > > ------------------------------------------------------------ > revno: 3403 > revision-id: timour(a)askmonty.org-20120511152503-zbjewctjclx0sajt > parent: knielsen(a)knielsen-hq.org-20120508122744-v5okh33kuolot3kb > fixes bug(s): https://launchpad.net/bugs/944706 > committer: timour(a)askmonty.org > branch nick: 5.5-lpb944706 > timestamp: Fri 2012-05-11 18:25:03 +0300 > message: > Fix for bug lp:944706, task MDEV-193 > > The patch enables back constant subquery execution during > query optimization after it was disabled during the development > of MWL#89 (cost-based choice of IN-TO-EXISTS vs MATERIALIZATION). > > The main idea is that constant subqueries are allowed to be executed > during optimization if their execution is not expensive. > > The approach is as follows: > - Constant subqueries are recursively optimized in the beginning of > JOIN::optimize of the outer query. This is done by the new method > JOIN::optimize_constant_subqueries(). This is done so that the cost > of executing these queries can be estimated. > - Optimization of the outer query proceeds normally. During this phase > the optimizer may request execution of non-expensive constant subqueries. > Each place where the optimizer may potentially execute an expensive > expression is guarded with the predicate Item::is_expensive(). > - The implementation of Item_subselect::is_expensive has been extended > to use the number of examined rows (estimated by the optimizer) as a > way to determine whether the subquery is expensive or not. > - The new system variable "expensive_subquery_limit" controls how many > examined rows are considered to be not expensive. The default is 100. > > In addition, multiple changes were needed to make this solution work > in the light of the changes made by MWL#89. These changes were needed > to fix various crashes and wrong results, and legacy bugs discovered > during development. > === modified file 'mysql-test/r/explain.result' > --- a/mysql-test/r/explain.result 2012-01-13 14:50:02 +0000 > +++ b/mysql-test/r/explain.result 2012-05-11 15:25:03 +0000 > @@ -260,7 +260,7 @@ FLUSH TABLES; > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT 1 FROM t1 t JOIN t2 WHERE b <= 1 AND t.a); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > DROP TABLE t1, t2; > # > # Bug #48573: difference of index selection between rpm binary and > @@ -287,7 +287,7 @@ WHERE t1.f1 GROUP BY t1.f1)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 2 SUBQUERY a system NULL NULL NULL NULL 1 > -2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 Using where > +2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 > PREPARE stmt FROM > 'EXPLAIN SELECT 1 FROM t1 > WHERE 1 > ALL((SELECT t1.f1 FROM t1 RIGHT OUTER JOIN t1 a > @@ -297,12 +297,12 @@ EXECUTE stmt; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 2 SUBQUERY a system NULL NULL NULL NULL 1 > -2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 Using where > +2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 > EXECUTE stmt; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 2 SUBQUERY a system NULL NULL NULL NULL 1 > -2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 Using where > +2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 > DEALLOCATE PREPARE stmt; > PREPARE stmt FROM > 'EXPLAIN SELECT 1 FROM t1 > @@ -313,12 +313,12 @@ EXECUTE stmt; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 2 SUBQUERY a system NULL NULL NULL NULL 1 > -2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 Using where > +2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 > EXECUTE stmt; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 2 SUBQUERY a system NULL NULL NULL NULL 1 > -2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 Using where > +2 SUBQUERY t1 fulltext f1_2,f1 f1 0 1 > DEALLOCATE PREPARE stmt; > DROP TABLE t1; > End of 5.1 tests. > > === modified file 'mysql-test/r/group_min_max.result' > --- a/mysql-test/r/group_min_max.result 2012-03-01 22:22:22 +0000 > +++ b/mysql-test/r/group_min_max.result 2012-05-11 15:25:03 +0000 > @@ -2398,12 +2398,12 @@ EXPLAIN SELECT 1 FROM t1 AS t1_outer WHE > (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1_outer index NULL a 10 NULL 15 Using index > -2 SUBQUERY t1 index NULL a 10 NULL 1 Using index > +2 SUBQUERY t1 index NULL a 10 NULL 15 Using index > EXPLAIN SELECT 1 FROM t1 AS t1_outer WHERE > (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2) > 12; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1_outer index NULL a 10 NULL 15 Using index > -2 SUBQUERY t1 range NULL a 5 NULL 8 Using index for group-by > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY t1 index NULL a 10 NULL 15 Using index > EXPLAIN SELECT 1 FROM t1 AS t1_outer WHERE > a IN (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2); > id select_type table type possible_keys key key_len ref rows Extra > @@ -2419,9 +2419,9 @@ EXPLAIN SELECT 1 FROM t1 AS t1_outer1 JO > ON t1_outer1.a = (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2) > AND t1_outer1.b = t1_outer2.b; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1_outer1 ref a a 5 const 2 Using where; Using index > +1 PRIMARY t1_outer1 ref a a 5 const 1 Using where; Using index > 1 PRIMARY t1_outer2 index NULL a 10 NULL 15 Using where; Using index; Using join buffer (flat, BNL join) > -2 SUBQUERY t1 range NULL a 5 NULL 8 Using index for group-by > +2 SUBQUERY t1 index NULL a 10 NULL 15 Using index > EXPLAIN SELECT (SELECT (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2) x > FROM t1 AS t1_outer) x2 FROM t1 AS t1_outer2; > id select_type table type possible_keys key key_len ref rows Extra > @@ -2756,8 +2756,8 @@ NULL > EXPLAIN > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > id select_type table type possible_keys key key_len ref rows Extra > -x x x x x x x x x Using where; Using index > -x x x x x x x x x Using where; Using index > +x x x x x x x x x Impossible WHERE noticed after reading const tables > +x x x x x x x x x Using index > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > MIN( a ) > NULL > @@ -2828,8 +2828,8 @@ NULL > EXPLAIN > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > id select_type table type possible_keys key key_len ref rows Extra > -x x x x x x x x x Using where; Using index > -x x x x x x x x x Using where; Using index > +x x x x x x x x x Impossible WHERE noticed after reading const tables > +x x x x x x x x x Using index > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > MIN( a ) > NULL > @@ -2907,8 +2907,8 @@ NULL > EXPLAIN > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > id select_type table type possible_keys key key_len ref rows Extra > -x x x x x x x x x Using where; Using index > -x x x x x x x x x Using where; Using index > +x x x x x x x x x Impossible WHERE noticed after reading const tables > +x x x x x x x x x Using index > SELECT MIN( a ) FROM t1 WHERE a = (SELECT a FROM t1 WHERE a < 0); > MIN( a ) > NULL > > === modified file 'mysql-test/r/join_outer.result' > --- a/mysql-test/r/join_outer.result 2012-04-07 13:58:46 +0000 > +++ b/mysql-test/r/join_outer.result 2012-05-11 15:25:03 +0000 > @@ -1871,10 +1871,10 @@ EXPLAIN EXTENDED > SELECT t2.a FROM t1 LEFT JOIN t2 ON (6) IN (SELECT a FROM t3); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -1 PRIMARY t2 ALL NULL NULL NULL NULL 1 100.00 Using where > +1 PRIMARY t2 system NULL NULL NULL NULL 1 100.00 > 2 DEPENDENT SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > -Note 1003 select `test`.`t2`.`a` AS `a` from `test`.`t2` where 1 > +Note 1003 select NULL AS `a` from `test`.`t2` where 1 > DROP TABLE t1,t2,t3; > # > # LP bug #817384 Wrong result with outer join + subquery in ON > > === modified file 'mysql-test/r/join_outer_jcl6.result' > --- a/mysql-test/r/join_outer_jcl6.result 2012-04-07 13:58:46 +0000 > +++ b/mysql-test/r/join_outer_jcl6.result 2012-05-11 15:25:03 +0000 > @@ -1882,10 +1882,10 @@ EXPLAIN EXTENDED > SELECT t2.a FROM t1 LEFT JOIN t2 ON (6) IN (SELECT a FROM t3); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > -1 PRIMARY t2 ALL NULL NULL NULL NULL 1 100.00 Using where > +1 PRIMARY t2 system NULL NULL NULL NULL 1 100.00 > 2 DEPENDENT SUBQUERY t3 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > -Note 1003 select `test`.`t2`.`a` AS `a` from `test`.`t2` where 1 > +Note 1003 select NULL AS `a` from `test`.`t2` where 1 > DROP TABLE t1,t2,t3; > # > # LP bug #817384 Wrong result with outer join + subquery in ON > > === modified file 'mysql-test/r/key.result' > --- a/mysql-test/r/key.result 2011-10-19 19:45:18 +0000 > +++ b/mysql-test/r/key.result 2012-05-11 15:25:03 +0000 > @@ -598,8 +598,8 @@ VALUES > EXPLAIN SELECT 1 FROM t1 AS t1_outer WHERE > (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2) > 12; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1_outer index NULL a 10 NULL 15 Using index > -2 SUBQUERY t1 range NULL a 5 NULL 8 Using index for group-by > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY t1 index NULL a 10 NULL 15 Using index > SELECT 1 as RES FROM t1 AS t1_outer WHERE > (SELECT max(b) FROM t1 GROUP BY a HAVING a < 2) > 12; > RES > > === modified file 'mysql-test/r/limit_rows_examined.result' > --- a/mysql-test/r/limit_rows_examined.result 2012-03-11 22:45:18 +0000 > +++ b/mysql-test/r/limit_rows_examined.result 2012-05-11 15:25:03 +0000 > @@ -679,7 +679,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 system NULL NULL NULL NULL 1 Using temporary > 1 PRIMARY t2 ALL NULL NULL NULL NULL 3 Distinct > 2 SUBQUERY t3 ALL NULL NULL NULL NULL 3 > -2 SUBQUERY t2 ALL NULL NULL NULL NULL 3 Using where; Using join buffer (flat, BNL join) > +2 SUBQUERY t2 ALL NULL NULL NULL NULL 3 > SELECT DISTINCT a AS field1 FROM t1, t2 > WHERE EXISTS (SELECT c FROM t3 LEFT JOIN t2 ON b = d) > HAVING field1 > 'aaa' LIMIT ROWS EXAMINED 20; > > === modified file 'mysql-test/r/myisam_mrr.result' > --- a/mysql-test/r/myisam_mrr.result 2012-02-25 15:13:24 +0000 > +++ b/mysql-test/r/myisam_mrr.result 2012-05-11 15:25:03 +0000 > @@ -349,10 +349,10 @@ WHERE t2.int_key IS NULL > GROUP BY t2.pk > ); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 100.00 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 SUBQUERY t2 ref int_key int_key 5 const 1 100.00 Using index condition > Warnings: > -Note 1003 select min(1) AS `MIN(t1.pk)` from dual where exists(select `test`.`t2`.`pk` from `test`.`t2` where isnull(`test`.`t2`.`int_key`)) > +Note 1003 select min(`test`.`t1`.`pk`) AS `MIN(t1.pk)` from `test`.`t1` where 0 > DROP TABLE t1, t2; > # > # BUG#42048 Discrepancy between MyISAM and Maria's ICP implementation > > === modified file 'mysql-test/r/mysqld--help.result' > --- a/mysql-test/r/mysqld--help.result 2012-04-10 06:28:13 +0000 > +++ b/mysql-test/r/mysqld--help.result 2012-05-11 15:25:03 +0000 > @@ -159,6 +159,9 @@ > Enable the event scheduler. Possible values are ON, OFF, > and DISABLED (keep the event scheduler completely > deactivated, it cannot be activated run-time) > + --expensive-subquery-limit=# > + The maximum number of rows a subquery examines in order > + to be considered non-expensive > --expire-logs-days=# > If non-zero, binary logs will be purged after > expire_logs_days days; possible purges happen at startup > @@ -885,6 +888,7 @@ delayed-queue-size 1000 > div-precision-increment 4 > engine-condition-pushdown FALSE > event-scheduler OFF > +expensive-subquery-limit 100 > expire-logs-days 0 > external-locking FALSE > extra-max-connections 1 > > === modified file 'mysql-test/r/subselect.result' > --- a/mysql-test/r/subselect.result 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/r/subselect.result 2012-05-11 15:25:03 +0000 > @@ -371,12 +371,12 @@ INSERT INTO t8 (pseudo,email) VALUES ('j > INSERT INTO t8 (pseudo,email) VALUES ('2joce1','2test1'); > EXPLAIN EXTENDED SELECT pseudo,(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce')) FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where; Using index > +1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > 4 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > -2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where > +2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 > 3 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > Warnings: > -Note 1003 select `test`.`t8`.`pseudo` AS `pseudo`,(select `test`.`t8`.`email` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1))) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1)) > +Note 1003 select 'joce' AS `pseudo`,(select 'test' from `test`.`t8` where 1) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where 1 > SELECT pseudo FROM t8 WHERE pseudo=(SELECT pseudo,email FROM > t8 WHERE pseudo='joce'); > ERROR 21000: Operand should contain 1 column(s) > @@ -517,6 +517,7 @@ ERROR 21000: Subquery returns more than > show warnings; > Level Code Message > Error 1242 Subquery returns more than 1 row > +Error 1028 Sort aborted: Subquery returns more than 1 row > drop table t1; > create table t1 (a int); > insert into t1 values (1),(2),(3); > @@ -547,11 +548,7 @@ SELECT numreponse, (SELECT numeropost FR > numreponse (SELECT numeropost FROM t1 HAVING numreponse=1) > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 const PRIMARY,numreponse PRIMARY 7 const,const 1 100.00 Using where; Using index > -2 SUBQUERY t1 ref PRIMARY PRIMARY 3 const 2 100.00 Using index > -Warnings: > -Note 1003 select `test`.`t1`.`numreponse` AS `numreponse` from `test`.`t1` where ((`test`.`t1`.`numeropost` = '1') and (`test`.`t1`.`numreponse` = (select 1 from `test`.`t1` where (`test`.`t1`.`numeropost` = '1')))) > +ERROR 21000: Subquery returns more than 1 row > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > ERROR 21000: Subquery returns more than 1 row > EXPLAIN EXTENDED SELECT MAX(numreponse) FROM t1 WHERE numeropost='1'; > @@ -1675,34 +1672,34 @@ select * from t3 where NULL >= any (sele > a > explain extended select * from t3 where NULL >= any (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= any (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= any (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2); > a > explain extended select * from t3 where NULL >= some (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= some (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > insert into t2 values (2,2), (2,1), (3,3), (3,1); > select * from t3 where a > all (select max(b) from t2 group by a); > a > @@ -1764,7 +1761,7 @@ id select_type table type possible_keys > 3 UNION t1 system NULL NULL NULL NULL 1 100.00 > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL NULL > Warnings: > -Note 1003 select 'e' AS `s1` from dual where <nop>(<in_optimizer>('f',(<min>(select 'e' from dual union select 'e' from dual) < 'f'))) > +Note 1003 select 'e' AS `s1` from dual where 1 > drop table t1; > CREATE TABLE t1 (number char(11) NOT NULL default '') ENGINE=MyISAM CHARSET=latin1; > INSERT INTO t1 VALUES ('69294728265'),('18621828126'),('89356874041'),('95895001874'); > @@ -3099,7 +3096,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3111,7 +3108,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3727,8 +3724,8 @@ explain select * from t1 where not exist > ((select t11.i from t1 t11) union (select t12.i from t1 t12)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t11 system NULL NULL NULL NULL 0 const row not found > -3 UNION t12 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 UNION NULL NULL NULL NULL NULL NULL NULL no matching row in const table > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > DROP TABLE t1; > CREATE TABLE t1 (a VARCHAR(250), b INT auto_increment, PRIMARY KEY (b)); > @@ -4198,8 +4195,8 @@ CREATE TABLE t1 (a int, b int, KEY (a)); > INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref a a 5 const 0 Using where; Using index > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > +2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > @@ -4626,7 +4623,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <nop>(<in_optimizer>(1,((select max(`test`.`t1`.`a1`) from `test`.`t1`) > 1))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > SELECT 1 FROM t1 WHERE 1 < SOME (SELECT a1 FROM t1); > 1 > 1 > @@ -5960,7 +5957,7 @@ id select_type table type possible_keys > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > DROP TABLE t1; > # > # BUG#12616253 - WRONG RESULT WITH EXISTS(SUBQUERY) (MISSING ROWS) > @@ -6121,7 +6118,7 @@ set optimizer_switch=@tmp_optimizer_swit > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > > DROP TABLE t1; > # > > === modified file 'mysql-test/r/subselect3.result' > --- a/mysql-test/r/subselect3.result 2012-04-07 13:58:46 +0000 > +++ b/mysql-test/r/subselect3.result 2012-05-11 15:25:03 +0000 > @@ -1479,7 +1479,7 @@ id select_type table type possible_keys > 3 MATERIALIZED t3 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > -Note 1003 select `test`.`t1`.`a` AS `a` from `test`.`t1` where (not(<expr_cache><(select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9))>(<in_optimizer>((select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9)),(select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9)) in ( <materialize> (select `test`.`t3`.`b` from `test`.`t3` ), <primary_index_lookup>(9 in <temporary table> on distinct_key where ((9 = `<subquery3>`.`b`)))))))) > +Note 1003 select `test`.`t1`.`a` AS `a` from `test`.`t1` where 1 > SELECT * FROM t1 > WHERE ( > ( SELECT a FROM t2 WHERE a = 9 ), > > === modified file 'mysql-test/r/subselect3_jcl6.result' > --- a/mysql-test/r/subselect3_jcl6.result 2012-04-07 13:58:46 +0000 > +++ b/mysql-test/r/subselect3_jcl6.result 2012-05-11 15:25:03 +0000 > @@ -1489,7 +1489,7 @@ id select_type table type possible_keys > 3 MATERIALIZED t3 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > -Note 1003 select `test`.`t1`.`a` AS `a` from `test`.`t1` where (not(<expr_cache><(select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9))>(<in_optimizer>((select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9)),(select `test`.`t2`.`a` from `test`.`t2` where (`test`.`t2`.`a` = 9)) in ( <materialize> (select `test`.`t3`.`b` from `test`.`t3` ), <primary_index_lookup>(9 in <temporary table> on distinct_key where ((9 = `<subquery3>`.`b`)))))))) > +Note 1003 select `test`.`t1`.`a` AS `a` from `test`.`t1` where 1 > SELECT * FROM t1 > WHERE ( > ( SELECT a FROM t2 WHERE a = 9 ), > > === modified file 'mysql-test/r/subselect4.result' > --- a/mysql-test/r/subselect4.result 2012-03-01 22:22:22 +0000 > +++ b/mysql-test/r/subselect4.result 2012-05-11 15:25:03 +0000 > @@ -562,7 +562,7 @@ WHERE f3 = ( > SELECT t1.f3 FROM t1 > WHERE ( t1.f10 ) IN ( SELECT f11 FROM t2 GROUP BY f11 )); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref f3 f3 5 const 0 Using where > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > 2 SUBQUERY <subquery3> eq_ref distinct_key distinct_key 4 func 1 > 3 MATERIALIZED t2 ALL NULL NULL NULL NULL 2 > @@ -577,7 +577,7 @@ WHERE f3 = ( > SELECT f3 FROM t1 > WHERE ( f10, f10 ) IN ( SELECT f11, f11 FROM t2 GROUP BY f11 )); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref f3 f3 5 const 0 Using where > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 > 2 SUBQUERY <subquery3> eq_ref distinct_key distinct_key 8 func,func 1 > 3 MATERIALIZED t2 ALL NULL NULL NULL NULL 2 > @@ -1220,6 +1220,13 @@ id select_type table type possible_keys > 2 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using where > 3 SUBQUERY SQ1_t1 index NULL f4 5 NULL 2 Using index; Using temporary > 3 SUBQUERY SQ1_t3 index f4 f4 5 NULL 2 Using where; Using index; Using join buffer (flat, BNL join) > +SELECT * FROM t1 WHERE > +(SELECT f2 FROM t2 > +WHERE f4 <= ALL > +(SELECT max(SQ1_t1.f4) > +FROM t3 AS SQ1_t1 JOIN t3 AS SQ1_t3 ON SQ1_t3.f4 > +GROUP BY SQ1_t1.f4)); > +ERROR 21000: Subquery returns more than 1 row > drop table t1, t2, t3; > # > # BUG#52317: Assertion failing in Field_varstring::store() > @@ -1250,8 +1257,7 @@ FROM t2 JOIN t1 ON t1.f3 > WHERE ('v') IN (SELECT f4 FROM t2) > GROUP BY f9; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t2 system NULL NULL NULL NULL 1 > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using where > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > @@ -1267,8 +1273,7 @@ FROM t2 JOIN t1 ON t1.f3 > WHERE ('v') IN (SELECT f4 FROM t2) > ORDER BY f9; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t2 system NULL NULL NULL NULL 1 > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 Using where > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > @@ -1277,7 +1282,7 @@ FROM t2 JOIN t1 ON t1.f3 > WHERE ('v') IN (SELECT f4 FROM t2) > ORDER BY f9; > COUNT(t2.f3) f9 > -0 2 > +0 NULL > EXPLAIN > SELECT COUNT(t2.f3), > (SELECT t2.f1 FROM t1 limit 1) AS f9 > @@ -1285,8 +1290,7 @@ FROM t2 JOIN t1 > WHERE ('v') IN (SELECT f4 FROM t2) > GROUP BY f9; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t2 system NULL NULL NULL NULL 1 > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > @@ -1302,8 +1306,7 @@ FROM t2 JOIN t1 > WHERE ('v') IN (SELECT f4 FROM t2) > ORDER BY f9; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t2 system NULL NULL NULL NULL 1 > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 3 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 > SELECT COUNT(t2.f3), > @@ -1325,7 +1328,7 @@ EXPLAIN > SELECT (SELECT f1 FROM t1) AS field1 FROM t2 GROUP BY field1; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 2 > -2 SUBQUERY t1 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT (SELECT f1 FROM t1) AS field1 FROM t2 GROUP BY field1; > field1 > NULL > @@ -1333,7 +1336,7 @@ EXPLAIN > SELECT (SELECT f1 FROM t1) AS field1 FROM t2 ORDER BY field1; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t2 ALL NULL NULL NULL NULL 2 > -2 SUBQUERY t1 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT (SELECT f1 FROM t1) AS field1 FROM t2 ORDER BY field1; > field1 > NULL > @@ -1631,7 +1634,7 @@ SET SESSION optimizer_switch='in_to_exis > EXPLAIN > SELECT * FROM t1 WHERE ( 6 ) NOT IN ( SELECT t2.f3 FROM t2 JOIN t3 ON t3.f10 = t2.f10); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 DEPENDENT SUBQUERY t3 system NULL NULL NULL NULL 1 > 2 DEPENDENT SUBQUERY t2 ref_or_null f10 f10 10 const,const 2 Using where; Using index > SELECT * FROM t1 WHERE ( 6 ) NOT IN ( SELECT t2.f3 FROM t2 JOIN t3 ON t3.f10 = t2.f10); > @@ -1852,8 +1855,8 @@ GROUP BY 1, 2; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t3 system NULL NULL NULL NULL 1 Using temporary; Using filesort > 1 PRIMARY t2 ALL NULL NULL NULL NULL 2 Using where > -2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -3 MATERIALIZED t1 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 MATERIALIZED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > PREPARE st1 FROM " > SELECT t2.f1, (SELECT f2 FROM t1 WHERE (7) IN (SELECT f1 FROM t1)) > FROM t2 JOIN t3 ON t3.f4 = t2.f4 > > === modified file 'mysql-test/r/subselect_cache.result' > --- a/mysql-test/r/subselect_cache.result 2012-02-15 17:08:08 +0000 > +++ b/mysql-test/r/subselect_cache.result 2012-05-11 15:25:03 +0000 > @@ -486,9 +486,9 @@ Handler_read_key 7 > Handler_read_last 0 > Handler_read_next 0 > Handler_read_prev 0 > -Handler_read_rnd 10 > +Handler_read_rnd 0 > Handler_read_rnd_deleted 0 > -Handler_read_rnd_next 42 > +Handler_read_rnd_next 31 > set optimizer_switch='subquery_cache=off'; > flush status; > select a from t1 ORDER BY (select d from t2 where b=c); > @@ -514,9 +514,9 @@ Handler_read_key 0 > Handler_read_last 0 > Handler_read_next 0 > Handler_read_prev 0 > -Handler_read_rnd 10 > +Handler_read_rnd 0 > Handler_read_rnd_deleted 0 > -Handler_read_rnd_next 72 > +Handler_read_rnd_next 61 > set optimizer_switch='subquery_cache=on'; > #single value subquery test (distinct ORDER BY) > flush status; > > === modified file 'mysql-test/r/subselect_innodb.result' > --- a/mysql-test/r/subselect_innodb.result 2012-01-13 14:50:02 +0000 > +++ b/mysql-test/r/subselect_innodb.result 2012-05-11 15:25:03 +0000 > @@ -272,8 +272,8 @@ FROM t2 > WHERE (SELECT DISTINCT b FROM t3) > 0); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 const PRIMARY PRIMARY 4 const 1 Using where; Using index > -2 SUBQUERY t2 ALL NULL NULL NULL NULL 1 > -3 SUBQUERY t3 ALL NULL NULL NULL NULL 1 Using temporary > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +3 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 0 Using temporary > SELECT * > FROM t1 > WHERE t1.a = ( > @@ -301,7 +301,7 @@ GROUP BY 1 > ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 1 > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 3 SUBQUERY t3 ALL NULL NULL NULL NULL 1 > SELECT MAX( f1 ) FROM t2 > WHERE f2 >= ( > > === modified file 'mysql-test/r/subselect_mat.result' > --- a/mysql-test/r/subselect_mat.result 2012-03-17 08:26:58 +0000 > +++ b/mysql-test/r/subselect_mat.result 2012-05-11 15:25:03 +0000 > @@ -1152,8 +1152,8 @@ create table t2 (b1 int); > insert into t1 values (5); > explain select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 MATERIALIZED t2 system NULL NULL NULL NULL 0 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 MATERIALIZED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > min(a1) > NULL > @@ -1162,8 +1162,8 @@ set @@optimizer_switch=@optimizer_switch > set @@optimizer_switch='materialization=off,in_to_exists=on'; > explain select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > min(a1) > NULL > @@ -1171,8 +1171,8 @@ set @@optimizer_switch=@optimizer_switch > set @@optimizer_switch='semijoin=off'; > explain select min(a1) from t1 where 7 in (select b1 from t2); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 MATERIALIZED t2 system NULL NULL NULL NULL 0 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 MATERIALIZED NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2); > min(a1) > NULL > @@ -1181,16 +1181,16 @@ set @@optimizer_switch='materialization= > # with MariaDB and MWL#90, this particular case is solved: > explain select min(a1) from t1 where 7 in (select b1 from t2); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2); > min(a1) > NULL > # but when we go around MWL#90 code, the problem still shows up: > explain select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > min(a1) > NULL > @@ -1932,7 +1932,7 @@ WHERE s2.d = s1.e AND s1.e = (SELECT MAX > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 1 PRIMARY t2 index NULL c 5 NULL 8 Using where; Using index > -2 MATERIALIZED s2 ref d d 4 const 1 Using where; Using index > +2 MATERIALIZED s2 ref d d 4 const 2 Using where; Using index > 2 MATERIALIZED s1 ALL NULL NULL NULL NULL 8 Using where; Using join buffer (flat, BNL join) > 3 SUBQUERY t2 ALL NULL NULL NULL NULL 8 > SELECT a, c FROM t1, t2 > @@ -1951,8 +1951,8 @@ WHERE s2.d = s1.e AND s1.e = (SELECT MAX > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 1 PRIMARY t2 index NULL c 5 NULL 8 Using where; Using index > -2 MATERIALIZED s2 ref d d 4 const 1 Using where; Using index > -2 MATERIALIZED s1 hash_ALL NULL #hash#$hj 5 test.s2.d 8 Using where; Using join buffer (flat, BNLH join) > +2 MATERIALIZED s2 ref d d 4 const 2 Using where; Using index > +2 MATERIALIZED s1 ALL NULL NULL NULL NULL 8 Using where; Using join buffer (flat, BNL join) > 3 SUBQUERY t2 ALL NULL NULL NULL NULL 8 > SELECT a, c FROM t1, t2 > WHERE (a, c) IN (SELECT s1.b, s1.c FROM t2 AS s1, t2 AS s2 > @@ -2153,7 +2153,7 @@ set @@optimizer_switch='materialization= > EXPLAIN > SELECT * FROM t2 WHERE (SELECT f3a FROM t3) NOT IN (SELECT f1a FROM t1); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t2 ALL NULL NULL NULL NULL 2 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 3 MATERIALIZED t1 ALL NULL NULL NULL NULL 2 > 2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT * FROM t2 WHERE (SELECT f3a FROM t3) NOT IN (SELECT f1a FROM t1); > @@ -2222,10 +2222,10 @@ NULL > EXPLAIN EXTENDED > SELECT MAX(t1.b) AS max_res FROM t1 WHERE (9) IN (SELECT a FROM t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 index NULL b 5 NULL 2 100.00 Using index > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 MATERIALIZED t2 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select max(`test`.`t1`.`b`) AS `max_res` from `test`.`t1` where <expr_cache><9>(<in_optimizer>(9,9 in ( <materialize> (select `test`.`t2`.`a` from `test`.`t2` ), <primary_index_lookup>(9 in <temporary table> on distinct_key where ((9 = `<subquery2>`.`a`)))))) > +Note 1003 select max(`test`.`t1`.`b`) AS `max_res` from `test`.`t1` where 0 > set @@optimizer_switch='materialization=off,in_to_exists=on,semijoin=off'; > SELECT MAX(t1.b) AS max_res FROM t1 WHERE (9) IN (SELECT a FROM t2); > max_res > @@ -2233,10 +2233,10 @@ NULL > EXPLAIN EXTENDED > SELECT MAX(t1.b) AS max_res FROM t1 WHERE (9) IN (SELECT a FROM t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 index NULL b 5 NULL 2 100.00 Using index > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 DEPENDENT SUBQUERY t2 ALL NULL NULL NULL NULL 2 100.00 Using where > Warnings: > -Note 1003 select max(`test`.`t1`.`b`) AS `max_res` from `test`.`t1` where <expr_cache><9>(<in_optimizer>(9,<exists>(select `test`.`t2`.`a` from `test`.`t2` where (9 = `test`.`t2`.`a`)))) > +Note 1003 select max(`test`.`t1`.`b`) AS `max_res` from `test`.`t1` where 0 > DROP TABLE t1,t2; > # > # LPBUG#825095: Wrong result with materialization and NOT IN with 2 expressions > > === modified file 'mysql-test/r/subselect_mat_cost_bugs.result' > --- a/mysql-test/r/subselect_mat_cost_bugs.result 2012-01-13 14:50:02 +0000 > +++ b/mysql-test/r/subselect_mat_cost_bugs.result 2012-05-11 15:25:03 +0000 > @@ -124,7 +124,7 @@ FROM t3 RIGHT JOIN t1 ON t1.pk = t3.f1 > WHERE t3.f3 OR ( 3 ) IN ( SELECT f2 FROM t2 ); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > SELECT t1.* > FROM t3 RIGHT JOIN t1 ON t1.pk = t3.f1 > WHERE t3.f3 OR ( 3 ) IN ( SELECT f2 FROM t2 ); > @@ -148,7 +148,7 @@ FROM t2 GROUP BY f1 > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > 2 SUBQUERY t1 system NULL NULL NULL NULL 1 > -3 SUBQUERY t2 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +3 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > drop table t1, t2, t3; > # > # LP BUG#715034 Item_sum_distinct::clear(): Assertion `tree != 0' failed > > === modified file 'mysql-test/r/subselect_no_mat.result' > --- a/mysql-test/r/subselect_no_mat.result 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/r/subselect_no_mat.result 2012-05-11 15:25:03 +0000 > @@ -378,12 +378,12 @@ INSERT INTO t8 (pseudo,email) VALUES ('j > INSERT INTO t8 (pseudo,email) VALUES ('2joce1','2test1'); > EXPLAIN EXTENDED SELECT pseudo,(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce')) FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where; Using index > +1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > 4 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > -2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where > +2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 > 3 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > Warnings: > -Note 1003 select `test`.`t8`.`pseudo` AS `pseudo`,(select `test`.`t8`.`email` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1))) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1)) > +Note 1003 select 'joce' AS `pseudo`,(select 'test' from `test`.`t8` where 1) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where 1 > SELECT pseudo FROM t8 WHERE pseudo=(SELECT pseudo,email FROM > t8 WHERE pseudo='joce'); > ERROR 21000: Operand should contain 1 column(s) > @@ -524,6 +524,7 @@ ERROR 21000: Subquery returns more than > show warnings; > Level Code Message > Error 1242 Subquery returns more than 1 row > +Error 1028 Sort aborted: Subquery returns more than 1 row > drop table t1; > create table t1 (a int); > insert into t1 values (1),(2),(3); > @@ -554,11 +555,7 @@ SELECT numreponse, (SELECT numeropost FR > numreponse (SELECT numeropost FROM t1 HAVING numreponse=1) > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 const PRIMARY,numreponse PRIMARY 7 const,const 1 100.00 Using where; Using index > -2 SUBQUERY t1 ref PRIMARY PRIMARY 3 const 2 100.00 Using index > -Warnings: > -Note 1003 select `test`.`t1`.`numreponse` AS `numreponse` from `test`.`t1` where ((`test`.`t1`.`numeropost` = '1') and (`test`.`t1`.`numreponse` = (select 1 from `test`.`t1` where (`test`.`t1`.`numeropost` = '1')))) > +ERROR 21000: Subquery returns more than 1 row > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > ERROR 21000: Subquery returns more than 1 row > EXPLAIN EXTENDED SELECT MAX(numreponse) FROM t1 WHERE numeropost='1'; > @@ -1682,34 +1679,34 @@ select * from t3 where NULL >= any (sele > a > explain extended select * from t3 where NULL >= any (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= any (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= any (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2); > a > explain extended select * from t3 where NULL >= some (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= some (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > insert into t2 values (2,2), (2,1), (3,3), (3,1); > select * from t3 where a > all (select max(b) from t2 group by a); > a > @@ -1771,7 +1768,7 @@ id select_type table type possible_keys > 3 UNION t1 system NULL NULL NULL NULL 1 100.00 > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL NULL > Warnings: > -Note 1003 select 'e' AS `s1` from dual where <nop>(<in_optimizer>('f',(<min>(select 'e' from dual union select 'e' from dual) < 'f'))) > +Note 1003 select 'e' AS `s1` from dual where 1 > drop table t1; > CREATE TABLE t1 (number char(11) NOT NULL default '') ENGINE=MyISAM CHARSET=latin1; > INSERT INTO t1 VALUES ('69294728265'),('18621828126'),('89356874041'),('95895001874'); > @@ -3105,7 +3102,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3117,7 +3114,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3731,8 +3728,8 @@ explain select * from t1 where not exist > ((select t11.i from t1 t11) union (select t12.i from t1 t12)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t11 system NULL NULL NULL NULL 0 const row not found > -3 UNION t12 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 UNION NULL NULL NULL NULL NULL NULL NULL no matching row in const table > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > DROP TABLE t1; > CREATE TABLE t1 (a VARCHAR(250), b INT auto_increment, PRIMARY KEY (b)); > @@ -4202,8 +4199,8 @@ CREATE TABLE t1 (a int, b int, KEY (a)); > INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref a a 5 const 0 Using where; Using index > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > +2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > @@ -4531,13 +4528,13 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <expr_cache><1>(<in_optimizer>(1,<exists>(select min(`test`.`t1`.`a`) from `test`.`t1` group by `test`.`t1`.`a` having (1 = <ref_null_helper>(min(`test`.`t1`.`a`)))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <expr_cache><1>(<in_optimizer>(1,<exists>(select min(`test`.`t1`.`a`) from `test`.`t1` where (`test`.`t1`.`a` > 3) group by `test`.`t1`.`a` having (1 = <ref_null_helper>(min(`test`.`t1`.`a`)))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > DROP TABLE t1; > # > @@ -4628,7 +4625,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <nop>(<in_optimizer>(1,((select max(`test`.`t1`.`a1`) from `test`.`t1`) > 1))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > SELECT 1 FROM t1 WHERE 1 < SOME (SELECT a1 FROM t1); > 1 > 1 > @@ -5961,7 +5958,7 @@ id select_type table type possible_keys > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > DROP TABLE t1; > # > # BUG#12616253 - WRONG RESULT WITH EXISTS(SUBQUERY) (MISSING ROWS) > @@ -6120,7 +6117,7 @@ set optimizer_switch=@tmp_optimizer_swit > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > > DROP TABLE t1; > # > > === modified file 'mysql-test/r/subselect_no_opts.result' > --- a/mysql-test/r/subselect_no_opts.result 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/r/subselect_no_opts.result 2012-05-11 15:25:03 +0000 > @@ -374,12 +374,12 @@ INSERT INTO t8 (pseudo,email) VALUES ('j > INSERT INTO t8 (pseudo,email) VALUES ('2joce1','2test1'); > EXPLAIN EXTENDED SELECT pseudo,(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce')) FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where; Using index > +1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > 4 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > -2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where > +2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 > 3 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > Warnings: > -Note 1003 select `test`.`t8`.`pseudo` AS `pseudo`,(select `test`.`t8`.`email` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1))) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1)) > +Note 1003 select 'joce' AS `pseudo`,(select 'test' from `test`.`t8` where 1) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where 1 > SELECT pseudo FROM t8 WHERE pseudo=(SELECT pseudo,email FROM > t8 WHERE pseudo='joce'); > ERROR 21000: Operand should contain 1 column(s) > @@ -520,6 +520,7 @@ ERROR 21000: Subquery returns more than > show warnings; > Level Code Message > Error 1242 Subquery returns more than 1 row > +Error 1028 Sort aborted: Subquery returns more than 1 row > drop table t1; > create table t1 (a int); > insert into t1 values (1),(2),(3); > @@ -550,11 +551,7 @@ SELECT numreponse, (SELECT numeropost FR > numreponse (SELECT numeropost FROM t1 HAVING numreponse=1) > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 const PRIMARY,numreponse PRIMARY 7 const,const 1 100.00 Using where; Using index > -2 SUBQUERY t1 ref PRIMARY PRIMARY 3 const 2 100.00 Using index > -Warnings: > -Note 1003 select `test`.`t1`.`numreponse` AS `numreponse` from `test`.`t1` where ((`test`.`t1`.`numeropost` = '1') and (`test`.`t1`.`numreponse` = (select 1 from `test`.`t1` where (`test`.`t1`.`numeropost` = '1')))) > +ERROR 21000: Subquery returns more than 1 row > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > ERROR 21000: Subquery returns more than 1 row > EXPLAIN EXTENDED SELECT MAX(numreponse) FROM t1 WHERE numeropost='1'; > @@ -1678,34 +1675,34 @@ select * from t3 where NULL >= any (sele > a > explain extended select * from t3 where NULL >= any (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= any (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= any (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2); > a > explain extended select * from t3 where NULL >= some (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= some (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > insert into t2 values (2,2), (2,1), (3,3), (3,1); > select * from t3 where a > all (select max(b) from t2 group by a); > a > @@ -1767,7 +1764,7 @@ id select_type table type possible_keys > 3 UNION t1 system NULL NULL NULL NULL 1 100.00 > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL NULL > Warnings: > -Note 1003 select 'e' AS `s1` from dual where <nop>(<in_optimizer>('f',(<min>(select 'e' from dual union select 'e' from dual) < 'f'))) > +Note 1003 select 'e' AS `s1` from dual where 1 > drop table t1; > CREATE TABLE t1 (number char(11) NOT NULL default '') ENGINE=MyISAM CHARSET=latin1; > INSERT INTO t1 VALUES ('69294728265'),('18621828126'),('89356874041'),('95895001874'); > @@ -3101,7 +3098,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3113,7 +3110,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3727,8 +3724,8 @@ explain select * from t1 where not exist > ((select t11.i from t1 t11) union (select t12.i from t1 t12)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t11 system NULL NULL NULL NULL 0 const row not found > -3 UNION t12 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 UNION NULL NULL NULL NULL NULL NULL NULL no matching row in const table > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > DROP TABLE t1; > CREATE TABLE t1 (a VARCHAR(250), b INT auto_increment, PRIMARY KEY (b)); > @@ -4198,8 +4195,8 @@ CREATE TABLE t1 (a int, b int, KEY (a)); > INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref a a 5 const 0 Using where; Using index > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > +2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > @@ -4527,13 +4524,13 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <in_optimizer>(1,<exists>(select min(`test`.`t1`.`a`) from `test`.`t1` group by `test`.`t1`.`a` having (1 = <ref_null_helper>(min(`test`.`t1`.`a`))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > 2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <in_optimizer>(1,<exists>(select min(`test`.`t1`.`a`) from `test`.`t1` where (`test`.`t1`.`a` > 3) group by `test`.`t1`.`a` having (1 = <ref_null_helper>(min(`test`.`t1`.`a`))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > DROP TABLE t1; > # > @@ -4624,7 +4621,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <nop>(<in_optimizer>(1,((select max(`test`.`t1`.`a1`) from `test`.`t1`) > 1))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > SELECT 1 FROM t1 WHERE 1 < SOME (SELECT a1 FROM t1); > 1 > 1 > @@ -5957,7 +5954,7 @@ id select_type table type possible_keys > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > DROP TABLE t1; > # > # BUG#12616253 - WRONG RESULT WITH EXISTS(SUBQUERY) (MISSING ROWS) > @@ -6116,7 +6113,7 @@ set optimizer_switch=@tmp_optimizer_swit > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > > DROP TABLE t1; > # > > === modified file 'mysql-test/r/subselect_no_scache.result' > --- a/mysql-test/r/subselect_no_scache.result 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/r/subselect_no_scache.result 2012-05-11 15:25:03 +0000 > @@ -377,12 +377,12 @@ INSERT INTO t8 (pseudo,email) VALUES ('j > INSERT INTO t8 (pseudo,email) VALUES ('2joce1','2test1'); > EXPLAIN EXTENDED SELECT pseudo,(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce')) FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where; Using index > +1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > 4 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > -2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where > +2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 > 3 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > Warnings: > -Note 1003 select `test`.`t8`.`pseudo` AS `pseudo`,(select `test`.`t8`.`email` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1))) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1)) > +Note 1003 select 'joce' AS `pseudo`,(select 'test' from `test`.`t8` where 1) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where 1 > SELECT pseudo FROM t8 WHERE pseudo=(SELECT pseudo,email FROM > t8 WHERE pseudo='joce'); > ERROR 21000: Operand should contain 1 column(s) > @@ -523,6 +523,7 @@ ERROR 21000: Subquery returns more than > show warnings; > Level Code Message > Error 1242 Subquery returns more than 1 row > +Error 1028 Sort aborted: Subquery returns more than 1 row > drop table t1; > create table t1 (a int); > insert into t1 values (1),(2),(3); > @@ -553,11 +554,7 @@ SELECT numreponse, (SELECT numeropost FR > numreponse (SELECT numeropost FROM t1 HAVING numreponse=1) > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 const PRIMARY,numreponse PRIMARY 7 const,const 1 100.00 Using where; Using index > -2 SUBQUERY t1 ref PRIMARY PRIMARY 3 const 2 100.00 Using index > -Warnings: > -Note 1003 select `test`.`t1`.`numreponse` AS `numreponse` from `test`.`t1` where ((`test`.`t1`.`numeropost` = '1') and (`test`.`t1`.`numreponse` = (select 1 from `test`.`t1` where (`test`.`t1`.`numeropost` = '1')))) > +ERROR 21000: Subquery returns more than 1 row > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > ERROR 21000: Subquery returns more than 1 row > EXPLAIN EXTENDED SELECT MAX(numreponse) FROM t1 WHERE numeropost='1'; > @@ -1681,34 +1678,34 @@ select * from t3 where NULL >= any (sele > a > explain extended select * from t3 where NULL >= any (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= any (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= any (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2); > a > explain extended select * from t3 where NULL >= some (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= some (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > insert into t2 values (2,2), (2,1), (3,3), (3,1); > select * from t3 where a > all (select max(b) from t2 group by a); > a > @@ -1770,7 +1767,7 @@ id select_type table type possible_keys > 3 UNION t1 system NULL NULL NULL NULL 1 100.00 > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL NULL > Warnings: > -Note 1003 select 'e' AS `s1` from dual where <nop>(<in_optimizer>('f',(<min>(select 'e' from dual union select 'e' from dual) < 'f'))) > +Note 1003 select 'e' AS `s1` from dual where 1 > drop table t1; > CREATE TABLE t1 (number char(11) NOT NULL default '') ENGINE=MyISAM CHARSET=latin1; > INSERT INTO t1 VALUES ('69294728265'),('18621828126'),('89356874041'),('95895001874'); > @@ -3105,7 +3102,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3117,7 +3114,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3733,8 +3730,8 @@ explain select * from t1 where not exist > ((select t11.i from t1 t11) union (select t12.i from t1 t12)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t11 system NULL NULL NULL NULL 0 const row not found > -3 UNION t12 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 UNION NULL NULL NULL NULL NULL NULL NULL no matching row in const table > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > DROP TABLE t1; > CREATE TABLE t1 (a VARCHAR(250), b INT auto_increment, PRIMARY KEY (b)); > @@ -4204,8 +4201,8 @@ CREATE TABLE t1 (a int, b int, KEY (a)); > INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref a a 5 const 0 Using where; Using index > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > +2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > @@ -4632,7 +4629,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <nop>(<in_optimizer>(1,((select max(`test`.`t1`.`a1`) from `test`.`t1`) > 1))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > SELECT 1 FROM t1 WHERE 1 < SOME (SELECT a1 FROM t1); > 1 > 1 > @@ -5966,7 +5963,7 @@ id select_type table type possible_keys > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > DROP TABLE t1; > # > # BUG#12616253 - WRONG RESULT WITH EXISTS(SUBQUERY) (MISSING ROWS) > @@ -6127,7 +6124,7 @@ set optimizer_switch=@tmp_optimizer_swit > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > > DROP TABLE t1; > # > > === modified file 'mysql-test/r/subselect_no_semijoin.result' > --- a/mysql-test/r/subselect_no_semijoin.result 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/r/subselect_no_semijoin.result 2012-05-11 15:25:03 +0000 > @@ -374,12 +374,12 @@ INSERT INTO t8 (pseudo,email) VALUES ('j > INSERT INTO t8 (pseudo,email) VALUES ('2joce1','2test1'); > EXPLAIN EXTENDED SELECT pseudo,(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce')) FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where; Using index > +1 PRIMARY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > 4 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > -2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using where > +2 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 > 3 SUBQUERY t8 const PRIMARY PRIMARY 37 const 1 100.00 Using index > Warnings: > -Note 1003 select `test`.`t8`.`pseudo` AS `pseudo`,(select `test`.`t8`.`email` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1))) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where (`test`.`t8`.`pseudo` = (select 'joce' from `test`.`t8` where 1)) > +Note 1003 select 'joce' AS `pseudo`,(select 'test' from `test`.`t8` where 1) AS `(SELECT email FROM t8 WHERE pseudo=(SELECT pseudo FROM t8 WHERE pseudo='joce'))` from `test`.`t8` where 1 > SELECT pseudo FROM t8 WHERE pseudo=(SELECT pseudo,email FROM > t8 WHERE pseudo='joce'); > ERROR 21000: Operand should contain 1 column(s) > @@ -520,6 +520,7 @@ ERROR 21000: Subquery returns more than > show warnings; > Level Code Message > Error 1242 Subquery returns more than 1 row > +Error 1028 Sort aborted: Subquery returns more than 1 row > drop table t1; > create table t1 (a int); > insert into t1 values (1),(2),(3); > @@ -550,11 +551,7 @@ SELECT numreponse, (SELECT numeropost FR > numreponse (SELECT numeropost FROM t1 HAVING numreponse=1) > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 const PRIMARY,numreponse PRIMARY 7 const,const 1 100.00 Using where; Using index > -2 SUBQUERY t1 ref PRIMARY PRIMARY 3 const 2 100.00 Using index > -Warnings: > -Note 1003 select `test`.`t1`.`numreponse` AS `numreponse` from `test`.`t1` where ((`test`.`t1`.`numeropost` = '1') and (`test`.`t1`.`numreponse` = (select 1 from `test`.`t1` where (`test`.`t1`.`numeropost` = '1')))) > +ERROR 21000: Subquery returns more than 1 row > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > ERROR 21000: Subquery returns more than 1 row > EXPLAIN EXTENDED SELECT MAX(numreponse) FROM t1 WHERE numeropost='1'; > @@ -1678,34 +1675,34 @@ select * from t3 where NULL >= any (sele > a > explain extended select * from t3 where NULL >= any (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= any (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= any (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2); > a > explain extended select * from t3 where NULL >= some (select b from t2); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > select * from t3 where NULL >= some (select b from t2 group by 1); > a > explain extended select * from t3 where NULL >= some (select b from t2 group by 1); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t3 ALL NULL NULL NULL NULL 3 100.00 > -2 SUBQUERY t2 system NULL NULL NULL NULL 0 0.00 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL NULL no matching row in const table > Warnings: > -Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where <nop>(<in_optimizer>(NULL,((select min(NULL) from `test`.`t2`) <= NULL))) > +Note 1003 select `test`.`t3`.`a` AS `a` from `test`.`t3` where 0 > insert into t2 values (2,2), (2,1), (3,3), (3,1); > select * from t3 where a > all (select max(b) from t2 group by a); > a > @@ -1767,7 +1764,7 @@ id select_type table type possible_keys > 3 UNION t1 system NULL NULL NULL NULL 1 100.00 > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL NULL > Warnings: > -Note 1003 select 'e' AS `s1` from dual where <nop>(<in_optimizer>('f',(<min>(select 'e' from dual union select 'e' from dual) < 'f'))) > +Note 1003 select 'e' AS `s1` from dual where 1 > drop table t1; > CREATE TABLE t1 (number char(11) NOT NULL default '') ENGINE=MyISAM CHARSET=latin1; > INSERT INTO t1 VALUES ('69294728265'),('18621828126'),('89356874041'),('95895001874'); > @@ -3101,7 +3098,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c DESC, t2.b DESC LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3113,7 +3110,7 @@ ON r.a = (SELECT t2.a FROM t2 WHERE t2.c > ORDER BY t2.c, t2.b LIMIT 1) WHERE t1.a = 10; > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 system PRIMARY NULL NULL NULL 1 > -1 PRIMARY r eq_ref PRIMARY PRIMARY 4 const 1 Using where > +1 PRIMARY r const PRIMARY PRIMARY 4 const 1 > 2 DEPENDENT SUBQUERY t2 range b b 40 NULL 2 Using index condition > SELECT sql_no_cache t1.a, r.a, r.b FROM t1 LEFT JOIN t2 r > ON r.a = (SELECT t2.a FROM t2 WHERE t2.c = t1.a AND t2.b <= '359899' > @@ -3727,8 +3724,8 @@ explain select * from t1 where not exist > ((select t11.i from t1 t11) union (select t12.i from t1 t12)); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > -2 SUBQUERY t11 system NULL NULL NULL NULL 0 const row not found > -3 UNION t12 system NULL NULL NULL NULL 0 const row not found > +2 SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > +3 UNION NULL NULL NULL NULL NULL NULL NULL no matching row in const table > NULL UNION RESULT <union2,3> ALL NULL NULL NULL NULL NULL > DROP TABLE t1; > CREATE TABLE t1 (a VARCHAR(250), b INT auto_increment, PRIMARY KEY (b)); > @@ -4198,8 +4195,8 @@ CREATE TABLE t1 (a int, b int, KEY (a)); > INSERT INTO t1 VALUES (1,1),(2,1); > EXPLAIN SELECT 1 FROM t1 WHERE a = (SELECT COUNT(*) FROM t1 GROUP BY b); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 ref a a 5 const 0 Using where; Using index > -2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 Using temporary; Using filesort > +1 PRIMARY t1 ref a a 5 const 1 Using where; Using index > +2 SUBQUERY internal_tmp_table ALL group_key NULL NULL NULL 1 Using temporary; Using filesort > DROP TABLE t1; > CREATE TABLE t1 (id int NOT NULL, st CHAR(2), INDEX idx(id)); > INSERT INTO t1 VALUES > @@ -4525,15 +4522,15 @@ SET join_cache_level=0; > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > -2 MATERIALIZED t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <expr_cache><1>(<in_optimizer>(1,1 in ( <materialize> (select min(`test`.`t1`.`a`) from `test`.`t1` group by `test`.`t1`.`a` ), <primary_index_lookup>(1 in <temporary table> on distinct_key where ((1 = `<subquery2>`.`min(a)`)))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > EXPLAIN EXTENDED SELECT 1 FROM t1 WHERE 1 IN (SELECT min(a) FROM t1 WHERE a > 3 GROUP BY a); > id select_type table type possible_keys key key_len ref rows filtered Extra > -1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > -2 MATERIALIZED t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE > +2 DEPENDENT SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 Using where; Using temporary > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <expr_cache><1>(<in_optimizer>(1,1 in ( <materialize> (select min(`test`.`t1`.`a`) from `test`.`t1` where (`test`.`t1`.`a` > 3) group by `test`.`t1`.`a` ), <primary_index_lookup>(1 in <temporary table> on distinct_key where ((1 = `<subquery2>`.`min(a)`)))))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 0 > SET join_cache_level=@save_join_cache_level; > DROP TABLE t1; > # > @@ -4624,7 +4621,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 ALL NULL NULL NULL NULL 2 100.00 > 2 SUBQUERY t1 ALL NULL NULL NULL NULL 2 100.00 > Warnings: > -Note 1003 select 1 AS `1` from `test`.`t1` where <nop>(<in_optimizer>(1,((select max(`test`.`t1`.`a1`) from `test`.`t1`) > 1))) > +Note 1003 select 1 AS `1` from `test`.`t1` where 1 > SELECT 1 FROM t1 WHERE 1 < SOME (SELECT a1 FROM t1); > 1 > 1 > @@ -5957,7 +5954,7 @@ id select_type table type possible_keys > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > DROP TABLE t1; > # > # BUG#12616253 - WRONG RESULT WITH EXISTS(SUBQUERY) (MISSING ROWS) > @@ -6116,7 +6113,7 @@ set optimizer_switch=@tmp_optimizer_swit > EXPLAIN SELECT * FROM t1 WHERE EXISTS (SELECT * FROM t1 WHERE a=7); > id select_type table type possible_keys key key_len ref rows Extra > 1 PRIMARY t1 ALL NULL NULL NULL NULL 4 > -2 SUBQUERY t1 ref a a 5 const 1 Using index > +2 SUBQUERY t1 ref a a 5 const 1 > > DROP TABLE t1; > # > > === modified file 'mysql-test/r/subselect_sj_mat.result' > --- a/mysql-test/r/subselect_sj_mat.result 2012-04-07 13:58:46 +0000 > +++ b/mysql-test/r/subselect_sj_mat.result 2012-05-11 15:25:03 +0000 > @@ -1197,8 +1197,8 @@ set @@optimizer_switch=@optimizer_switch > set @@optimizer_switch='materialization=off,in_to_exists=on'; > explain select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY t2 system NULL NULL NULL NULL 0 const row not found > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select max(b1) from t2 group by b1); > min(a1) > NULL > @@ -1206,8 +1206,8 @@ set @@optimizer_switch=@optimizer_switch > set @@optimizer_switch='semijoin=off'; > explain select min(a1) from t1 where 7 in (select b1 from t2); > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2); > min(a1) > NULL > @@ -1223,8 +1223,8 @@ NULL > # but when we go around MWL#90 code, the problem still shows up: > explain select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > id select_type table type possible_keys key key_len ref rows Extra > -1 PRIMARY t1 system NULL NULL NULL NULL 1 > -2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +1 PRIMARY NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables > +2 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL no matching row in const table > select min(a1) from t1 where 7 in (select b1 from t2) or 2> 4; > min(a1) > NULL > @@ -1971,7 +1971,7 @@ id select_type table type possible_keys > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 1 PRIMARY t2 index c c 5 NULL 8 Using index > 1 PRIMARY <subquery2> eq_ref distinct_key distinct_key 8 func,func 1 > -2 MATERIALIZED s2 ref d d 4 const 1 Using where; Using index > +2 MATERIALIZED s2 ref d d 4 const 2 Using where; Using index > 2 MATERIALIZED s1 ALL c NULL NULL NULL 8 Using where; Using join buffer (flat, BNL join) > 3 SUBQUERY t2 ALL NULL NULL NULL NULL 8 > SELECT a, c FROM t1, t2 > @@ -1991,8 +1991,8 @@ id select_type table type possible_keys > 1 PRIMARY t1 system NULL NULL NULL NULL 1 > 1 PRIMARY t2 index c c 5 NULL 8 Using index > 1 PRIMARY <subquery2> eq_ref distinct_key distinct_key 8 func,func 1 > -2 MATERIALIZED s2 ref d d 4 const 1 Using where; Using index > -2 MATERIALIZED s1 hash_ALL c #hash#$hj 10 const,test.s2.d 8 Using where; Using join buffer (flat, BNLH join) > +2 MATERIALIZED s2 ref d d 4 const 2 Using where; Using index > +2 MATERIALIZED s1 hash_ALL c #hash#$hj 5 const 8 Using where; Using join buffer (flat, BNLH join) > 3 SUBQUERY t2 ALL NULL NULL NULL NULL 8 > SELECT a, c FROM t1, t2 > WHERE (a, c) IN (SELECT s1.b, s1.c FROM t2 AS s1, t2 AS s2 > > === modified file 'mysql-test/suite/sys_vars/r/all_vars.result' > --- a/mysql-test/suite/sys_vars/r/all_vars.result 2012-03-09 07:06:59 +0000 > +++ b/mysql-test/suite/sys_vars/r/all_vars.result 2012-05-11 15:25:03 +0000 > @@ -10,5 +10,6 @@ where length(variable_name) > 50; > select distinct variable_name as `there should be *no* variables listed below:` from t2 > left join t1 on variable_name=test_name where test_name is null; > there should be *no* variables listed below: > +expensive_subquery_limit > drop table t1; > drop table t2; > > === modified file 'mysql-test/t/subselect.test' > --- a/mysql-test/t/subselect.test 2012-05-04 05:16:38 +0000 > +++ b/mysql-test/t/subselect.test 2012-05-11 15:25:03 +0000 > @@ -304,6 +304,7 @@ SELECT (SELECT numeropost FROM t1 HAVING > SELECT numreponse, (SELECT numeropost FROM t1 HAVING numreponse=a) FROM (SELECT * FROM t1) as a; > SELECT numreponse, (SELECT numeropost FROM t1 HAVING numreponse=1) FROM (SELECT * FROM t1) as a; > INSERT INTO t1 (numeropost,numreponse,pseudo) VALUES (1,1,'joce'),(1,2,'joce'),(1,3,'test'); > +-- error ER_SUBQUERY_NO_1_ROW > EXPLAIN EXTENDED SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > -- error ER_SUBQUERY_NO_1_ROW > SELECT numreponse FROM t1 WHERE numeropost='1' AND numreponse=(SELECT 1 FROM t1 WHERE numeropost='1'); > > === modified file 'mysql-test/t/subselect4.test' > --- a/mysql-test/t/subselect4.test 2012-03-01 22:22:22 +0000 > +++ b/mysql-test/t/subselect4.test 2012-05-11 15:25:03 +0000 > @@ -977,6 +977,14 @@ SELECT * FROM t1 WHERE > FROM t3 AS SQ1_t1 JOIN t3 AS SQ1_t3 ON SQ1_t3.f4 > GROUP BY SQ1_t1.f4)); > > +--error ER_SUBQUERY_NO_1_ROW > +SELECT * FROM t1 WHERE > +(SELECT f2 FROM t2 > + WHERE f4 <= ALL > + (SELECT max(SQ1_t1.f4) > + FROM t3 AS SQ1_t1 JOIN t3 AS SQ1_t3 ON SQ1_t3.f4 > + GROUP BY SQ1_t1.f4)); > + > drop table t1, t2, t3; > > --echo # > > === modified file 'sql/item.h' > --- a/sql/item.h 2012-04-10 06:28:13 +0000 > +++ b/sql/item.h 2012-05-11 15:25:03 +0000 > @@ -1395,21 +1395,21 @@ class Item { > { > return cmp_context == IMPOSSIBLE_RESULT || item->cmp_context == cmp_context; > } > - /* > + /** > Test whether an expression is expensive to compute. Used during > optimization to avoid computing expensive expressions during this > phase. Also used to force temp tables when sorting on expensive > functions. > - TODO: > + @todo > Normally we should have a method: > cost Item::execution_cost(), > where 'cost' is either 'double' or some structure of various cost > parameters. > > - NOTE > - This function is now used to prevent evaluation of materialized IN > - subquery predicates before it is allowed. grep for > - DontEvaluateMaterializedSubqueryTooEarly to see the uses. > + @note > + This function is now used to prevent evaluation of expensive subquery > + predicates during the optimization phase. It also prevents evaluation > + of predicates that are not computable at this moment (infinite cost). Could you elaborate more on the "infinite cost" part? (let's discuss it on irc) > */ > virtual bool is_expensive() > { > > === modified file 'sql/item_cmpfunc.cc' > --- a/sql/item_cmpfunc.cc 2012-03-28 18:25:31 +0000 > +++ b/sql/item_cmpfunc.cc 2012-05-11 15:25:03 +0000 > @@ -5539,7 +5539,8 @@ void Item_equal::add_const(Item *c, Item > else > { > Item_func_eq *func= new Item_func_eq(c, const_item); > - func->set_cmp_func(); > + if(func->set_cmp_func()) > + return; Please add space after "if", Please add a comment about what kind of condition can cause the problem. As far as I understand, this is a kind of error? > func->quick_fix_field(); > cond_false= !func->val_int(); > } > > === modified file 'sql/item_cmpfunc.h' > --- a/sql/item_cmpfunc.h 2012-05-04 05:16:38 +0000 > +++ b/sql/item_cmpfunc.h 2012-05-11 15:25:03 +0000 > @@ -370,9 +370,9 @@ class Item_bool_func2 :public Item_int_f > Item_bool_func2(Item *a,Item *b) > :Item_int_func(a,b), cmp(tmp_arg, tmp_arg+1), abort_on_null(FALSE) {} > void fix_length_and_dec(); > - void set_cmp_func() > + int set_cmp_func() > { > - cmp.set_cmp_func(this, tmp_arg, tmp_arg+1, TRUE); > + return cmp.set_cmp_func(this, tmp_arg, tmp_arg+1, TRUE); > } > optimize_type select_optimize() const { return OPTIMIZE_OP; } > virtual enum Functype rev_functype() const { return UNKNOWN_FUNC; } > > === modified file 'sql/item_strfunc.h' > --- a/sql/item_strfunc.h 2012-03-06 19:46:07 +0000 > +++ b/sql/item_strfunc.h 2012-05-11 15:25:03 +0000 > @@ -829,7 +829,7 @@ class Item_func_conv_charset :public Ite > { > DBUG_ASSERT(args[0]->fixed); > conv_charset= cs; > - if (cache_if_const && args[0]->const_item() && !args[0]->with_subselect) > + if (cache_if_const && args[0]->const_item() && !args[0]->is_expensive()) > { > uint errors= 0; > String tmp, *str= args[0]->val_str(&tmp); > > === modified file 'sql/item_subselect.cc' > --- a/sql/item_subselect.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/item_subselect.cc 2012-05-11 15:25:03 +0000 > @@ -522,6 +522,48 @@ void Item_subselect::recalc_used_tables( > */ > } > > + > +/** > + Determine if a subquery is expensive to execute during query optimization. > + > + @details The cost of execution of a subquery is estimated based on an > + estimate of the number of rows the subquery will access during execution. > + This measure is used instead of JOIN::read_time, because it is considered > + to be much more reliable than the cost estimate. > + > + @return true if the subquery is expensive > + @return false otherwise > +*/ > +bool Item_subselect::is_expensive() > +{ > + double examined_rows= 0; > + > + for (SELECT_LEX *sl= unit->first_select(); sl; sl= sl->next_select()) > + { > + JOIN *cur_join= sl->join; > + if (!cur_join) > + continue; > + > + /* If a subquery is not optimized we cannot estimate its cost. */ > + if (!cur_join->join_tab) > + return true; > + > + if (sl->first_inner_unit()) > + { > + /* > + Subqueries that contain subqueries are considered expensive. > + @todo: accumulate the cost of subqueries. > + */ > + return true; > + } > + > + examined_rows+= cur_join->get_examined_rows(); > + } > + > + return (examined_rows > thd->variables.expensive_subquery_limit); > +} > + > + > bool Item_subselect::walk(Item_processor processor, bool walk_subquery, > uchar *argument) > { > > === modified file 'sql/item_subselect.h' > --- a/sql/item_subselect.h 2012-05-04 05:16:38 +0000 > +++ b/sql/item_subselect.h 2012-05-11 15:25:03 +0000 > @@ -191,6 +191,7 @@ class Item_subselect :public Item_result > table_map used_tables() const; > table_map not_null_tables() const { return 0; } > bool const_item() const; > + virtual bool const_pred() const { return const_item(); } What is this function for? It seems to have the same semantics as const_item(), but it is a separate function, which implies there is some difference. Could you add a comment about this? > inline table_map get_used_tables_cache() { return used_tables_cache; } > Item *get_tmp_table_item(THD *thd); > void update_used_tables(); > @@ -209,7 +210,7 @@ class Item_subselect :public Item_result > */ > bool is_evaluated() const; > bool is_uncacheable() const; > - bool is_expensive() { return TRUE; } > + bool is_expensive(); > > /* > Used by max/min subquery to initialize value presence registration > @@ -235,7 +236,7 @@ class Item_subselect :public Item_result > @retval TRUE if the predicate is expensive > @retval FALSE otherwise > */ > - bool is_expensive_processor(uchar *arg) { return TRUE; } > + bool is_expensive_processor(uchar *arg) { return is_expensive(); } > > /** > Get the SELECT_LEX structure associated with this Item. > @@ -581,6 +582,7 @@ class Item_in_subselect :public Item_exi > bool fix_fields(THD *thd, Item **ref); > void fix_length_and_dec(); > void fix_after_pullout(st_select_lex *new_parent, Item **ref); > + bool const_pred() const { return const_item() && left_expr->const_item(); } > void update_used_tables(); > bool setup_mat_engine(); > bool init_left_expr_cache(); > > === modified file 'sql/opt_subselect.cc' > --- a/sql/opt_subselect.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/opt_subselect.cc 2012-05-11 15:25:03 +0000 > @@ -4872,7 +4872,43 @@ static void remove_subq_pushed_predicate > > bool JOIN::optimize_unflattened_subqueries() > { > - return select_lex->optimize_unflattened_subqueries(); > + return select_lex->optimize_unflattened_subqueries(false); > +} > + > +/** > + Optimize all constant subqueries of a query that were not flattened into > + a semijoin. > + > + @details > + Similar to other constant conditions, constant subqueries can be used in > + various constant optimizations. Having optimized constant subqueries before > + these constant optimizations, makes it possible to estimate if a subquery > + is "cheap" enough to be executed during the optimization phase. > + > + Constant subqueries can be optimized and evaluated independent of the outer > + query, therefore if const_only = true, this method can be called early in > + the optimization phase of the outer query. > + > + @return Operation status > + @retval FALSE success. > + @retval TRUE error occurred. > +*/ > + > +bool JOIN::optimize_constant_subqueries() > +{ > + ulonglong save_options= select_lex->options; > + bool res; > + /* > + Constant subqueries may be executed during the optimization phase. > + In EXPLAIN mode the optimizer doesn't initialize many of the data structures > + needed for execution. In order to make it possible to execute subqueries > + during optimization, constant subqueries must be optimized for execution, > + not for EXPLAIN. > + */ > + select_lex->options&= ~SELECT_DESCRIBE; > + res= select_lex->optimize_unflattened_subqueries(true); > + select_lex->options= save_options; > + return res; > } > > > @@ -5273,7 +5309,14 @@ bool JOIN::choose_subquery_plan(table_ma > by the IN predicate. > */ > outer_join= unit->outer_select() ? unit->outer_select()->join : NULL; > - if (outer_join && outer_join->table_count > 0) > + /* > + Get the cost of the outer join if: > + (1) It has at least one table, and > + (2) It has been already optimized (if there is no join_tab, then the > + outer join has not been optimized yet). > + */ > + if (outer_join && outer_join->table_count > 0 && // (1) > + outer_join->join_tab) // (2) > { > /* > TODO: > > === modified file 'sql/sql_class.h' > --- a/sql/sql_class.h 2012-03-11 22:45:18 +0000 > +++ b/sql/sql_class.h 2012-05-11 15:25:03 +0000 > @@ -482,6 +482,7 @@ typedef struct system_variables > ulonglong group_concat_max_len; > ha_rows select_limit; > ha_rows max_join_size; > + ha_rows expensive_subquery_limit; > ulong auto_increment_increment, auto_increment_offset; > ulong lock_wait_timeout; > ulong join_cache_level; > > === modified file 'sql/sql_delete.cc' > --- a/sql/sql_delete.cc 2012-03-24 17:21:22 +0000 > +++ b/sql/sql_delete.cc 2012-05-11 15:25:03 +0000 > @@ -120,7 +120,7 @@ bool mysql_delete(THD *thd, TABLE_LIST * > } > > /* Apply the IN=>EXISTS transformation to all subqueries and optimize them. */ > - if (select_lex->optimize_unflattened_subqueries()) > + if (select_lex->optimize_unflattened_subqueries(false)) > DBUG_RETURN(TRUE); > > const_cond= (!conds || conds->const_item()); > > === modified file 'sql/sql_lex.cc' > --- a/sql/sql_lex.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/sql_lex.cc 2012-05-11 15:25:03 +0000 > @@ -3405,7 +3405,23 @@ bool st_select_lex::add_index_hint (THD > } > > > -bool st_select_lex::optimize_unflattened_subqueries() > +/** > + Optimize all subqueries that have not been flattened into semi-joins. > + > + @details > + This functionality is a method of SELECT_LEX instead of JOIN because > + SQL statements as DELETE/UPDATE do not have a corresponding JOIN object. > + > + @see JOIN::optimize_unflattened_subqueries > + > + @param const_only Restrict subquery optimization to constant subqueries > + > + @return Operation status > + @retval FALSE success. > + @retval TRUE error occurred. > +*/ > + > +bool st_select_lex::optimize_unflattened_subqueries(bool const_only) > { > for (SELECT_LEX_UNIT *un= first_inner_unit(); un; un= un->next_unit()) > { > @@ -3415,11 +3431,17 @@ bool st_select_lex::optimize_unflattened > { > if (subquery_predicate->substype() == Item_subselect::IN_SUBS) > { > - Item_in_subselect *in_subs=(Item_in_subselect*)subquery_predicate; > + Item_in_subselect *in_subs= (Item_in_subselect*) subquery_predicate; > if (in_subs->is_jtbm_merged) > continue; > } > > + if (const_only && !subquery_predicate->const_pred()) > + { > + /* Skip non-constant subqueries if the caller asked so. */ > + continue; > + } > + > bool empty_union_result= true; > /* > If the subquery is a UNION, optimize all the subqueries in the UNION. If > > === modified file 'sql/sql_lex.h' > --- a/sql/sql_lex.h 2012-04-10 06:28:13 +0000 > +++ b/sql/sql_lex.h 2012-05-11 15:25:03 +0000 > @@ -997,12 +997,7 @@ class st_select_lex: public st_select_le > > void clear_index_hints(void) { index_hints= NULL; } > bool is_part_of_union() { return master_unit()->is_union(); } > - /* > - Optimize all subqueries that have not been flattened into semi-joins. > - This functionality is a method of SELECT_LEX instead of JOIN because > - some SQL statements as DELETE do not have a corresponding JOIN object. > - */ > - bool optimize_unflattened_subqueries(); > + bool optimize_unflattened_subqueries(bool const_only); > /* Set the EXPLAIN type for this subquery. */ > void set_explain_type(); > bool handle_derived(LEX *lex, uint phases); > > === modified file 'sql/sql_select.cc' > --- a/sql/sql_select.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/sql_select.cc 2012-05-11 15:25:03 +0000 > @@ -986,7 +986,10 @@ JOIN::optimize() > } > > eval_select_list_used_tables(); > - > + > + if (optimize_constant_subqueries()) > + DBUG_RETURN(1); > + > table_count= select_lex->leaf_tables.elements; > > if (setup_ftfuncs(select_lex)) /* should be after having->fix_fields */ > @@ -1273,6 +1276,12 @@ JOIN::optimize() > { > conds= substitute_for_best_equal_field(NO_PARTICULAR_TAB, conds, > cond_equal, map2table); > + if (thd->is_error()) > + { > + error= 1; > + DBUG_PRINT("error",("Error from substitute_for_best_equal")); > + DBUG_RETURN(1); > + } > conds->update_used_tables(); > DBUG_EXECUTE("where", > print_where(conds, > @@ -1293,6 +1302,12 @@ JOIN::optimize() > *tab->on_expr_ref, > tab->cond_equal, > map2table); > + if (thd->is_error()) > + { > + error= 1; > + DBUG_PRINT("error",("Error from substitute_for_best_equal")); > + DBUG_RETURN(1); > + } > (*tab->on_expr_ref)->update_used_tables(); > } > } > @@ -6592,6 +6607,36 @@ void JOIN::get_prefix_cost_and_fanout(ui > > > /** > + Estimate the number of rows that query execution will read. > + > + @todo This is a very pessimistic upper bound. Use join selectivity > + when available to produce a more realistic number. > +*/ > + > +double JOIN::get_examined_rows() > +{ > + /* Each constant table examines one row, and the result is at most one row. */ > + ha_rows examined_rows= const_tables; > + uint i= const_tables; > + double prev_fanout; > + > + if (table_count == const_tables) > + return examined_rows; > + > + examined_rows+= join_tab[i++].get_examined_rows(); > + for (; i < table_count ; i++) > + { > + if (join_tab[i].type == JT_EQ_REF) > + prev_fanout= 1; > + else > + prev_fanout= best_positions[i-1].records_read; This looks wrong. Declaration of POSITION::records_read has this comment: /* The "fanout": number of output rows that will be produced (after pushed down selection condition is applied) per each row combination of previous tables. */ note the "PER EACH ROW COMBINATION .." part. I would expect that this function would calculate a product of records_read values. > + examined_rows+= join_tab[i].get_examined_rows() * prev_fanout; > + } > + return examined_rows; > +} > + > + > +/** > Find a good, possibly optimal, query execution plan (QEP) by a possibly > exhaustive search. > > @@ -8011,36 +8056,15 @@ JOIN::make_simple_join(JOIN *parent, TAB > row_limit= unit->select_limit_cnt; > do_send_rows= row_limit ? 1 : 0; > > - join_tab->use_join_cache= FALSE; > - join_tab->cache=0; /* No caching */ > + bzero(join_tab, sizeof(JOIN_TAB)); > join_tab->table=temp_table; > - join_tab->cache_select= 0; > - join_tab->select=0; > - join_tab->select_cond= 0; // Avoid valgrind warning > join_tab->set_select_cond(NULL, __LINE__); > - join_tab->quick=0; > join_tab->type= JT_ALL; /* Map through all records */ > join_tab->keys.init(); > join_tab->keys.set_all(); /* test everything in quick */ > - join_tab->info=0; > - join_tab->on_expr_ref=0; > - join_tab->last_inner= 0; > - join_tab->first_unmatched= 0; > join_tab->ref.key = -1; > - join_tab->not_used_in_distinct=0; > join_tab->read_first_record= join_init_read_record; > - join_tab->preread_init_done= FALSE; > join_tab->join= this; > - join_tab->ref.key_parts= 0; > - join_tab->keep_current_rowid= FALSE; > - join_tab->flush_weedout_table= join_tab->check_weed_out_table= NULL; > - join_tab->do_firstmatch= NULL; > - join_tab->loosescan_match_tab= NULL; > - join_tab->emb_sj_nest= NULL; > - join_tab->pre_idx_push_select_cond= NULL; > - join_tab->bush_root_tab= NULL; > - join_tab->bush_children= NULL; > - join_tab->last_leaf_in_bush= FALSE; > bzero((char*) &join_tab->read_record,sizeof(join_tab->read_record)); > temp_table->status=0; > temp_table->null_row=0; > @@ -10225,6 +10249,51 @@ double JOIN_TAB::scan_time() > return res; > } > > + > +/** > + Estimate the number of rows that a an access method will read from a table. > + > + @todo: why not use JOIN_TAB::found_records > +*/ > + > +ha_rows JOIN_TAB::get_examined_rows() > +{ > + ha_rows examined_rows; > + > + if (select && select->quick) > + examined_rows= select->quick->records; > + else if (type == JT_NEXT || type == JT_ALL || > + type == JT_HASH || type ==JT_HASH_NEXT) > + { > + if (limit) > + { > + /* > + @todo This estimate is wrong, a LIMIT query may examine much more rows > + than the LIMIT itself. > + */ > + examined_rows= limit; > + } > + else > + { > + if (table->is_filled_at_execution()) > + examined_rows= records; > + else > + { > + /* > + handler->info(HA_STATUS_VARIABLE) has been called in > + make_join_statistics() > + */ > + examined_rows= table->file->stats.records; > + } > + } > + } > + else > + examined_rows= (ha_rows) records_read; > + > + return examined_rows; > +} > + > + > /** > Initialize the join_tab before reading. > Currently only derived table/view materialization is done here. > @@ -11204,9 +11273,9 @@ static bool check_simple_equality(Item * > if (!item) > { > Item_func_eq *eq_item; > - if ((eq_item= new Item_func_eq(orig_left_item, orig_right_item))) > + if (!(eq_item= new Item_func_eq(orig_left_item, orig_right_item)) || > + eq_item->set_cmp_func()) > return FALSE; > - eq_item->set_cmp_func(); > eq_item->quick_fix_field(); > item= eq_item; > } > @@ -11299,9 +11368,9 @@ static bool check_row_equality(THD *thd, > if (!is_converted) > { > Item_func_eq *eq_item; > - if (!(eq_item= new Item_func_eq(left_item, right_item))) > + if (!(eq_item= new Item_func_eq(left_item, right_item)) || > + eq_item->set_cmp_func()) > return FALSE; > - eq_item->set_cmp_func(); > eq_item->quick_fix_field(); > eq_list->push_back(eq_item); > } > @@ -11987,9 +12056,8 @@ Item *eliminate_item_equal(COND *cond, C > > eq_item= new Item_func_eq(field_item->real_item(), head_item); > > - if (!eq_item) > + if (!eq_item || eq_item->set_cmp_func()) > return 0; > - eq_item->set_cmp_func(); > eq_item->quick_fix_field(); > } > current_sjm= field_sjm; > @@ -12076,7 +12144,7 @@ Item *eliminate_item_equal(COND *cond, C > Item_equal::get_first() for details. > > @return > - The transformed condition > + The transformed condition, or NULL in case of error > */ > > static COND* substitute_for_best_equal_field(JOIN_TAB *context_tab, > @@ -18544,6 +18612,7 @@ check_reverse_order: > tab->ref.key_parts= 0; > if (select_limit < table->file->stats.records) > tab->limit= select_limit; > + table->disable_keyread(); ^^^ This change looks weird. Let's discuss it. > } > } > else if (tab->type != JT_ALL) > @@ -21269,10 +21338,17 @@ static void select_describe(JOIN *join, > } > else > { > - TABLE_LIST *real_table= table->pos_in_table_list; > - item_list.push_back(new Item_string(real_table->alias, > - strlen(real_table->alias), > - cs)); > + TABLE_LIST *real_table= table->pos_in_table_list; > + /* > + Internal temporary tables have no corresponding table reference > + object. Such a table may appear in EXPLAIN when a subquery that needs > + a temporary table has been executed, and JOIN::exec replaced the > + original JOIN with a plan to access the data in the temp table > + (made by JOIN::make_simple_join). > + */ > + const char *tab_name= real_table ? real_table->alias : > + "internal_tmp_table"; > + item_list.push_back(new Item_string(tab_name, strlen(tab_name), cs)); > } > /* "partitions" column */ > if (join->thd->lex->describe & DESCRIBE_PARTITIONS) > @@ -21430,32 +21506,8 @@ static void select_describe(JOIN *join, > } > else > { > - ha_rows examined_rows; > - if (tab->select && tab->select->quick) > - examined_rows= tab->select->quick->records; > - else if (tab->type == JT_NEXT || tab->type == JT_ALL || is_hj) > - { > - if (tab->limit) > - examined_rows= tab->limit; > - else > - { > - if (tab->table->is_filled_at_execution()) > - { > - examined_rows= tab->records; > - } > - else > - { > - /* > - handler->info(HA_STATUS_VARIABLE) has been called in > - make_join_statistics() > - */ > - examined_rows= tab->table->file->stats.records; > - } > - } > - } > - else > - examined_rows=(ha_rows)tab->records_read; > - > + ha_rows examined_rows= tab->get_examined_rows(); > + > item_list.push_back(new Item_int((longlong) (ulonglong) examined_rows, > MY_INT64_NUM_DECIMAL_DIGITS)); > > > === modified file 'sql/sql_select.h' > --- a/sql/sql_select.h 2012-05-04 05:16:38 +0000 > +++ b/sql/sql_select.h 2012-05-11 15:25:03 +0000 > @@ -512,6 +512,7 @@ typedef struct st_join_table { > return (is_hash_join_key_no(key) ? hj_key : table->key_info+key); > } > double scan_time(); > + ha_rows get_examined_rows(); > bool preread_init(); > > bool is_sjm_nest() { return test(bush_children); } > @@ -1281,6 +1282,7 @@ class JOIN :public Sql_alloc > bool alloc_func_list(); > bool flatten_subqueries(); > bool optimize_unflattened_subqueries(); > + bool optimize_constant_subqueries(); > bool make_sum_func_list(List<Item> &all_fields, List<Item> &send_fields, > bool before_group_by, bool recompute= FALSE); > > @@ -1380,6 +1382,7 @@ class JOIN :public Sql_alloc > void get_prefix_cost_and_fanout(uint n_tables, > double *read_time_arg, > double *record_count_arg); > + double get_examined_rows(); > /* defined in opt_subselect.cc */ > bool transform_max_min_subquery(); > /* True if this JOIN is a subquery under an IN predicate. */ > > === modified file 'sql/sql_update.cc' > --- a/sql/sql_update.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/sql_update.cc 2012-05-11 15:25:03 +0000 > @@ -368,7 +368,7 @@ int mysql_update(THD *thd, > } > > /* Apply the IN=>EXISTS transformation to all subqueries and optimize them. */ > - if (select_lex->optimize_unflattened_subqueries()) > + if (select_lex->optimize_unflattened_subqueries(false)) > DBUG_RETURN(TRUE); > > if (select_lex->inner_refs_list.elements && > > === modified file 'sql/sys_vars.cc' > --- a/sql/sys_vars.cc 2012-04-19 14:00:13 +0000 > +++ b/sql/sys_vars.cc 2012-05-11 15:25:03 +0000 > @@ -3780,4 +3780,9 @@ static Sys_var_ulong Sys_debug_binlog_fs > CMD_LINE(REQUIRED_ARG), > VALID_RANGE(0, UINT_MAX), DEFAULT(0), BLOCK_SIZE(1)); > #endif > - > +static Sys_var_harows Sys_expensive_subquery_limit( > + "expensive_subquery_limit", > + "The maximum number of rows a subquery examines in order to be " > + "considered non-expensive", > + SESSION_VAR(expensive_subquery_limit), CMD_LINE(REQUIRED_ARG), > + VALID_RANGE(0, HA_POS_ERROR), DEFAULT(100), BLOCK_SIZE(1)); > > === modified file 'sql/table.cc' > --- a/sql/table.cc 2012-05-04 05:16:38 +0000 > +++ b/sql/table.cc 2012-05-11 15:25:03 +0000 > @@ -5985,7 +5985,8 @@ void TABLE::use_index(int key_to_save) > > bool TABLE::is_filled_at_execution() > { > - return test(pos_in_table_list->jtbm_subselect || > + return test(!pos_in_table_list || > + pos_in_table_list->jtbm_subselect || > pos_in_table_list->is_active_sjm()); Could you please add a comment for pos_in_table_list that it can be NULL for some tables? (I assume you still remember what kind of tables have pos_in_table_list==NULL?) > } > > A general question: was this expected to work: MariaDB [j3]> explain extended select * from t1 where 33 in (select b from five) or c > 11; +------+--------------------+-------+------+---------------+------+---------+------+------+----------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +------+--------------------+-------+------+---------------+------+---------+------+------+----------+-------------+ | 1 | PRIMARY | t1 | ALL | NULL | NULL | NULL | NULL | 10 | 100.00 | Using where | | 2 | DEPENDENT SUBQUERY | five | ALL | NULL | NULL | NULL | NULL | 5 | 100.00 | Using where | +------+--------------------+-------+------+---------------+------+---------+------+------+----------+-------------+ 2 rows in set, 1 warning (22 min 25.44 sec) The subuqery is constant, table `five` has 5 records, and none of them has b=33. When debugging, I see JOIN::get_examined_rows() to be invoked and it returns 5 rows, but EXPLAIN still doesn't show "Impossible WHERE"? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 1

[Maria-developers] Question about Query plan
by Markus Pilman 14 May '12

14 May '12

Hello all, At ETH Zurich we are working on a new storage engine, that allows us to test several new architectures for transactional databases. So far we worked with MySQL, but we had massive performance issues. After some investigation we figured out, that MySQL generates different query plans for InnoDB than for our engine. One query which killed our performance was the following (this is a query from the TPC-W benchmark): SELECT ol2.ol_i_id, SUM(ol2.ol_qty) AS sum_ol FROM order_line ol, order_line ol2, (SELECT o_id FROM orders ORDER BY o_date DESC LIMIT 10000) AS t WHERE ol.ol_o_id = t.o_id AND ol.ol_i_id = 10 AND ol2.ol_o_id = t.o_id AND ol2.ol_i_id <> 10 GROUP BY ol2.ol_i_id ORDER BY sum_ol DESC LIMIT 0,5 MySQL generated the following plan for InnoDB: +----+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ | 1 | PRIMARY | <derived2> | ALL | NULL | NULL | NULL | NULL | 10000 | Using temporary; Using filesort | | 1 | PRIMARY | ol | ref | orderline_o_id,orderline_i_id | orderline_o_id | 8 | t.o_id | 1 | Using where | | 1 | PRIMARY | ol2 | ref | orderline_o_id,orderline_i_id | orderline_o_id | 8 | tpcw.ol.OL_O_ID | 1 | Using where | | 2 | DERIVED | orders | index | NULL | orders_o_date | 4 | NULL | 10000 | Using index | +----+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ while it generated the following one for our storage engine: +----+-------------+------------+-------+-------------------------------+----------------+---------+-------+-------+----------------------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+------------+-------+-------------------------------+----------------+---------+-------+-------+----------------------------------------------+ | 1 | PRIMARY | ol | ref | orderline_o_id,orderline_i_id | orderline_i_id | 5 | const | 10 | Using where; Using temporary; Using filesort | | 1 | PRIMARY | ol2 | range | orderline_o_id,orderline_i_id | orderline_i_id | 5 | NULL | 20 | Using where; Using join buffer | | 1 | PRIMARY | <derived2> | ALL | NULL | NULL | NULL | NULL | 10000 | Using where; Using join buffer | | 2 | DERIVED | orders | index | NULL | orders_o_date | 4 | NULL | 10000 | | +----+-------------+------------+-------+-------------------------------+----------------+---------+-------+-------+----------------------------------------------+ The second one is obviously a very bad one. So we decided to try with MariaDB, which generates the following query plan: +------+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +------+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ | 1 | PRIMARY | ol | ref | orderline_o_id,orderline_i_id | orderline_i_id | 5 | const | 10 | Using temporary; Using filesort | | 1 | PRIMARY | ol2 | ref | orderline_o_id,orderline_i_id | orderline_o_id | 8 | test.ol.OL_O_ID | 11 | Using where | | 1 | PRIMARY | <derived2> | ref | key0 | key0 | 8 | test.ol.OL_O_ID | 10 | | | 2 | DERIVED | orders | index | NULL | orders_o_date | 4 | NULL | 10000 | | +------+-------------+------------+-------+-------------------------------+----------------+---------+-----------------+-------+---------------------------------+ The query plan from MariaDB looks sane to me, and the numbers approve this (the query runs on a middle sized data set about 200 times faster with MariaDB than with MySQL). So we will continue our work with MariaDB. But I have a question to these query plans: why are we getting this differences in MySQL between our storage engine and InnoDB? Is there a feature in our storage engine missing (we first thought we need the ability to support HA_KEYREAD_ONLY - but implementing this feature did not change the query plan)? Or does MySQL some kind of "cheating"? We should understand this issue to be able to present our results we get later (may be we will compare MariaDB and MySQL, but in a paper we would have to explain why MySQL sucks that much). And btw: good work with MariaDB!! The optimizer seems to do a much better job than MySQL - even with InnoDB/XtraDB (we had to rewrite some queries in MySQL to force it to generate sane query plans - with MariaDB this does not seem to be necessary anymore). Thanks for your help in advance and best regards Markus

2 4

Re: [Maria-developers] Review request for: [Commits] Rev 3403: Fix for bug lp:944706, task MDEV-193 in file:///home/tsk/mprog/src/5.5-lpb944706/
by Sergei Petrunia 14 May '12

14 May '12

On Sun, May 13, 2012 at 02:11:15PM +0300, Timour Katchaounov wrote: > === modified file 'sql/sys_vars.cc' > --- a/sql/sys_vars.cc 2012-04-19 14:00:13 +0000 > +++ b/sql/sys_vars.cc 2012-05-11 15:25:03 +0000 > @@ -3780,4 +3780,9 @@ static Sys_var_ulong Sys_debug_binlog_fs > CMD_LINE(REQUIRED_ARG), > VALID_RANGE(0, UINT_MAX), DEFAULT(0), BLOCK_SIZE(1)); > #endif > - > +static Sys_var_harows Sys_expensive_subquery_limit( > + "expensive_subquery_limit", > + "The maximum number of rows a subquery examines in order to be " > + "considered non-expensive", > + SESSION_VAR(expensive_subquery_limit), CMD_LINE(REQUIRED_ARG), > + VALID_RANGE(0, HA_POS_ERROR), DEFAULT(100), BLOCK_SIZE(1)); "considered non-expensive" may not mean much for the outside user. Please change the wording to explicitly say that subqueries that enumerate less than expensive_subquery_limit rows may be evaluated at the optimization phase. Which version will this fix be pushed to? 5.5-ga? BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] federated bug: Bug #61526 create table .. like .. federated table crashes with auto_increment column
by xiaobing jiang 14 May '12

14 May '12

hi all: I try to fix the bug: http://bugs.mysql.com/bug.php?id=61526. but after dig the code, I found that the federated seems don't support "create table xx like yy". the function miss the "share->connect_string", so it always report: ERROR 1 (HY000): server name: '' doesn't exist! from the page: http://kb.askmonty.org/en/about-federatedx, it says: The FederatedX Storage Engine is a fork of the Federated Storage Engine, the latter of the two no longer being maintained by Oracle. The purpose of FederatedX is to keep this storage engine's development progressing-- to both add new features as well as fix old bugs. is it really? so will oracle accept the patch? Thanks!

3 4

Re: [Maria-developers] Added --continue-on-error to mysqltest and mysql-test-run in lp:maria/5.5
by Sergei Golubchik 12 May '12

12 May '12

Hi, Monty! On May 04, Michael Widenius wrote: > At lp:maria/5.5 > > ------------------------------------------------------------ > revno: 3393 > revision-id: monty(a)askmonty.org-20120504133709-w1l2t1vhvpqupu3v > parent: monty(a)askmonty.org-20120503130041-fb9myku0uh1qybty > committer: Michael Widenius <monty(a)askmonty.org> > branch nick: maria-5.5 > timestamp: Fri 2012-05-04 16:37:09 +0300 > message: > Added --continue-on-error to mysqltest and mysql-test-run > This will contune the test case even if there was an error. This makes it easier to run a test that contains many sub tests against one engine. First, you forgot to test the new feature. see mysqltest.test for examples > === modified file 'client/mysqltest.cc' > --- a/client/mysqltest.cc 2012-02-23 06:50:11 +0000 > +++ b/client/mysqltest.cc 2012-05-04 13:37:09 +0000 > @@ -1746,9 +1791,15 @@ static int diff_check(const char *diff_n > my_snprintf(buf, sizeof(buf), "%s -v", diff_name); > > if (!(res_file= popen(buf, "r"))) > - die("popen(\"%s\", \"r\") failed", buf); > + { > + report_or_die("popen(\"%s\", \"r\") failed", buf); > + return; > + } I doubt it makes sense to continue in this case > - /* if diff is not present, nothing will be in stdout to increment have_diff */ > + /* > + if diff is not present, nothing will be in stdout to increment > + have_diff > + */ > if (fgets(buf, sizeof(buf), res_file)) > have_diff= 1; > > @@ -2069,9 +2120,13 @@ void check_result() > > switch (compare_files(log_file.file_name(), result_file_name)) { > case RESULT_OK: > - break; /* ok */ > + if (!error_count) > + break; /* ok */ > + mess= "Got errors while running test"; I wonder whether you need it. As the error goes into the result, it'll inevitably end up in the .reject file. one can get RESULT_OK only if the same error is present in the .result file. But in this case you won't increment error_count, because an error in the result file means --error directive. > + /* Fallthrough */ > case RESULT_LENGTH_MISMATCH: > - mess= "Result length mismatch\n"; > + if (!mess) > + mess= "Result length mismatch\n"; > /* Fallthrough */ > case RESULT_CONTENT_MISMATCH: > { > @@ -3240,8 +3319,10 @@ void do_exec(struct st_command *command) > log_msg("exec of '%s' failed, error: %d, status: %d, errno: %d", > ds_cmd.str, error, status, errno); > dynstr_free(&ds_cmd); > - die("command \"%s\" failed\n\nOutput from before failure:\n%s\n", > - command->first_argument, ds_res.str); > + if (!opt_continue_on_error) > + die("command \"%s\" failed\n\nOutput from before failure:\n%s\n", > + command->first_argument, ds_res.str); why not report_or_die ? > + return; > } > > DBUG_PRINT("info", > @@ -3941,12 +4030,12 @@ void read_until_delimiter(DYNAMIC_STRING > No characters except \n are allowed on > the same line as the command > */ > - die("Trailing characters found after command"); > + report_or_die("Trailing characters found after command"); I think --continue-on-error should not affect syntax errors in the test script itself. They should always abort the execution. in general, you've documented the new option as "This is mostly useful when testing a storage engine and one wants to see how much of a multi-test file it can execute." which is also what BigFish wanted it to do. So, most (if not all) mysqltest failures not related to a storage engine should not be ignored. > } > > if (feof(cur_file->file)) > - die("End of file encountered before '%s' delimiter was found", > - ds_delimiter->str); > + report_or_die("End of file encountered before '%s' delimiter was found", > + ds_delimiter->str); > > if (match_delimiter(c, ds_delimiter->str, ds_delimiter->length)) > { > === modified file 'mysql-test/mysql-test-run.pl' > --- a/mysql-test/mysql-test-run.pl 2012-05-03 13:00:41 +0000 > +++ b/mysql-test/mysql-test-run.pl 2012-05-04 13:37:09 +0000 > @@ -5805,6 +5807,11 @@ sub start_mysqltest ($) { > mtr_add_arg($args, "--max-connections=%d", $opt_max_connections); > } > > + if ($opt_continue_on_error) > + { > + mtr_add_arg($args, "--continue-on-error"); I think this should enable --force too. or, perhaps, --force should enable this and one does not need a separate --continue-on-error in mtr. > + } > + > if ( $opt_embedded_server ) > { > Regards, Sergei

1 0

[Maria-developers] Request for testing: MWL#182 SHOW EXPLAIN
by Sergei Petrunia 11 May '12

11 May '12

Hi Elena, I need testing for MWL#182 SHOW EXPLAIN. The code is here: lp:~maria-captains/maria/5.3-show-explain2 (note the '2' at the end) The branch is currently based on 5.3, although will move to 5.5 soon (I suppose this doesn't cause any problems for testing). Currently, I can think of two things that need testing: === "Asynchronous Procedure Call" module === This is a module that allows one thread to schedule a call in the other thread, and wait until the call has been invoked. Suggested way to test: - run a set of queries that look at each other with "SHOW EXPLAIN FOR <thread>" statements. - have some "victim" threads which run long queries, and have "attacker" threads which submit a lot of "SHOW EXPLAIN FOR <victim_thread>" queries. === On-the-fly EXPLAIN printing functionality === Before this WL, the code that produced EXPLAIN output - Required that optimizer produces a slightly different structures when EXPLAIN is running - Did destructive modifications to the query plan (that is, it was not possible to call these functions two times to get two EXPLAIN outputs for the same planned query) Suggested way to test: - Run various complicated queries, with UNIONs, JOINs, etc. - While they are running, do SHOW EXPLAIN FOR command for them, expecting it to crash or produce nonsensical output at some point. If SHOW EXPLAIN crashes, the crash report will show the query it was trying to obtain the EXPLAIN for. My current suggestion is to use some "runner" threads to run the queries and use another set of threads to hit the "runner" threads with SHOW EXPLAIN request. When creating tests in mysql-test/t/show_explain.test, I made a framework that allows to stop query execution at some particular point and then wait until some number of SHOW EXPLAIN requests were arrived and served. This requires insertion of "probes" into the source code, though. Perhaps, we could use something like that for your testing: to have the query pause at some "interesting" points and wait for SHOW EXPLAIN requests to arrive and be served. I wasn't able to come up with any intelligent way to find interesting point though, so this is still just an idea. BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 1

Re: [Maria-developers] Added --continue-on-error to mysqltest and mysql-test-run
by Sergei Golubchik 07 May '12

07 May '12

Hi, Monty! On May 04, Michael Widenius wrote: > At lp:maria/5.5 > > ------------------------------------------------------------ > revno: 3393 > committer: Michael Widenius <monty(a)askmonty.org> > branch nick: maria-5.5 > message: > Added --continue-on-error to mysqltest and mysql-test-run > This will contune the test case even if there was an error. This > makes it easier to run a test that contains many sub tests against > one engine. Monty, why did you do your parallel implementation of this feature, instead of using the existing abort_on_error flag? Regards, Sergei

1 0

Re: [Maria-developers] MDEV-255: Compile handlersocket plugin in 5.5
by Sergei Golubchik 07 May '12

07 May '12

Hi, Vladislav! On May 05, Vladislav Vaintroub wrote: > ------------------------------------------------------------ > revno: 3398 > revision-id: wlad(a)montyprogram.com-20120505003610-rz89d70v8c6n5v0c > parent: wlad(a)montyprogram.com-20120504152240-ptie9ox0vjqqukd4 > committer: Vladislav Vaintroub <wlad(a)montyprogram.com> > branch nick: 5.5 > timestamp: Sat 2012-05-05 02:36:10 +0200 > message: > MDEV-255: Compile handlersocket plugin in 5.5 > === added file 'plugin/handler_socket/CMakeLists.txt' > --- a/plugin/handler_socket/CMakeLists.txt 1970-01-01 00:00:00 +0000 > +++ b/plugin/handler_socket/CMakeLists.txt 2012-05-05 00:36:10 +0000 > @@ -0,0 +1,41 @@ > + > +IF(WIN32 OR CMAKE_SYSTEM_NAME MATCHES "FreeBSD") > + # Handlersocket does not compile on Windows, compiles but does > + # not start on FreeBSD. Why does it not start of BSD? > + RETURN() > +ENDIF() > + > +#Remove -fno-implicit-templates from compiler flags(handlersocket would not work with it) > +IF(CMAKE_COMPILER_IS_GNUCXX) > + STRING(REPLACE "-fno-implicit-templates" "" CMAKE_CXX_FLAGS ${CMAKE_CXX_FLAGS}) Do we still compile with -fno-implicit-templates? I thought we don't use it anymore. We shouldn't. > +ENDIF() > + > === modified file 'plugin/handler_socket/handlersocket/mysql_incl.hpp' > --- a/plugin/handler_socket/handlersocket/mysql_incl.hpp 2011-06-07 11:19:49 +0000 > +++ b/plugin/handler_socket/handlersocket/mysql_incl.hpp 2012-05-05 00:36:10 +0000 > @@ -13,7 +13,10 @@ > #define HAVE_CONFIG_H > #endif > > +#ifndef MYSQL_DYNAMIC_PLUGIN > #define MYSQL_DYNAMIC_PLUGIN > +#endif I prefer not to have #ifdef, but instead write #define MYSQL_DYNAMIC_PLUGIN 1 you apparently got macro redefinition warning, right? Because there was -DMYSQL_DYNAMIC_PLUGIN on the gcc command line? It defines MYSQL_DYNAMIC_PLUGIN to be 1, and when it's redefined in the source to the empty string you get a warning. "Redefining" it to the same value of 1 is okay, there will be no warning for it. That's why all such symbols should be defined to 1, not to an empty string. > + > #define MYSQL_SERVER 1 > Regards, Sergei

2 2

[Maria-developers] KB Question: Ubuntu 10.04 LTS Server and MariaDB 5.5.23
by Daniel Bartholomew 07 May '12

07 May '12

A question has been posted to the Knowledgebase: http://kb.askmonty.org/en/ubuntu-1004-lts-server-and-mariadb-5523 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Ubuntu 10.04 LTS Server and MariaDB 5.5.23 After installing fresh mariadb, I'm having troubles using some software compiled with the mysql support. For ex. mysql2 ruby gem: Incorrect MySQL client library version! This gem was compiled for 5.5.23-MariaDB but the client library is 5.1.61 Quick investigation took me into this conclusions: mariadb-client-5.5 introduces new libmariadbclient18, but depends on libdbd-mysql-perl which depends on libmysqlclient16. So after install we have 2 libraries: libmysqlclient.so.16(from original mysql) and libmysqlclient.so.18(from mariadb). And default libmysqlclient.so is symlinked to libmysqlclient.so.16 Can this be solved in some way? - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

2 1

[Maria-developers] A question about JOIN re-execution by subquery code
by Sergei Petrunia 07 May '12

07 May '12

Hi Sanja, As far I understand, JOIN structure has two "modules" to support join re-execution: M1. save_join_tab() saves/restores the contents of JOIN::join_tab array, as if execution process changed JOIN_TAB structures M2. JOIN::tmp_join is used to save/restore the contents of JOIN structure itself. What puzzles me are two questions: Q1: Why do we have separate functions to M1 and M2, instead of one "save everything" call, and one "restore everything" call? Q2: Why does M1 copy the whole JOIN_TAB structures? I've made a patch which makes it to save/restore JOIN::join_tab pointer instead (attached), and it passed all tests in the main test suite. (I am asking all this in connection with bug MDEV-238 on MWL#182 code) BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] MariaDB 5.3.7 release prep
by Daniel Bartholomew 04 May '12

04 May '12

All, The packages and other files for MariaDB 5.3.7 have been uploaded to the primary mirror and we're waiting for the other mirrors to pick up the new files before announcing the release. I've created preliminary changelog and release notes pages for the release here: http://kb.askmonty.org/en/mariadb-537-release-notes http://kb.askmonty.org/en/mariadb-537-changelog From the changelog, it appears this is primarily a bug-fix release. If there are any particularly notable bugs or features of this release which I should make mention of in the release notes, let me know ASAP. Thanks. -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

[Maria-developers] proposed patch for handler APIs for range query information
by Zardosht Kasheff 02 May '12

02 May '12

Hello all, After talking with Colin Charles last night, I have been encouraged to send more patches for MariaDB. Here is a hopefully simple one that benefits all storage engines. Attached is a proposed patch of new handler APIs. The purpose of this patch is to give storage engines more information about impending range queries. The idea is if MySQL is about to perform a range query from a to b, then the handler is informed that a range query is about to happen. It extends functionality like handler::prepare_index_scan to other scenarios. We use it to know how much data to prefetch and prelock before doing a range query. Any storage engine can use this as a heuristic for optimizing range query performance. I would love to hear feedback -Zardosht

2 2

[Maria-developers] KB Question: Aria features ?
by Daniel Bartholomew 30 Apr '12

30 Apr '12

A question has been posted to the Knowledgebase: http://kb.askmonty.org/en/aria-features - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Aria features ? When Aria Engine is a new generation MyISAM Engine , whether the new features: - parallel index creating? - merge table? - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Thanks. -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

[Maria-developers] KB question: parallel index creating
by Daniel Bartholomew 30 Apr '12

30 Apr '12

A question has been posted to the Knowledgebase: http://kb.askmonty.org/en/parallel-index-creating - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - parallel index creating? Hi all can we expect the creation of parallel index creating in MyISAM engine? - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Thanks. -- Daniel Bartholomew MariaDB - http://mariadb.org Monty Program - http://montyprogram.com AskMonty Knowledgebase - http://kb.askmonty.org

1 0

Re: [Maria-developers] Request for testing: MWL#182 SHOW EXPLAIN
by Sergei Petrunia 27 Apr '12

27 Apr '12

On Fri, Apr 27, 2012 at 01:31:08PM +0400, Elena Stepanova wrote: > Hi Sergey, > > OK, I will start working on this. > > > Before this WL, the code that produced EXPLAIN output > > - Required that optimizer produces a slightly different structures > > when EXPLAIN is running > > - Did destructive modifications to the query plan > > Does this mean that the traditional EXPLAIN (preceding a statement or > subsequent to it) is also affected? > Yes. I've changed regular EXPLAIN code so that it can be ran multiple times. That way, traditional EXPLAIN is affected. There is one known difference, currently only documented in this e-mail (will forward it): Date: Sat, 27 Aug 2011 10:14:13 +0400 From: Sergey Petrunya <psergey(a)askmonty.org> To: maria-developers(a)lists.launchpad.net Subject: [Maria-developers] MWL#182: a problem select_type column, SIMPLE vs PRIMARY > By the way, s there any standard way of naming the new EXPLAIN, as > opposed to the old EXPLAIN? (e.g. Online EXPLAIN? Asynchronous? Running? > Parallel?) > This is just to simplify further talks. We call it "SHOW EXPLAIN", following the SQL syntax. Nobody has made an explicit decision on the "official" way to name it, but it looks like this name will stick. > Regards, > /E > > > On 4/27/2012 9:13 AM, Sergei Petrunia wrote: >> Hi Elena, >> >> I need testing for MWL#182 SHOW EXPLAIN. >> >> The code is here: lp:~maria-captains/maria/5.3-show-explain2 (note the '2' at >> the end) >> >> The branch is currently based on 5.3, although will move to 5.5 soon (I suppose >> this doesn't cause any problems for testing). >> >> Currently, I can think of two things that need testing: >> >> === "Asynchronous Procedure Call" module === >> This is a module that allows one thread to schedule a call in the other thread, >> and wait until the call has been invoked. >> >> Suggested way to test: >> - run a set of queries that look at each other with "SHOW EXPLAIN FOR<thread>" >> statements. >> - have some "victim" threads which run long queries, and have "attacker" >> threads which submit a lot of "SHOW EXPLAIN FOR<victim_thread>" queries. >> >> === On-the-fly EXPLAIN printing functionality === >> Before this WL, the code that produced EXPLAIN output >> - Required that optimizer produces a slightly different structures when EXPLAIN >> is running >> - Did destructive modifications to the query plan (that is, it was not possible >> to call these functions two times to get two EXPLAIN outputs for the same >> planned query) >> >> Suggested way to test: >> - Run various complicated queries, with UNIONs, JOINs, etc. >> - While they are running, do SHOW EXPLAIN FOR command for them, expecting >> it to crash or produce nonsensical output at some point. >> If SHOW EXPLAIN crashes, the crash report will show the query it was trying >> to obtain the EXPLAIN for. >> >> My current suggestion is to use some "runner" threads to run the queries and >> use another set of threads to hit the "runner" threads with SHOW EXPLAIN >> request. >> >> When creating tests in mysql-test/t/show_explain.test, I made a framework that >> allows to stop query execution at some particular point and then wait until >> some number of SHOW EXPLAIN requests were arrived and served. >> >> This requires insertion of "probes" into the source code, though. >> >> Perhaps, we could use something like that for your testing: to have the query >> pause at some "interesting" points and wait for SHOW EXPLAIN requests to arrive >> and be served. I wasn't able to come up with any intelligent way to find >> interesting point though, so this is still just an idea. >> >> BR >> Sergei -- BR Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

Re: [Maria-developers] [Bug 988694] Re: Client software depends on server PAM plugin setup
by sergii＠pisem.net 27 Apr '12

27 Apr '12

Hi, Predrag! On Apr 27, Predrag Zecevic wrote: > @Sergei, > > is that MariaDB PAM plugin portable to MySQL? > > Regards. > > BTW, thanks for clarifications about PAM plugin usage for MySQL. I'm taking this out of percona-pam-plugin bug report, hope it's ok. Personally, I didn't try it with MySQL. But because pluggable authentication in MySQL is pretty much the same as in MariaDB (in fact, we've implemented it, and contributed to Oracle), our pam plugin should work. It might need minor cosmetic changes to compile though. Alternatively, you can simply use MariaDB :) it has everything that MySQL does, plus much faster subqueries, much faster replication, and tons of other goodies. Regards, Sergei

2 1

[Maria-developers] lp:maria -> 5.3 or 5.5?
by Sergei Golubchik 27 Apr '12

27 Apr '12

Hi, Currently lp:maria is the same as lp:maria/5.3 - our main 5.3 branch. Typically for projects on launchpad, the short url points to a trunk. In our case it's an old (not even the latest) GA branch. I suggest to move lp:maria to be lp:maria/5.5 at once. And change to lp:maria/trunk as soon as we'll have it. Objections? Regards, Sergei

2 1