April 2010 - developers - lists.mariadb.org

Re: [Maria-developers] Fix windows compiler errors
by Bo Thorsen 10 Apr '10

10 Apr '10

Den 09-04-2010 11:58, Michael Widenius skrev: > > Hi! > >>>>>> "Bo" == Bo Thorsen<bo(a)askmonty.org> writes: > > Bo> Hi all, > Bo> When compiling lp:maria on windows, there are a couple of errors. The > Bo> attached patch fixes those. > > Bo> Part of this patch fixes https://bugs.launchpad.net/maria/+bug/550974, > Bo> and it's done the way Sanja suggested. > > Bo> Bo. > > > === modified file storage/maria/ma_loghandler.c > --- storage/maria/ma_loghandler.c 2010-03-30 12:36:49 +0000 > +++ storage/maria/ma_loghandler.c 2010-04-07 12:12:49 +0000 > @@ -1211,9 +1211,10 @@ > for (file= from_file; file<= to_file; file++) > { > LOGHANDLER_FILE_INFO info; > + File fd; > LINT_INIT(info.max_lsn); > > - File fd= open_logfile_by_number_no_cache(file); > + fd= open_logfile_by_number_no_cache(file); > if ((fd< 0) || > ((translog_read_file_header(&info, fd) || > (cmp_translog_addr(lsn, info.max_lsn)> 0&& > > @@ -1396,8 +1397,9 @@ > > { > LOGHANDLER_FILE_INFO info; > + File fd; > LINT_INIT_STRUCT(info); > - File fd= open_logfile_by_number_no_cache(file); > + fd= open_logfile_by_number_no_cache(file); > if ((fd< 0) || > (translog_read_file_header(&info, fd) | my_close(fd, MYF(MY_WME)))) > { > > Part of the above is already fixed in 5.1 (the same way). > > Please be careful regarding this when you do a merge. Yes, I did see that you fixed one of these already. I removed that hunk from the patch before pushing. Bo.

1 0

[Maria-developers] Rev 34: Extracted tests to run into host specific config files. in file:///Users/hakan/work/monty_program/mariadb-tools/
by Hakan Kuecuekyilmaz 09 Apr '10

09 Apr '10

At file:///Users/hakan/work/monty_program/mariadb-tools/ ------------------------------------------------------------ revno: 34 revision-id: hakan(a)askmonty.org-20100409191359-w2dpenia3bnmlypr parent: hakan(a)askmonty.org-20100408002114-36ml405ax1h61x60 committer: Hakan Kuecuekyilmaz <hakan(a)askmonty.org> branch nick: mariadb-tools timestamp: Fri 2010-04-09 21:13:59 +0200 message: Extracted tests to run into host specific config files. Added OProfile hook. === modified file 'sysbench/conf/au0013.inc' --- a/sysbench/conf/au0013.inc 2010-03-04 02:03:03 +0000 +++ b/sysbench/conf/au0013.inc 2010-04-09 19:13:59 +0000 @@ -27,9 +27,38 @@ # Note: sar works different on Mac OS X. SAR='/usr/bin/sar' +# Other binaries. +SUDO=/usr/bin/sudo + # Directories. TEMP_DIR='/tmp' DATA_DIR="${TEMP_DIR}/data" # Where to compile the sources. -WORK_DIR='/tmp' \ No newline at end of file +WORK_DIR='/tmp' + +# Which concurrency level(s) we want to profile with OProfile +# +# Example: profile 8 and 32 +# DO_OPROFILE='8 32' +#DO_OPROFILE='64' +# There is no OProfile on Mac OS X. +DO_OPROFILE='' + +# sysbench tests to run. +SYSBENCH_TESTS[0]="delete.lua" +SYSBENCH_TESTS[1]="insert.lua" +SYSBENCH_TESTS[2]="oltp_complex_ro.lua" +SYSBENCH_TESTS[3]="oltp_complex_rw.lua" +SYSBENCH_TESTS[4]="oltp_simple.lua" +SYSBENCH_TESTS[5]="select.lua" + +# Default option is --random-points=10. +SYSBENCH_TESTS[6]="select_random_points.lua" + +# Default options are +# --number-of-ranges=10 +# --random-ranges-delta=5. +SYSBENCH_TESTS[7]="select_random_ranges.lua" +SYSBENCH_TESTS[8]="update_index.lua" +SYSBENCH_TESTS[9]="update_non_index.lua" === modified file 'sysbench/conf/lu0012.inc' --- a/sysbench/conf/lu0012.inc 2010-03-09 23:31:27 +0000 +++ b/sysbench/conf/lu0012.inc 2010-04-09 19:13:59 +0000 @@ -25,9 +25,36 @@ # For CPU utilization statistics. MPSTAT='/usr/bin/mpstat' +# Other binaries. +SUDO=/usr/bin/sudo + # Directories. TEMP_DIR='/tmp' DATA_DIR="${TEMP_DIR}/data" # Where to compile the sources. WORK_DIR='/tmp' + +# Which concurrency level(s) we want to profile with OProfile +# +# Example: profile 8 and 32 +# DO_OPROFILE='8 32' +DO_OPROFILE='1' + +# sysbench tests to run. +SYSBENCH_TESTS[0]="delete.lua" +SYSBENCH_TESTS[1]="insert.lua" +SYSBENCH_TESTS[2]="oltp_complex_ro.lua" +SYSBENCH_TESTS[3]="oltp_complex_rw.lua" +SYSBENCH_TESTS[4]="oltp_simple.lua" +SYSBENCH_TESTS[5]="select.lua" + +# Default option is --random-points=10. +SYSBENCH_TESTS[6]="select_random_points.lua" + +# Default options are +# --number-of-ranges=10 +# --random-ranges-delta=5. +SYSBENCH_TESTS[7]="select_random_ranges.lua" +SYSBENCH_TESTS[8]="update_index.lua" +SYSBENCH_TESTS[9]="update_non_index.lua" === modified file 'sysbench/conf/perro.inc' --- a/sysbench/conf/perro.inc 2010-03-09 23:31:27 +0000 +++ b/sysbench/conf/perro.inc 2010-04-09 19:13:59 +0000 @@ -34,3 +34,27 @@ # Where to compile the sources. WORK_DIR='/tmp' + +# Which concurrency level(s) we want to profile with OProfile +# +# Example: profile 8 and 32 +# DO_OPROFILE='8 32' +DO_OPROFILE='64' + +# sysbench tests to run. +SYSBENCH_TESTS[0]="delete.lua" +SYSBENCH_TESTS[1]="insert.lua" +SYSBENCH_TESTS[2]="oltp_complex_ro.lua" +SYSBENCH_TESTS[3]="oltp_complex_rw.lua" +SYSBENCH_TESTS[4]="oltp_simple.lua" +SYSBENCH_TESTS[5]="select.lua" + +# Default option is --random-points=10. +SYSBENCH_TESTS[6]="select_random_points.lua" + +# Default options are +# --number-of-ranges=10 +# --random-ranges-delta=5. +SYSBENCH_TESTS[7]="select_random_ranges.lua" +SYSBENCH_TESTS[8]="update_index.lua" +SYSBENCH_TESTS[9]="update_non_index.lua" === modified file 'sysbench/conf/work.inc' --- a/sysbench/conf/work.inc 2010-03-09 23:31:27 +0000 +++ b/sysbench/conf/work.inc 2010-04-09 19:13:59 +0000 @@ -34,3 +34,27 @@ # Where to compile the sources. WORK_DIR='/tmp' + +# Which concurrency level(s) we want to profile with OProfile +# +# Example: profile 8 and 32 +# DO_OPROFILE='8 32' +DO_OPROFILE='64' + +# sysbench tests to run. +SYSBENCH_TESTS[0]="delete.lua" +SYSBENCH_TESTS[1]="insert.lua" +SYSBENCH_TESTS[2]="oltp_complex_ro.lua" +SYSBENCH_TESTS[3]="oltp_complex_rw.lua" +SYSBENCH_TESTS[4]="oltp_simple.lua" +SYSBENCH_TESTS[5]="select.lua" + +# Default option is --random-points=10. +SYSBENCH_TESTS[6]="select_random_points.lua" + +# Default options are +# --number-of-ranges=10 +# --random-ranges-delta=5. +SYSBENCH_TESTS[7]="select_random_ranges.lua" +SYSBENCH_TESTS[8]="update_index.lua" +SYSBENCH_TESTS[9]="update_non_index.lua" === modified file 'sysbench/run-sysbench-myisam.sh' --- a/sysbench/run-sysbench-myisam.sh 2010-04-08 00:21:14 +0000 +++ b/sysbench/run-sysbench-myisam.sh 2010-04-09 19:13:59 +0000 @@ -10,6 +10,11 @@ # reference please check # http://www.mysqlperformanceblog.com/2009/01/30/linux-schedulers-in-tpcc-lik… # +# For proper work we need these two commands to be run via sudo +# with no password. Example: +# hakan ALL=NOPASSWD: /usr/bin/opcontrol +# hakan ALL=NOPASSWD: /usr/bin/tee /proc/sys/vm/drop_caches +# # Index sizes for 20 mio rows (--table-size=20000000). # * delete.lua 313M sbtest.MYI # * insert.lua 4.0K sbtest.MYI @@ -110,23 +115,6 @@ # CPU utilization, and such. MONITOR_INTERVAL=10 -SYSBENCH_TESTS[0]="delete.lua" -SYSBENCH_TESTS[1]="insert.lua" -SYSBENCH_TESTS[2]="oltp_complex_ro.lua" -SYSBENCH_TESTS[3]="oltp_complex_rw.lua" -SYSBENCH_TESTS[4]="oltp_simple.lua" -SYSBENCH_TESTS[5]="select.lua" - -# Default option is --random-points=10. -SYSBENCH_TESTS[6]="select_random_points.lua" - -# Default options are -# --number-of-ranges=10 -# --random-ranges-delta=5. -SYSBENCH_TESTS[7]="select_random_ranges.lua" -SYSBENCH_TESTS[8]="update_index.lua" -SYSBENCH_TESTS[9]="update_non_index.lua" - # # Note: myisam-max-rows has to match or exceed oltp-table-size # otherwise we get a table full error while preparing the run. @@ -216,6 +204,9 @@ exit 1 fi +# Location of our mysqld binary. +MYSQLD_BINARY="${TEMP_DIR}/build/sql/mysqld" + # # Compile sources. # TODO: Add platform detection and choose proper build script accordingly. @@ -296,11 +287,14 @@ echo "Run time is: $RUN_TIME" >> ${RESULT_DIR}/${TODAY}/${PRODUCT}/sysbench_options.txt # -# Kill possibly left over monitoring processes. +# Clean up possibly left over monitoring processes. # killall -9 iostat killall -9 mpstat killall -9 mysqladmin +$SUDO opcontrol --stop +$SUDO opcontrol --deinit +$SUDO opcontrol --reset for (( i = 0 ; i < ${#SYSBENCH_TESTS[@]} ; i++ )) do @@ -340,6 +334,16 @@ SYSBENCH_OPTIONS_WARM_UP="${SYSBENCH_OPTIONS} --num-threads=3 --max-time=$WARM_UP_TIME" SYSBENCH_OPTIONS_RUN="${SYSBENCH_OPTIONS} --num-threads=$THREADS --max-time=$RUN_TIME" + # Check whether we want a profiled run. + PROFILE_IT=-1 + for l in $DO_OPROFILE + do + if [ x"$l" = x"$THREADS" ]; then + PROFILE_IT=1 + break + fi + done + k=0 while [ $k -lt $LOOP_COUNT ] do @@ -374,8 +378,35 @@ $MYSQLADMIN $MYSQLADMIN_OPTIONS --sleep $MONITOR_INTERVAL status > ${THIS_RESULT_DIR}/server_status${k}.txt 2>&1 & SERVER_STATUS_PID=$! - + + if [ $PROFILE_IT -eq 1 ]; then + $SUDO opcontrol --setup --separate=lib,kernel,thread --no-vmlinux + $SUDO opcontrol --start-daemon + if [ $? != 0 ]; then + echo "[WARNING]: Could not start oprofile daemonl." + echo " Please check your OProfile installation." + fi + + $SUDO opcontrol --start + echo "[$(date "+%Y-%m-%d %H:%M:%S")] This is an OProfile'd sysbench run." + fi + $SYSBENCH $SYSBENCH_OPTIONS_RUN run > ${THIS_RESULT_DIR}/result${k}.txt 2>&1 + + if [ $PROFILE_IT -eq 1 ]; then + PROFILE_IT=-1 + $SUDO opcontrol --dump + $SUDO opcontrol --stop + + opreport --demangle=smart --threshold 0.5 --symbols --long-filenames --merge tgid $MYSQLD_BINARY > ${THIS_RESULT_DIR}/oprofile${k}.txt 2>&1 + + $SUDO opcontrol --deinit + $SUDO opcontrol --reset + fi + + # Copy mysqld error log for future reference. + # TODO: add chrash detection. + cp ${DATA_DIR}/${HOSTNAME}.err ${THIS_RESULT_DIR}/${HOSTNAME}${k}.err sync; sync; sync sleep 1 === modified file 'sysbench/run-sysbench.sh' --- a/sysbench/run-sysbench.sh 2010-04-08 00:21:14 +0000 +++ b/sysbench/run-sysbench.sh 2010-04-09 19:13:59 +0000 @@ -10,6 +10,11 @@ # reference please check # http://www.mysqlperformanceblog.com/2009/01/30/linux-schedulers-in-tpcc-lik… # +# For proper work we need these two commands to be run via sudo +# with no password. Example: +# hakan ALL=NOPASSWD: /usr/bin/opcontrol +# hakan ALL=NOPASSWD: /usr/bin/tee /proc/sys/vm/drop_caches +# # Hakan Kuecuekyilmaz <hakan at askmonty dot org> 2010-02-19. # @@ -107,23 +112,6 @@ # CPU utilization, and such. MONITOR_INTERVAL=10 -SYSBENCH_TESTS[0]="delete.lua" -SYSBENCH_TESTS[1]="insert.lua" -SYSBENCH_TESTS[2]="oltp_complex_ro.lua" -SYSBENCH_TESTS[3]="oltp_complex_rw.lua" -SYSBENCH_TESTS[4]="oltp_simple.lua" -SYSBENCH_TESTS[5]="select.lua" - -# Default option is --random-points=10. -SYSBENCH_TESTS[6]="select_random_points.lua" - -# Default options are -# --number-of-ranges=10 -# --random-ranges-delta=5. -SYSBENCH_TESTS[7]="select_random_ranges.lua" -SYSBENCH_TESTS[8]="update_index.lua" -SYSBENCH_TESTS[9]="update_non_index.lua" - SYSBENCH_OPTIONS="--oltp-table-size=$TABLE_SIZE \ --max-requests=0 \ --mysql-table-engine=InnoDB \ @@ -208,6 +196,9 @@ exit 1 fi +# Location of our mysqld binary. +MYSQLD_BINARY="${TEMP_DIR}/build/sql/mysqld" + # # Compile sources. # TODO: Add platform detection and choose proper build script accordingly. @@ -288,11 +279,14 @@ echo "Run time is: $RUN_TIME" >> ${RESULT_DIR}/${TODAY}/${PRODUCT}/sysbench_options.txt # -# Kill possibly left over monitoring processes. +# Clean up possibly left over monitoring processes. # killall -9 iostat killall -9 mpstat killall -9 mysqladmin +$SUDO opcontrol --stop +$SUDO opcontrol --deinit +$SUDO opcontrol --reset for (( i = 0 ; i < ${#SYSBENCH_TESTS[@]} ; i++ )) do @@ -332,6 +326,16 @@ SYSBENCH_OPTIONS_WARM_UP="${SYSBENCH_OPTIONS} --num-threads=1 --max-time=$WARM_UP_TIME" SYSBENCH_OPTIONS_RUN="${SYSBENCH_OPTIONS} --num-threads=$THREADS --max-time=$RUN_TIME" + # Check whether we want a profiled run. + PROFILE_IT=-1 + for l in $DO_OPROFILE + do + if [ x"$l" = x"$THREADS" ]; then + PROFILE_IT=1 + break + fi + done + k=0 while [ $k -lt $LOOP_COUNT ] do @@ -366,8 +370,35 @@ $MYSQLADMIN $MYSQLADMIN_OPTIONS --sleep $MONITOR_INTERVAL status > ${THIS_RESULT_DIR}/server_status${k}.txt 2>&1 & SERVER_STATUS_PID=$! + if [ $PROFILE_IT -eq 1 ]; then + $SUDO opcontrol --setup --separate=lib,kernel,thread --no-vmlinux + $SUDO opcontrol --start-daemon + if [ $? != 0 ]; then + echo "[WARNING]: Could not start oprofile daemonl." + echo " Please check your OProfile installation." + fi + + $SUDO opcontrol --start + echo "[$(date "+%Y-%m-%d %H:%M:%S")] This is an OProfile'd sysbench run." + fi + $SYSBENCH $SYSBENCH_OPTIONS_RUN run > ${THIS_RESULT_DIR}/result${k}.txt 2>&1 + if [ $PROFILE_IT -eq 1 ]; then + PROFILE_IT=-1 + $SUDO opcontrol --dump + $SUDO opcontrol --stop + + opreport --demangle=smart --threshold 0.5 --symbols --long-filenames --merge tgid $MYSQLD_BINARY > ${THIS_RESULT_DIR}/oprofile${k}.txt 2>&1 + + $SUDO opcontrol --deinit + $SUDO opcontrol --reset + fi + + # Copy mysqld error log for future reference. + # TODO: add chrash detection. + cp ${DATA_DIR}/${HOSTNAME}.err ${THIS_RESULT_DIR}/${HOSTNAME}${k}.err + sync; sync; sync sleep 1

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2847: Remove unnecessary redefinition of TAILQ_EMPTY
by noreply＠launchpad.net 09 Apr '10

09 Apr '10

Merge authors: Bo Thorsen (bo.thorsen) ------------------------------------------------------------ revno: 2847 [merge] committer: bo.thorsen <bo(a)sonofthor.dk> branch nick: trunk timestamp: Fri 2010-04-09 12:45:19 +0200 message: Remove unnecessary redefinition of TAILQ_EMPTY modified: extra/libevent/event-internal.h -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription

1 0

[Maria-developers] Removing a bunch of windows compilation warnings
by Bo Thorsen 09 Apr '10

09 Apr '10

Hi everyone, I have been working on getting rid of the compiler warnings on Windows. I'm not sure if I have the proper fix for this one, so I separated it from the others. TAILQ_EMPTY(head) is defined in extra\libevent\event-internal.h as this: #define TAILQ_EMPTY(head) ((head)->tqh_first == NULL) It's also defined in extra\libevent\compat\sys\queue.h as: #define TAILQ_FIRST(head) ((head)->tqh_first) #define TAILQ_END(head) NULL #define TAILQ_EMPTY(head) \ (TAILQ_FIRST(head) == TAILQ_END(head)) As you can see, these two are actually identical, but the Visual C++ compiler still complains about them, because it doesn't understand that they are identical. There are several ways to fix it. The proper way would be to avoid having the same define in two different places, but I don't know why it's there in the first place. The second is to guard both defines with an #ifndef TAILQ_EMPTY, which works fine, of course. The one I've chosen is to change the definition in queue.h to be identical (character wise) to the other, and this shuts up the compiler. This is the patch: === modified file 'extra/libevent/compat/sys/queue.h' --- extra/libevent/compat/sys/queue.h 2009-03-12 22:27:35 +0000 +++ extra/libevent/compat/sys/queue.h 2010-04-09 09:34:31 +0000 @@ -298,9 +298,7 @@ /* XXX */ #define TAILQ_PREV(elm, headname, field) \ (*(((struct headname *)((elm)->field.tqe_prev))->tqh_last)) -#define TAILQ_EMPTY(head) \ - (TAILQ_FIRST(head) == TAILQ_END(head)) - +#define TAILQ_EMPTY(head) ((head)->tqh_first == NULL) #define TAILQ_FOREACH(var, head, field) \ for((var) = TAILQ_FIRST(head); \ (var) != TAILQ_END(head); \ Ok to push to lp:maria? Bo.

2 2

[Maria-developers] bzr commit into Mariadb 5.2, with Maria 2.0:maria/5.2 branch (knielsen:2763)
by knielsen＠knielsen-hq.org 09 Apr '10

09 Apr '10

#At lp:maria/5.2 2763 knielsen(a)knielsen-hq.org 2010-04-09 [merge] Merge Oqgraph with MariaDB 5.2 trunk added: mysql-test/r/create_options.result mysql-test/r/select_pkeycache.result mysql-test/suite/binlog/t/binlog_index-master.opt mysql-test/suite/pbxt/t/multi_statement-master.opt mysql-test/t/create_options.test mysql-test/t/select_pkeycache-master.opt mysql-test/t/select_pkeycache.test sql/create_options.cc sql/create_options.h modified: .bzrignore BUILD/compile-solaris-sparc CMakeLists.txt Docs/sp-imp-spec.txt client/mysql.cc cmd-line-utils/libedit/filecomplete.c config/ac-macros/plugins.m4 configure.in extra/libevent/evbuffer.c extra/libevent/event.c extra/libevent/signal.c include/keycache.h include/m_ctype.h include/my_base.h include/my_global.h include/my_sys.h include/mysql/plugin.h include/mysql/plugin_auth.h.pp include/sql_common.h libmysqld/CMakeLists.txt libmysqld/Makefile.am mysql-test/extra/rpl_tests/rpl_insert_id_pk.test mysql-test/include/maria_empty_logs.inc mysql-test/lib/My/SafeProcess/safe_process.cc mysql-test/mysql-test-run.pl mysql-test/r/connect.result mysql-test/r/events_bugs.result mysql-test/r/func_time.result mysql-test/r/information_schema.result mysql-test/r/information_schema_all_engines.result mysql-test/r/key_cache.result mysql-test/r/plugin.result mysql-test/r/sp.result mysql-test/r/status_user.result mysql-test/r/system_mysql_db.result mysql-test/suite/binlog/r/binlog_index.result mysql-test/suite/binlog/t/binlog_index.test mysql-test/suite/funcs_1/datadict/processlist_priv.inc mysql-test/suite/funcs_1/r/is_columns_is.result mysql-test/suite/funcs_1/r/is_columns_mysql.result mysql-test/suite/funcs_1/r/is_columns_mysql_embedded.result mysql-test/suite/funcs_1/r/is_tables_is.result mysql-test/suite/pbxt/r/multi_statement.result mysql-test/suite/pbxt/r/mysqlshow.result mysql-test/suite/pbxt/t/status.test mysql-test/suite/rpl/r/rpl_optimize.result mysql-test/suite/rpl/t/rpl_optimize.test mysql-test/t/connect.test mysql-test/t/events_bugs.test mysql-test/t/func_time.test mysql-test/t/key_cache.test mysql-test/t/plugin.test mysql-test/t/sp.test mysys/mf_keycache.c mysys/my_gethostbyname.c plugin/auth/auth_socket.c plugin/auth/dialog.c plugin/daemon_example/daemon_example.cc plugin/fulltext/plugin_example.c scripts/mysql_system_tables.sql scripts/mysql_system_tables_fix.sql server-tools/CMakeLists.txt server-tools/instance-manager/CMakeLists.txt sql-common/client.c sql/CMakeLists.txt sql/Makefile.am sql/client_settings.h sql/event_db_repository.cc sql/events.cc sql/field.cc sql/field.h sql/ha_ndbcluster.cc sql/ha_partition.cc sql/handler.cc sql/handler.h sql/log.cc sql/log_event.cc sql/log_event.h sql/mysql_priv.h sql/mysqld.cc sql/set_var.cc sql/set_var.h sql/share/errmsg.txt sql/slave.cc sql/sp.cc sql/sp_head.cc sql/sql_acl.cc sql/sql_builtin.cc.in sql/sql_class.cc sql/sql_class.h sql/sql_insert.cc sql/sql_lex.h sql/sql_parse.cc sql/sql_plugin.cc sql/sql_plugin.h sql/sql_select.cc sql/sql_show.cc sql/sql_table.cc sql/sql_test.cc sql/sql_view.cc sql/sql_yacc.yy sql/structs.h sql/table.cc sql/table.h sql/unireg.cc storage/archive/ha_archive.cc storage/blackhole/ha_blackhole.cc storage/csv/ha_tina.cc storage/example/ha_example.cc storage/example/ha_example.h storage/federated/ha_federated.cc storage/federatedx/ha_federatedx.cc storage/heap/ha_heap.cc storage/ibmdb2i/ha_ibmdb2i.cc storage/innobase/handler/ha_innodb.cc storage/innodb_plugin/handler/i_s.cc storage/maria/ha_maria.cc storage/maria/ma_loghandler.c storage/myisam/ha_myisam.cc storage/myisam/mi_check.c storage/myisam/mi_close.c storage/myisam/mi_delete_all.c storage/myisam/mi_extra.c storage/myisam/mi_keycache.c storage/myisam/mi_locking.c storage/myisam/mi_page.c storage/myisam/mi_panic.c storage/myisam/mi_preload.c storage/myisam/mi_test1.c storage/myisam/mi_test2.c storage/myisam/mi_test3.c storage/myisam/myisam_ftdump.c storage/myisam/myisamchk.c storage/myisam/myisamdef.h storage/myisam/myisamlog.c storage/myisammrg/ha_myisammrg.cc storage/pbxt/src/discover_xt.cc storage/pbxt/src/ha_pbxt.cc storage/xtradb/handler/ha_innodb.cc storage/xtradb/handler/i_s.cc storage/xtradb/handler/i_s.h storage/xtradb/include/ut0lst.h storage/xtradb/log/log0recv.c strings/my_vsnprintf.c support-files/compiler_warnings.supp === modified file '.bzrignore' --- a/.bzrignore 2010-03-29 15:13:53 +0000 +++ b/.bzrignore 2010-04-08 12:10:05 +0000 @@ -1934,3 +1934,4 @@ client/sql_list.h libmysqld/client_plugin.c sql/client_plugin.c *.dgcov +libmysqld/create_options.cc === modified file 'BUILD/compile-solaris-sparc' --- a/BUILD/compile-solaris-sparc 2008-09-30 20:57:48 +0000 +++ b/BUILD/compile-solaris-sparc 2010-03-31 19:12:21 +0000 @@ -9,6 +9,6 @@ PATH=$PATH:/usr/ccs/bin:/usr/local/bin path=`dirname $0` . "$path/autorun.sh" -CFLAGS="-g -Wimplicit -Wreturn-type -Wswitch -Wtrigraphs -Wcomment -W -Wchar-subscripts -Wformat -Wparentheses -Wsign-compare -Wwrite-strings -Wunused -O3 -fno-omit-frame-pointer -mcpu=v8 -Wa,-xarch=v8plusa" CXX=gcc CXXFLAGS="-Wimplicit -Wreturn-type -Wswitch -Wtrigraphs -Wcomment -W -Wchar-subscripts -Wformat -Wparentheses -Wsign-compare -Wwrite-strings -Woverloaded-virtual -Wsign-promo -Wreorder -Wctor-dtor-privacy -Wnon-virtual-dtor -felide-constructors -fno-exceptions -fno-rtti -O3 -fno-omit-frame-pointer -mcpu=v8 -Wa,-xarch=v8plusa -g" LIBS="-lmtmalloc" ./configure --prefix=/usr/local/mysql --enable-assembler --with-extra-charsets=complex --enable-thread-safe-client +CFLAGS="-g -Wimplicit -Wreturn-type -Wswitch -Wtrigraphs -Wcomment -W -Wchar-subscripts -Wformat -Wparentheses -Wsign-compare -Wwrite-strings -Wunused -O3 -fno-omit-frame-pointer -mcpu=v8 -Wa,-xarch=v8plusa $EXTRA_FLAGS $EXTRA_CFLAGS" CXX=gcc CXXFLAGS="-Wimplicit -Wreturn-type -Wswitch -Wtrigraphs -Wcomment -W -Wchar-subscripts -Wformat -Wparentheses -Wsign-compare -Wwrite-strings -Woverloaded-virtual -Wsign-promo -Wreorder -Wctor-dtor-privacy -Wnon-virtual-dtor -felide-constructors -fno-exceptions -fno-rtti -O3 -fno-omit-frame-pointer -mcpu=v8 -Wa,-xarch=v8plusa -g $EXTRA_FLAGS $EXTRA_CXXFLAGS" LIBS="-lmtmalloc" ./configure --prefix=/usr/local/mysql --enable-assembler --with-extra-charsets=complex --enable-thread-safe-client make -j 4 === modified file 'CMakeLists.txt' --- a/CMakeLists.txt 2010-03-15 11:51:23 +0000 +++ b/CMakeLists.txt 2010-04-02 09:20:09 +0000 @@ -38,6 +38,7 @@ ADD_DEFINITIONS(-DMYSQL_DATADIR="c:/Prog ADD_DEFINITIONS(-DDEFAULT_CHARSET_HOME="c:/Program Files/MySQL/MySQL Server ${MYSQL_BASE_VERSION}/") ADD_DEFINITIONS(-DPACKAGE=mysql) ADD_DEFINITIONS(-DSHAREDIR="share") +ADD_DEFINITIONS(-DPLUGINDIR="lib/plugin") # Set debug options SET(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} -DFORCE_INIT_OF_VARS") @@ -260,7 +261,7 @@ FOREACH(SUBDIR ${STORAGE_SUBDIRS}) ENDIF(WITH_${ENGINE}_STORAGE_ENGINE AND MYSQL_PLUGIN_STATIC) IF (ENGINE_BUILD_TYPE STREQUAL "STATIC") - SET (mysql_plugin_defs "${mysql_plugin_defs},builtin_${PLUGIN_NAME}_plugin") + SET (maria_plugin_defs "${maria_plugin_defs},builtin_maria_${PLUGIN_NAME}_plugin") SET (MYSQLD_STATIC_ENGINE_LIBS ${MYSQLD_STATIC_ENGINE_LIBS} ${PLUGIN_NAME}) SET (STORAGE_ENGINE_DEFS "${STORAGE_ENGINE_DEFS} -DWITH_${ENGINE}_STORAGE_ENGINE") SET (WITH_${ENGINE}_STORAGE_ENGINE TRUE) @@ -279,7 +280,7 @@ ENDFOREACH(SUBDIR ${STORAGE_SUBDIRS}) # Special handling for partition(not really pluggable) IF(NOT WITHOUT_PARTITION_STORAGE_ENGINE) SET (STORAGE_ENGINE_DEFS "${STORAGE_ENGINE_DEFS} -DWITH_PARTITION_STORAGE_ENGINE") - SET (mysql_plugin_defs "${mysql_plugin_defs},builtin_partition_plugin") + SET (maria_plugin_defs "${maria_plugin_defs},builtin_maria_partition_plugin") ENDIF(NOT WITHOUT_PARTITION_STORAGE_ENGINE) # Special handling for tmp tables with the maria engine === modified file 'Docs/sp-imp-spec.txt' --- a/Docs/sp-imp-spec.txt 2004-03-23 11:04:40 +0000 +++ b/Docs/sp-imp-spec.txt 2010-04-08 12:10:05 +0000 @@ -1075,7 +1075,7 @@ 'PIPES_AS_CONCAT', 'ANSI_QUOTES', 'IGNORE_SPACE', - 'NOT_USED', + 'IGNORE_BAD_TABLE_OPTIONS', 'ONLY_FULL_GROUP_BY', 'NO_UNSIGNED_SUBTRACTION', 'NO_DIR_IN_CREATE', @@ -1097,4 +1097,4 @@ ) comment='Stored Procedures'; -- - \ No newline at end of file + === modified file 'client/mysql.cc' --- a/client/mysql.cc 2010-03-29 15:13:53 +0000 +++ b/client/mysql.cc 2010-04-08 13:02:49 +0000 @@ -4288,8 +4288,9 @@ extern "C" char *mysql_authentication_di } else { - fgets(buf, buf_len-1, stdin); - if (buf[0] && (s= strend(buf))[-1] == '\n') + if (!fgets(buf, buf_len-1, stdin)) + buf[0]= 0; + else if (buf[0] && (s= strend(buf))[-1] == '\n') s[-1]= 0; } === modified file 'cmd-line-utils/libedit/filecomplete.c' --- a/cmd-line-utils/libedit/filecomplete.c 2009-04-30 11:53:30 +0000 +++ b/cmd-line-utils/libedit/filecomplete.c 2010-03-31 19:12:21 +0000 @@ -95,10 +95,9 @@ static char break_chars[] = { ' ', '\t', char * fn_tilde_expand(const char *txt) { - struct passwd pwres, *pass; + struct passwd *pass; char *temp; size_t len = 0; - char pwbuf[1024]; if (txt[0] != '~') return (strdup(txt)); === modified file 'config/ac-macros/plugins.m4' --- a/config/ac-macros/plugins.m4 2010-03-29 15:14:28 +0000 +++ b/config/ac-macros/plugins.m4 2010-04-01 14:34:51 +0000 @@ -460,7 +460,7 @@ dnl Although this is "pretty", it breaks ]) ]) ]) - mysql_plugin_defs="$mysql_plugin_defs, [builtin_]$2[_plugin]" + maria_plugin_defs="$maria_plugin_defs, [builtin_maria_]$2[_plugin]" [with_plugin_]$2=yes AC_MSG_RESULT([yes]) m4_ifdef([$11], [ === modified file 'configure.in' --- a/configure.in 2010-03-29 15:13:53 +0000 +++ b/configure.in 2010-04-08 13:02:49 +0000 @@ -12,7 +12,7 @@ AC_PREREQ(2.59) # MySQL version number. # # Note: the following line must be parseable by win/configure.js:GetVersion() -AC_INIT([MariaDB Server], [5.2.0-MariaDB-alpha], [], [mysql]) +AC_INIT([MariaDB Server], [5.2.0-MariaDB-beta], [], [mysql]) AC_CONFIG_SRCDIR([sql/mysqld.cc]) AC_CANONICAL_SYSTEM # USTAR format gives us the possibility to store longer path names in @@ -2872,7 +2872,7 @@ AC_SUBST(server_scripts) AC_SUBST(mysql_plugin_dirs) AC_SUBST(mysql_plugin_libs) -AC_SUBST(mysql_plugin_defs) +AC_SUBST(maria_plugin_defs) # Now that sql_client_dirs and sql_server_dirs are stable, determine the union. === modified file 'extra/libevent/evbuffer.c' --- a/extra/libevent/evbuffer.c 2010-02-01 06:14:12 +0000 +++ b/extra/libevent/evbuffer.c 2010-04-02 09:20:09 +0000 @@ -75,8 +75,7 @@ bufferevent_add(struct event *ev, int ti */ void -bufferevent_read_pressure_cb(struct evbuffer *buf, - size_t old __attribute__((unused)), size_t now, +bufferevent_read_pressure_cb(struct evbuffer *buf, size_t old, size_t now, void *arg) { struct bufferevent *bufev = arg; /* === modified file 'extra/libevent/event.c' --- a/extra/libevent/event.c 2010-01-06 19:20:16 +0000 +++ b/extra/libevent/event.c 2010-04-02 09:20:09 +0000 @@ -394,8 +394,7 @@ event_base_get_method(struct event_base } static void -event_loopexit_cb(int fd __attribute__((unused)), - short what __attribute__((unused)), void *arg) +event_loopexit_cb(int fd, short what, void *arg) { struct event_base *base = arg; base->event_gotterm = 1; === modified file 'extra/libevent/signal.c' --- a/extra/libevent/signal.c 2010-01-06 19:20:16 +0000 +++ b/extra/libevent/signal.c 2010-04-02 09:20:09 +0000 @@ -69,7 +69,7 @@ static void evsignal_handler(int sig); /* Callback for when the signal handler write a byte to our signaling socket */ static void -evsignal_cb(int fd, short what __attribute((unused)), void *arg __attribute((unused))) +evsignal_cb(int fd, short what, void *arg) { static char signals[100]; #ifdef WIN32 === modified file 'include/keycache.h' --- a/include/keycache.h 2009-12-03 11:19:05 +0000 +++ b/include/keycache.h 2010-04-06 02:08:10 +0000 @@ -19,96 +19,156 @@ #define _keycache_h C_MODE_START -/* declare structures that is used by st_key_cache */ -struct st_block_link; -typedef struct st_block_link BLOCK_LINK; -struct st_keycache_page; -typedef struct st_keycache_page KEYCACHE_PAGE; -struct st_hash_link; -typedef struct st_hash_link HASH_LINK; -/* info about requests in a waiting queue */ -typedef struct st_keycache_wqueue +/* + Currently the default key cache is created as non-partitioned at + the start of the server unless the server is started with the parameter + --key-cache-partitions that is greater than 0 +*/ + +#define DEFAULT_KEY_CACHE_PARTITIONS 0 + +/* + MAX_KEY_CACHE_PARTITIONS cannot be greater than + sizeof(MYISAM_SHARE::dirty_part_map) + Currently sizeof(MYISAM_SHARE::dirty_part_map)=sizeof(ulonglong) +*/ + +#define MAX_KEY_CACHE_PARTITIONS 64 + +/* The structure to get statistical data about a key cache */ + +typedef struct st_key_cache_statistics { - struct st_my_thread_var *last_thread; /* circular list of waiting threads */ -} KEYCACHE_WQUEUE; + ulonglong mem_size; /* memory for cache buffers/auxiliary structures */ + ulonglong block_size; /* size of the each buffers in the key cache */ + ulonglong blocks_used; /* maximum number of used blocks/buffers */ + ulonglong blocks_unused; /* number of currently unused blocks */ + ulonglong blocks_changed; /* number of currently dirty blocks */ + ulonglong read_requests; /* number of read requests (read hits) */ + ulonglong reads; /* number of actual reads from files into buffers */ + ulonglong write_requests; /* number of write requests (write hits) */ + ulonglong writes; /* number of actual writes from buffers into files */ +} KEY_CACHE_STATISTICS; + +#define NUM_LONG_KEY_CACHE_STAT_VARIABLES 3 -#define CHANGED_BLOCKS_HASH 128 /* must be power of 2 */ +/* The type of a key cache object */ +typedef enum key_cache_type +{ + SIMPLE_KEY_CACHE, + PARTITIONED_KEY_CACHE +} KEY_CACHE_TYPE; + + +typedef + int (*INIT_KEY_CACHE) + (void *, uint key_cache_block_size, + size_t use_mem, uint division_limit, uint age_threshold); +typedef + int (*RESIZE_KEY_CACHE) + (void *, uint key_cache_block_size, + size_t use_mem, uint division_limit, uint age_threshold); +typedef + void (*CHANGE_KEY_CACHE_PARAM) + (void *keycache_cb, + uint division_limit, uint age_threshold); +typedef + uchar* (*KEY_CACHE_READ) + (void *keycache_cb, + File file, my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length, int return_buffer); +typedef + int (*KEY_CACHE_INSERT) + (void *keycache_cb, + File file, my_off_t filepos, int level, + uchar *buff, uint length); +typedef + int (*KEY_CACHE_WRITE) + (void *keycache_cb, + File file, void *file_extra, + my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length, int force_write); +typedef + int (*FLUSH_KEY_BLOCKS) + (void *keycache_cb, + int file, void *file_extra, + enum flush_type type); +typedef + int (*RESET_KEY_CACHE_COUNTERS) + (const char *name, void *keycache_cb); +typedef + void (*END_KEY_CACHE) + (void *keycache_cb, my_bool cleanup); +typedef + void (*GET_KEY_CACHE_STATISTICS) + (void *keycache_cb, uint partition_no, + KEY_CACHE_STATISTICS *key_cache_stats); +typedef + ulonglong (*GET_KEY_CACHE_STAT_VALUE) + (void *keycache_cb, uint var_no); /* - The key cache structure - It also contains read-only statistics parameters. + An object of the type KEY_CACHE_FUNCS contains pointers to all functions + from the key cache interface. + Currently a key cache can be of two types: simple and partitioned. + For each of them its own static structure of the type KEY_CACHE_FUNCS is + defined . The structures contain the pointers to the implementations of + the interface functions used by simple key caches and partitioned key + caches respectively. Pointers to these structures are assigned to key cache + objects at the time of their creation. */ +typedef struct st_key_cache_funcs +{ + INIT_KEY_CACHE init; + RESIZE_KEY_CACHE resize; + CHANGE_KEY_CACHE_PARAM change_param; + KEY_CACHE_READ read; + KEY_CACHE_INSERT insert; + KEY_CACHE_WRITE write; + FLUSH_KEY_BLOCKS flush; + RESET_KEY_CACHE_COUNTERS reset_counters; + END_KEY_CACHE end; + GET_KEY_CACHE_STATISTICS get_stats; + GET_KEY_CACHE_STAT_VALUE get_stat_val; +} KEY_CACHE_FUNCS; + + typedef struct st_key_cache { - my_bool key_cache_inited; - my_bool in_resize; /* true during resize operation */ - my_bool resize_in_flush; /* true during flush of resize operation */ + KEY_CACHE_TYPE key_cache_type; /* type of the key cache used for debugging */ + void *keycache_cb; /* control block of the used key cache */ + KEY_CACHE_FUNCS *interface_funcs; /* interface functions of the key cache */ + ulonglong param_buff_size; /* size the memory allocated for the cache */ + ulong param_block_size; /* size of the blocks in the key cache */ + ulong param_division_limit; /* min. percentage of warm blocks */ + ulong param_age_threshold; /* determines when hot block is downgraded */ + ulong param_partitions; /* number of the key cache partitions */ + my_bool key_cache_inited; /* <=> key cache has been created */ my_bool can_be_used; /* usage of cache for read/write is allowed */ - size_t key_cache_mem_size; /* specified size of the cache memory */ - uint key_cache_block_size; /* size of the page buffer of a cache block */ - ulong min_warm_blocks; /* min number of warm blocks; */ - ulong age_threshold; /* age threshold for hot blocks */ - ulonglong keycache_time; /* total number of block link operations */ - uint hash_entries; /* max number of entries in the hash table */ - int hash_links; /* max number of hash links */ - int hash_links_used; /* number of hash links currently used */ - int disk_blocks; /* max number of blocks in the cache */ - ulong blocks_used; /* maximum number of concurrently used blocks */ - ulong blocks_unused; /* number of currently unused blocks */ - ulong blocks_changed; /* number of currently dirty blocks */ - ulong warm_blocks; /* number of blocks in warm sub-chain */ - ulong cnt_for_resize_op; /* counter to block resize operation */ - long blocks_available; /* number of blocks available in the LRU chain */ - HASH_LINK **hash_root; /* arr. of entries into hash table buckets */ - HASH_LINK *hash_link_root; /* memory for hash table links */ - HASH_LINK *free_hash_list; /* list of free hash links */ - BLOCK_LINK *free_block_list; /* list of free blocks */ - BLOCK_LINK *block_root; /* memory for block links */ - uchar HUGE_PTR *block_mem; /* memory for block buffers */ - BLOCK_LINK *used_last; /* ptr to the last block of the LRU chain */ - BLOCK_LINK *used_ins; /* ptr to the insertion block in LRU chain */ - pthread_mutex_t cache_lock; /* to lock access to the cache structure */ - KEYCACHE_WQUEUE resize_queue; /* threads waiting during resize operation */ - /* - Waiting for a zero resize count. Using a queue for symmetry though - only one thread can wait here. - */ - KEYCACHE_WQUEUE waiting_for_resize_cnt; - KEYCACHE_WQUEUE waiting_for_hash_link; /* waiting for a free hash link */ - KEYCACHE_WQUEUE waiting_for_block; /* requests waiting for a free block */ - BLOCK_LINK *changed_blocks[CHANGED_BLOCKS_HASH]; /* hash for dirty file bl.*/ - BLOCK_LINK *file_blocks[CHANGED_BLOCKS_HASH]; /* hash for other file bl.*/ - - /* - The following variables are and variables used to hold parameters for - initializing the key cache. - */ - - ulonglong param_buff_size; /* size the memory allocated for the cache */ - ulong param_block_size; /* size of the blocks in the key cache */ - ulong param_division_limit; /* min. percentage of warm blocks */ - ulong param_age_threshold; /* determines when hot block is downgraded */ - - /* Statistics variables. These are reset in reset_key_cache_counters(). */ - ulong global_blocks_changed; /* number of currently dirty blocks */ + my_bool in_init; /* Set to 1 in MySQL during init/resize */ + uint partitions; /* actual number of partitions */ + size_t key_cache_mem_size; /* specified size of the cache memory */ + ulong blocks_used; /* maximum number of concurrently used blocks */ + ulong blocks_unused; /* number of currently unused blocks */ + ulong global_blocks_changed; /* number of currently dirty blocks */ ulonglong global_cache_w_requests;/* number of write requests (write hits) */ ulonglong global_cache_write; /* number of writes from cache to files */ ulonglong global_cache_r_requests;/* number of read requests (read hits) */ ulonglong global_cache_read; /* number of reads from files to cache */ - - int blocks; /* max number of blocks in the cache */ - my_bool in_init; /* Set to 1 in MySQL during init/resize */ } KEY_CACHE; + /* The default key cache */ extern KEY_CACHE dflt_key_cache_var, *dflt_key_cache; extern int init_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, size_t use_mem, uint division_limit, - uint age_threshold); + uint age_threshold, uint partitions); extern int resize_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, size_t use_mem, uint division_limit, uint age_threshold); @@ -122,12 +182,18 @@ extern int key_cache_insert(KEY_CACHE *k File file, my_off_t filepos, int level, uchar *buff, uint length); extern int key_cache_write(KEY_CACHE *keycache, - File file, my_off_t filepos, int level, + File file, void *file_extra, + my_off_t filepos, int level, uchar *buff, uint length, - uint block_length,int force_write); + uint block_length, int force_write); extern int flush_key_blocks(KEY_CACHE *keycache, - int file, enum flush_type type); + int file, void *file_extra, + enum flush_type type); extern void end_key_cache(KEY_CACHE *keycache, my_bool cleanup); +extern void get_key_cache_statistics(KEY_CACHE *keycache, + uint partition_no, + KEY_CACHE_STATISTICS *key_cache_stats); +extern ulonglong get_key_cache_stat_value(KEY_CACHE *keycache, uint var_no); /* Functions to handle multiple key caches */ extern my_bool multi_keycache_init(void); @@ -140,5 +206,11 @@ extern void multi_key_cache_change(KEY_C KEY_CACHE *new_data); extern int reset_key_cache_counters(const char *name, KEY_CACHE *key_cache); +extern int repartition_key_cache(KEY_CACHE *keycache, + uint key_cache_block_size, + size_t use_mem, + uint division_limit, + uint age_threshold, + uint partitions); C_MODE_END #endif /* _keycache_h */ === modified file 'include/m_ctype.h' --- a/include/m_ctype.h 2010-01-06 19:20:16 +0000 +++ b/include/m_ctype.h 2010-04-02 09:20:09 +0000 @@ -291,7 +291,7 @@ struct charset_info_st const uint16 *const *sort_order_big; const uint16 *tab_to_uni; MY_UNI_IDX *tab_from_uni; - const MY_UNICASE_INFO *const *caseinfo; + MY_UNICASE_INFO *const *caseinfo; const uchar *state_map; const uchar *ident_map; uint strxfrm_multiply; @@ -311,7 +311,9 @@ struct charset_info_st #define ILLEGAL_CHARSET_INFO_NUMBER (~0U) extern MYSQL_PLUGIN_IMPORT struct charset_info_st my_charset_bin; -extern struct charset_info_st my_charset_bin; +extern MYSQL_PLUGIN_IMPORT struct charset_info_st my_charset_latin1; +extern MYSQL_PLUGIN_IMPORT struct charset_info_st my_charset_filename; + extern struct charset_info_st my_charset_big5_chinese_ci; extern struct charset_info_st my_charset_big5_bin; extern struct charset_info_st my_charset_cp932_japanese_ci; @@ -324,7 +326,6 @@ extern struct charset_info_st my_charset extern struct charset_info_st my_charset_gb2312_bin; extern struct charset_info_st my_charset_gbk_chinese_ci; extern struct charset_info_st my_charset_gbk_bin; -extern struct charset_info_st my_charset_latin1; extern struct charset_info_st my_charset_latin1_german2_ci; extern struct charset_info_st my_charset_latin1_bin; extern struct charset_info_st my_charset_latin2_czech_ci; @@ -341,7 +342,6 @@ extern struct charset_info_st my_charset extern struct charset_info_st my_charset_utf8_unicode_ci; extern struct charset_info_st my_charset_utf8_bin; extern struct charset_info_st my_charset_cp1250_czech_ci; -extern struct charset_info_st my_charset_filename; /* declarations for simple charsets */ extern size_t my_strnxfrm_simple(CHARSET_INFO *, uchar *, size_t, === modified file 'include/my_base.h' --- a/include/my_base.h 2010-02-10 19:06:24 +0000 +++ b/include/my_base.h 2010-04-08 12:10:05 +0000 @@ -314,6 +314,8 @@ enum ha_base_keytype { #define HA_OPTION_RELIES_ON_SQL_LAYER 512 #define HA_OPTION_NULL_FIELDS 1024 #define HA_OPTION_PAGE_CHECKSUM 2048 +/* .frm has extra create options in linked-list format */ +#define HA_OPTION_TEXT_CREATE_OPTIONS (1L << 14) #define HA_OPTION_TEMP_COMPRESS_RECORD (1L << 15) /* set by isamchk */ #define HA_OPTION_READ_ONLY_DATA (1L << 16) /* Set by isamchk */ #define HA_OPTION_NO_CHECKSUM (1L << 17) === modified file 'include/my_global.h' --- a/include/my_global.h 2010-03-29 15:13:53 +0000 +++ b/include/my_global.h 2010-04-02 09:20:09 +0000 @@ -1522,6 +1522,9 @@ do { doubleget_union _tmp; \ #elif defined(HAVE_DLFCN_H) #include <dlfcn.h> #endif +#ifndef HAVE_DLERROR +#define dlerror() "" +#endif #endif /* FreeBSD 2.2.2 does not define RTLD_NOW) */ @@ -1529,7 +1532,7 @@ do { doubleget_union _tmp; \ #define RTLD_NOW 1 #endif -#ifndef HAVE_DLERROR +#ifndef HAVE_DLOPEN #define dlerror() "No support for dynamic loading (static build?)" #define dlopen(A,B) 0 #define dlsym(A,B) 0 === modified file 'include/my_sys.h' --- a/include/my_sys.h 2010-03-29 15:13:53 +0000 +++ b/include/my_sys.h 2010-04-08 12:10:05 +0000 @@ -208,10 +208,14 @@ extern void my_large_free(uchar * ptr, m #define alloca __builtin_alloca #endif /* GNUC */ #define my_alloca(SZ) alloca((size_t) (SZ)) -#define my_afree(PTR) {} +#define my_afree(PTR) ((void)0) +#define my_safe_alloca(size, min_length) ((size <= min_length) ? my_alloca(size) : my_malloc(size,MYF(MY_FAE))) +#define my_safe_afree(ptr, size, min_length) ((size <= min_length) ? (void)0 : my_free(ptr,MYF(MY_WME))) #else #define my_alloca(SZ) my_malloc(SZ,MYF(MY_FAE)) #define my_afree(PTR) my_free(PTR,MYF(MY_WME)) +#define my_safe_alloca(size, min_length) my_alloca(size) +#define my_safe_afree(ptr, size, min_length) my_afree(ptr) #endif /* HAVE_ALLOCA */ #ifndef errno /* did we already get it? */ === modified file 'include/mysql/plugin.h' --- a/include/mysql/plugin.h 2010-03-29 15:13:53 +0000 +++ b/include/mysql/plugin.h 2010-04-01 14:34:51 +0000 @@ -57,7 +57,10 @@ typedef struct st_mysql_xid MYSQL_XID; Plugin API. Common for all plugin types. */ +/* MySQL plugin interface version */ #define MYSQL_PLUGIN_INTERFACE_VERSION 0x0101 +/* MariaDB plugin interface version */ +#define MARIA_PLUGIN_INTERFACE_VERSION 0x0100 /* The allowable types of plugins @@ -81,6 +84,14 @@ typedef struct st_mysql_xid MYSQL_XID; #define PLUGIN_LICENSE_GPL_STRING "GPL" #define PLUGIN_LICENSE_BSD_STRING "BSD" +/* definitions of code maturity for plugins */ +#define MariaDB_PLUGIN_MATURITY_UNKNOWN 0 +#define MariaDB_PLUGIN_MATURITY_EXPERIMENTAL 1 +#define MariaDB_PLUGIN_MATURITY_ALPHA 2 +#define MariaDB_PLUGIN_MATURITY_BETA 3 +#define MariaDB_PLUGIN_MATURITY_GAMMA 4 +#define MariaDB_PLUGIN_MATURITY_STABLE 5 + /* Macros for beginning and ending plugin declarations. Between mysql_declare_plugin and mysql_declare_plugin_end there should @@ -93,11 +104,24 @@ typedef struct st_mysql_xid MYSQL_XID; int VERSION= MYSQL_PLUGIN_INTERFACE_VERSION; \ int PSIZE= sizeof(struct st_mysql_plugin); \ struct st_mysql_plugin DECLS[]= { + +#define MARIA_DECLARE_PLUGIN__(NAME, VERSION, PSIZE, DECLS) \ +int VERSION= MARIA_PLUGIN_INTERFACE_VERSION; \ +int PSIZE= sizeof(struct st_maria_plugin); \ +struct st_maria_plugin DECLS[]= { + #else + #define __MYSQL_DECLARE_PLUGIN(NAME, VERSION, PSIZE, DECLS) \ MYSQL_PLUGIN_EXPORT int _mysql_plugin_interface_version_= MYSQL_PLUGIN_INTERFACE_VERSION; \ MYSQL_PLUGIN_EXPORT int _mysql_sizeof_struct_st_plugin_= sizeof(struct st_mysql_plugin); \ MYSQL_PLUGIN_EXPORT struct st_mysql_plugin _mysql_plugin_declarations_[]= { + +#define MARIA_DECLARE_PLUGIN__(NAME, VERSION, PSIZE, DECLS) \ +MYSQL_PLUGIN_EXPORT int _maria_plugin_interface_version_= MARIA_PLUGIN_INTERFACE_VERSION; \ +MYSQL_PLUGIN_EXPORT int _maria_sizeof_struct_st_plugin_= sizeof(struct st_maria_plugin); \ +MYSQL_PLUGIN_EXPORT struct st_maria_plugin _maria_plugin_declarations_[]= { + #endif #define mysql_declare_plugin(NAME) \ @@ -106,7 +130,14 @@ __MYSQL_DECLARE_PLUGIN(NAME, \ builtin_ ## NAME ## _sizeof_struct_st_plugin, \ builtin_ ## NAME ## _plugin) +#define maria_declare_plugin(NAME) \ +MARIA_DECLARE_PLUGIN__(NAME, \ + builtin_maria_ ## NAME ## _plugin_interface_version, \ + builtin_maria_ ## NAME ## _sizeof_struct_st_plugin, \ + builtin_maria_ ## NAME ## _plugin) + #define mysql_declare_plugin_end ,{0,0,0,0,0,0,0,0,0,0,0,0}} +#define maria_declare_plugin_end ,{0,0,0,0,0,0,0,0,0,0,0,0,0}} /* declarations for SHOW STATUS support in plugins @@ -403,6 +434,30 @@ struct st_mysql_plugin void * __reserved1; /* reserved for dependency checking */ }; +/* + MariaDB extension for plugins declaration structure. + + It also copy current MySQL plugin fields to have more independency + in plugins extension +*/ + +struct st_maria_plugin +{ + int type; /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + void *info; /* pointer to type-specific plugin descriptor */ + const char *name; /* plugin name */ + const char *author; /* plugin author (for SHOW PLUGINS) */ + const char *descr; /* general descriptive text (for SHOW PLUGINS ) */ + int license; /* the plugin license (PLUGIN_LICENSE_XXX) */ + int (*init)(void *); /* the function to invoke when plugin is loaded */ + int (*deinit)(void *);/* the function to invoke when plugin is unloaded */ + unsigned int version; /* plugin version (for SHOW PLUGINS) */ + struct st_mysql_show_var *status_vars; + struct st_mysql_sys_var **system_vars; + const char *version_info; /* plugin version string */ + int maturity; /* MariaDB_PLUGIN_MATURITY_XXX */ +}; + /************************************************************************* API for Full-text parser plugin. (MYSQL_FTPARSER_PLUGIN) */ === modified file 'include/mysql/plugin_auth.h.pp' --- a/include/mysql/plugin_auth.h.pp 2010-03-29 15:13:53 +0000 +++ b/include/mysql/plugin_auth.h.pp 2010-04-01 14:34:51 +0000 @@ -77,6 +77,22 @@ struct st_mysql_plugin struct st_mysql_sys_var **system_vars; void * __reserved1; }; +struct st_maria_plugin +{ + int type; + void *info; + const char *name; + const char *author; + const char *descr; + int license; + int (*init)(void *); + int (*deinit)(void *); + unsigned int version; + struct st_mysql_show_var *status_vars; + struct st_mysql_sys_var **system_vars; + const char *version_info; + int maturity; +}; enum enum_ftparser_mode { MYSQL_FTPARSER_SIMPLE_MODE= 0, === modified file 'include/sql_common.h' --- a/include/sql_common.h 2010-03-29 15:13:53 +0000 +++ b/include/sql_common.h 2010-04-01 09:04:26 +0000 @@ -96,7 +96,7 @@ void set_mysql_extended_error(MYSQL *mys struct st_plugin_vio_info; void mpvio_info(Vio *vio, struct st_plugin_vio_info *info); int run_plugin_auth(MYSQL *mysql, char *data, uint data_len, - char *data_plugin, const char *db); + const char *data_plugin, const char *db); int mysql_client_plugin_init(); void mysql_client_plugin_deinit(); struct st_mysql_client_plugin; === modified file 'libmysqld/CMakeLists.txt' --- a/libmysqld/CMakeLists.txt 2010-01-31 09:13:21 +0000 +++ b/libmysqld/CMakeLists.txt 2010-04-08 12:10:05 +0000 @@ -104,7 +104,7 @@ SET(LIBMYSQLD_SOURCES emb_qcache.cc libm ../sql-common/client.c ../sql-common/my_time.c ../sql-common/my_user.c ../sql-common/pack.c ../sql/password.c ../sql/discover.cc ../sql/derror.cc - ../sql/field.cc ../sql/field_conv.cc + ../sql/field.cc ../sql/field_conv.cc ../sql-common/client_plugin.c ../sql/filesort.cc ../sql/gstream.cc ../sql/ha_partition.cc ../sql/handler.cc ../sql/hash_filo.cc ../sql/hostname.cc ../sql/init.cc ../sql/item_buff.cc ../sql/item_cmpfunc.cc @@ -139,7 +139,8 @@ SET(LIBMYSQLD_SOURCES emb_qcache.cc libm ../sql/strfunc.cc ../sql/table.cc ../sql/thr_malloc.cc ../sql/time.cc ../sql/tztime.cc ../sql/uniques.cc ../sql/unireg.cc ../sql/partition_info.cc ../sql/sql_connect.cc - ../sql/scheduler.cc ../sql/event_parse_data.cc + ../sql/scheduler.cc ../sql/event_parse_data.cc + ../sql/create_options.cc ${GEN_SOURCES} ${LIB_SOURCES}) === modified file 'libmysqld/Makefile.am' --- a/libmysqld/Makefile.am 2010-03-29 15:13:53 +0000 +++ b/libmysqld/Makefile.am 2010-04-08 12:10:05 +0000 @@ -75,7 +75,7 @@ sqlsources = derror.cc field.cc field_co parse_file.cc sql_view.cc sql_trigger.cc my_decimal.cc \ rpl_filter.cc sql_partition.cc sql_builtin.cc sql_plugin.cc \ debug_sync.cc \ - sql_tablespace.cc \ + sql_tablespace.cc create_options.cc \ rpl_injector.cc my_user.c partition_info.cc \ sql_servers.cc event_parse_data.cc opt_table_elimination.cc === modified file 'mysql-test/extra/rpl_tests/rpl_insert_id_pk.test' --- a/mysql-test/extra/rpl_tests/rpl_insert_id_pk.test 2007-06-06 17:57:07 +0000 +++ b/mysql-test/extra/rpl_tests/rpl_insert_id_pk.test 2010-04-06 22:47:08 +0000 @@ -82,7 +82,6 @@ SET FOREIGN_KEY_CHECKS=0; # Duplicate Key Errors codes --error 1022, ER_DUP_ENTRY INSERT INTO t1 VALUES (1),(1); -sync_slave_with_master; -connection master; drop table t1; +sync_slave_with_master; # End of 4.1 tests === modified file 'mysql-test/include/maria_empty_logs.inc' --- a/mysql-test/include/maria_empty_logs.inc 2009-10-13 12:47:17 +0000 +++ b/mysql-test/include/maria_empty_logs.inc 2010-04-06 22:47:08 +0000 @@ -11,9 +11,10 @@ connection default; let $default_db=`select database()`; let $MYSQLD_DATADIR= `SELECT @@datadir`; -# it will used at end of test for wait_for_status_var.inc primitive -let $status_var= Threads_connected; -let $status_var_value= query_get_value(SHOW STATUS LIKE 'Threads_connected', Value, 1); + +#it will used at end of test for wait_for_status_var.inc primitive +#let $status_var= Threads_connected; +#let $status_var_value= query_get_value(SHOW STATUS LIKE 'Threads_connected', Value, 1); connection admin; -- echo * shut down mysqld, removed logs, restarted it @@ -77,12 +78,15 @@ append_file $MYSQLTEST_VARDIR/tmp/mysqld restart-maria_empty_logs.inc EOF +connection default; --source include/wait_until_connected_again.inc -connection default; # Make sure that all connections are restored ---source include/wait_for_status_var.inc +# This is disabled as 'Threads_connected' can't be trusted' +# (It may be affected by 'check_testcase()') +# --source include/wait_for_status_var.inc # Restore current database as the effect of "use" was lost after restart + --disable_query_log eval use $default_db; --enable_query_log === modified file 'mysql-test/lib/My/SafeProcess/safe_process.cc' --- a/mysql-test/lib/My/SafeProcess/safe_process.cc 2010-03-28 18:10:00 +0000 +++ b/mysql-test/lib/My/SafeProcess/safe_process.cc 2010-04-08 11:56:57 +0000 @@ -85,17 +85,18 @@ static void die(const char* fmt, ...) va_end(args); if (int last_err= errno) fprintf(stderr, "error: %d, %s\n", last_err, strerror(last_err)); - exit(1); + exit(6); } -static void kill_child(void) +static void kill_child(bool was_killed) { int status= 0; message("Killing child: %d", child_pid); // Terminate whole process group - kill(-child_pid, SIGKILL); + if (! was_killed) + kill(-child_pid, SIGKILL); pid_t ret_pid= waitpid(child_pid, &status, 0); if (ret_pid == child_pid) @@ -115,7 +116,7 @@ static void kill_child(void) exit(exit_code); } - exit(1); + exit(5); } @@ -135,7 +136,7 @@ extern "C" void handle_signal(int sig) terminated= 1; if (child_pid > 0) - kill_child(); + kill_child(sig == SIGCHLD); // Ignore further signals signal(SIGTERM, SIG_IGN); @@ -240,8 +241,8 @@ int main(int argc, char* const argv[] ) // Close write end close(pfd[1]); - if (execvp(child_argv[0], child_argv) < 0) - die("Failed to exec child"); + execvp(child_argv[0], child_argv); + die("Failed to exec child"); } close(pfd[1]); // Close unused write end @@ -257,39 +258,19 @@ int main(int argc, char* const argv[] ) /* Monitor loop */ message("Started child %d, terminated: %d", child_pid, terminated); - while(!terminated) + while (!terminated) { // Check if parent is still alive - if (kill(parent_pid, 0) != 0){ + if (kill(parent_pid, 0) != 0) + { message("Parent is not alive anymore"); break; } - - // Check if child has exited, normally this will be - // detected immediately with SIGCHLD handler - int status= 0; - pid_t ret_pid= waitpid(child_pid, &status, WNOHANG); - if (ret_pid == child_pid) - { - int ret_code= 2; - if (WIFEXITED(status)) - { - // Process has exited, collect return status - ret_code= WEXITSTATUS(status); - message("Child exit: %d", ret_code); - // Exit with exit status of the child - exit(ret_code); - } - - if (WIFSIGNALED(status)) - message("Child killed by signal: %d", WTERMSIG(status)); - - exit(ret_code); - } + /* Wait for parent or child to die */ sleep(1); } - kill_child(); + kill_child(0); - return 1; + return 4; } === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2010-03-31 08:17:03 +0000 +++ b/mysql-test/mysql-test-run.pl 2010-04-09 09:26:17 +0000 @@ -4039,7 +4039,10 @@ sub extract_warning_lines ($) { qr/Slave I\/O: Get master COLLATION_SERVER failed with error:.*/, qr/Slave I\/O: Get master TIME_ZONE failed with error:.*/, qr/Slave I\/O: error reconnecting to master '.*' - retry-time: [1-3] retries/, + qr/Slave I\/0: Master command COM_BINLOG_DUMP failed/, qr/Error reading packet/, + qr/Lost connection to MySQL server at 'reading initial communication packet'/, + qr/Failed on request_dump/, qr/Slave: Can't drop database.* database doesn't exist/, qr/Slave: Operation DROP USER failed for 'create_rout_db'/, qr|Checking table: '\./mtr/test_suppressions'|, === modified file 'mysql-test/r/connect.result' --- a/mysql-test/r/connect.result 2009-04-25 09:04:38 +0000 +++ b/mysql-test/r/connect.result 2010-04-06 18:36:06 +0000 @@ -225,3 +225,17 @@ Connection on extra port 2 ok # ------------------------------------------------------------------ # -- End of 5.1 tests # ------------------------------------------------------------------ +CREATE USER mysqltest_up1 IDENTIFIED VIA mysql_native_password using '*E8D46CE25265E545D225A8A6F1BAF642FEBEE5CB'; +CREATE USER mysqltest_up2 IDENTIFIED VIA mysql_old_password using '09301740536db389'; +connect(localhost,mysqltest_up1,foo,test,MASTER_PORT,MASTER_SOCKET); +ERROR 28000: Access denied for user 'mysqltest_up1'@'localhost' (using password: YES) +select user(), current_user(); +user() current_user() +mysqltest_up1@localhost mysqltest_up1@% +connect(localhost,mysqltest_up2,newpw,test,MASTER_PORT,MASTER_SOCKET); +ERROR 28000: Access denied for user 'mysqltest_up2'@'localhost' (using password: YES) +select user(), current_user(); +user() current_user() +mysqltest_up2@localhost mysqltest_up2@% +DROP USER mysqltest_up1@'%'; +DROP USER mysqltest_up2@'%'; === added file 'mysql-test/r/create_options.result' --- a/mysql-test/r/create_options.result 1970-01-01 00:00:00 +0000 +++ b/mysql-test/r/create_options.result 2010-04-08 12:10:05 +0000 @@ -0,0 +1,182 @@ +drop table if exists t1; +SET @OLD_SQL_MODE=@@SQL_MODE; +SET SQL_MODE='IGNORE_BAD_TABLE_OPTIONS'; +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1='1v1'; +Warnings: +Warning 1650 Unknown option 'fkey' +Warning 1650 Unknown option 'dff' +Warning 1650 Unknown option 'tkey1' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey`=vvv, + KEY `akey` (`a`) `dff`=vvv +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey1`='1v1' +drop table t1; +#reassiginig options in the same line +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1=1v1 TKEY1=DEFAULT tkey1=1v2 tkey2=2v1; +Warnings: +Warning 1650 Unknown option 'fkey' +Warning 1650 Unknown option 'dff' +Warning 1650 Unknown option 'tkey1' +Warning 1650 Unknown option 'tkey2' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey`=vvv, + KEY `akey` (`a`) `dff`=vvv +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey1`=1v2 `tkey2`=2v1 +#add option +alter table t1 tkey4=4v1; +Warnings: +Warning 1650 Unknown option 'tkey4' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey`=vvv, + KEY `akey` (`a`) `dff`=vvv +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey1`=1v2 `tkey2`=2v1 `tkey4`=4v1 +#remove options +alter table t1 tkey3=DEFAULT tkey4=DEFAULT; +Warnings: +Warning 1650 Unknown option 'tkey3' +Warning 1650 Unknown option 'tkey4' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey`=vvv, + KEY `akey` (`a`) `dff`=vvv +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey1`=1v2 `tkey2`=2v1 +drop table t1; +create table t1 (a int fkey1=v1, key akey (a) kkey1=v1) tkey1=1v1 tkey1=1v2 TKEY1=DEFAULT tkey2=2v1 tkey3=3v1; +Warnings: +Warning 1650 Unknown option 'fkey1' +Warning 1650 Unknown option 'kkey1' +Warning 1650 Unknown option 'TKEY1' +Warning 1650 Unknown option 'tkey2' +Warning 1650 Unknown option 'tkey3' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v1, + KEY `akey` (`a`) `kkey1`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#change field with option with the same value +alter table t1 change a a int `FKEY1`='v1'; +Warnings: +Warning 1650 Unknown option 'FKEY1' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `FKEY1`='v1', + KEY `akey` (`a`) `kkey1`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#change field with option with a different value +alter table t1 change a a int fkey1=v2; +Warnings: +Warning 1650 Unknown option 'fkey1' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + KEY `akey` (`a`) `kkey1`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#new column no options +alter table t1 add column b int; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `b` int(11) DEFAULT NULL, + KEY `akey` (`a`) `kkey1`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#new key with options +alter table t1 add key bkey (b) kkey2=v1; +Warnings: +Warning 1650 Unknown option 'kkey2' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `b` int(11) DEFAULT NULL, + KEY `akey` (`a`) `kkey1`=v1, + KEY `bkey` (`b`) `kkey2`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#new column with options +alter table t1 add column c int fkey1=v1 fkey2=v2; +Warnings: +Warning 1650 Unknown option 'fkey1' +Warning 1650 Unknown option 'fkey2' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `b` int(11) DEFAULT NULL, + `c` int(11) DEFAULT NULL `fkey1`=v1 `fkey2`=v2, + KEY `akey` (`a`) `kkey1`=v1, + KEY `bkey` (`b`) `kkey2`=v1 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#new key no options +alter table t1 add key ckey (c); +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `b` int(11) DEFAULT NULL, + `c` int(11) DEFAULT NULL `fkey1`=v1 `fkey2`=v2, + KEY `akey` (`a`) `kkey1`=v1, + KEY `bkey` (`b`) `kkey2`=v1, + KEY `ckey` (`c`) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#drop column +alter table t1 drop b; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `c` int(11) DEFAULT NULL `fkey1`=v1 `fkey2`=v2, + KEY `akey` (`a`) `kkey1`=v1, + KEY `ckey` (`c`) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#add column with options after delete +alter table t1 add column b int fkey2=v1; +Warnings: +Warning 1650 Unknown option 'fkey2' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `c` int(11) DEFAULT NULL `fkey1`=v1 `fkey2`=v2, + `b` int(11) DEFAULT NULL `fkey2`=v1, + KEY `akey` (`a`) `kkey1`=v1, + KEY `ckey` (`c`) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +#add key +alter table t1 add key bkey (b) kkey2=v2; +Warnings: +Warning 1650 Unknown option 'kkey2' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `fkey1`=v2, + `c` int(11) DEFAULT NULL `fkey1`=v1 `fkey2`=v2, + `b` int(11) DEFAULT NULL `fkey2`=v1, + KEY `akey` (`a`) `kkey1`=v1, + KEY `ckey` (`c`), + KEY `bkey` (`b`) `kkey2`=v2 +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey2`=2v1 `tkey3`=3v1 +drop table t1; +create table t1 (a int) tkey1=100; +Warnings: +Warning 1650 Unknown option 'tkey1' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL +) ENGINE=MyISAM DEFAULT CHARSET=latin1 `tkey1`=100 +drop table t1; +#error on unknown option +SET SQL_MODE=''; +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1=1v1; +ERROR HY000: Unknown option 'fkey' +SET @@SQL_MODE=@OLD_SQL_MODE; === modified file 'mysql-test/r/events_bugs.result' --- a/mysql-test/r/events_bugs.result 2009-03-11 20:30:56 +0000 +++ b/mysql-test/r/events_bugs.result 2010-04-08 12:10:05 +0000 @@ -729,9 +729,8 @@ set @@sql_mode= pow(2,32)-1; create event e1 on schedule every 1 day do select 1; select @@sql_mode; @@sql_mode -REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,?,ONLY_FULL_GROUP_BY,NO_UNSIGNED_SUBTRACTION,NO_DIR_IN_CREATE,POSTGRESQL,ORACLE,MSSQL,DB2,MAXDB,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,MYSQL323,MYSQL40,ANSI,NO_AUTO_VALUE_ON_ZERO,NO_BACKSLASH_ESCAPES,STRICT_TRANS_TABLES,STRICT_ALL_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ALLOW_INVALID_DATES,ERROR_FOR_DIVISION_BY_ZERO,TRADITIONAL,NO_AUTO_CREATE_USER,HIGH_NOT_PRECEDENCE,NO_ENGINE_SUBSTITUTION,PAD_CHAR_TO_FULL_LENGTH +REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,IGNORE_BAD_TABLE_OPTIONS,ONLY_FULL_GROUP_BY,NO_UNSIGNED_SUBTRACTION,NO_DIR_IN_CREATE,POSTGRESQL,ORACLE,MSSQL,DB2,MAXDB,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,MYSQL323,MYSQL40,ANSI,NO_AUTO_VALUE_ON_ZERO,NO_BACKSLASH_ESCAPES,STRICT_TRANS_TABLES,STRICT_ALL_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ALLOW_INVALID_DATES,ERROR_FOR_DIVISION_BY_ZERO,TRADITIONAL,NO_AUTO_CREATE_USER,HIGH_NOT_PRECEDENCE,NO_ENGINE_SUBSTITUTION,PAD_CHAR_TO_FULL_LENGTH set @@sql_mode= @old_mode; -select replace(@full_mode, '?', 'NOT_USED') into @full_mode; select replace(@full_mode, 'ALLOW_INVALID_DATES', 'INVALID_DATES') into @full_mode; select name from mysql.event where name = 'p' and sql_mode = @full_mode; name === modified file 'mysql-test/r/func_time.result' --- a/mysql-test/r/func_time.result 2010-01-21 08:10:05 +0000 +++ b/mysql-test/r/func_time.result 2010-04-08 09:47:31 +0000 @@ -1120,7 +1120,7 @@ drop function t_slow_sysdate; drop table t1; SET GLOBAL log_bin_trust_function_creators = 0; create table t1 (a datetime, i int, b datetime); -insert into t1 select sysdate(), sleep(1), sysdate() from dual; +insert into t1 select sysdate(), sleep(2), sysdate() from dual; select a != b from t1; a != b 1 === modified file 'mysql-test/r/information_schema.result' --- a/mysql-test/r/information_schema.result 2010-03-15 11:51:23 +0000 +++ b/mysql-test/r/information_schema.result 2010-04-08 12:10:05 +0000 @@ -69,6 +69,7 @@ INNODB_LOCK_WAITS INNODB_RSEG INNODB_TABLE_STATS INNODB_TRX +KEY_CACHES KEY_COLUMN_USAGE PARTITIONS PLUGINS @@ -615,7 +616,7 @@ proc body longblob proc definer char(77) proc created timestamp proc modified timestamp -proc sql_mode set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +proc sql_mode set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') proc comment char(64) proc character_set_client char(32) proc collation_connection char(32) === modified file 'mysql-test/r/information_schema_all_engines.result' --- a/mysql-test/r/information_schema_all_engines.result 2010-02-01 06:14:12 +0000 +++ b/mysql-test/r/information_schema_all_engines.result 2010-02-16 16:41:11 +0000 @@ -13,6 +13,7 @@ FILES GLOBAL_STATUS GLOBAL_VARIABLES INDEX_STATISTICS +KEY_CACHES KEY_COLUMN_USAGE PARTITIONS PLUGINS @@ -76,6 +77,7 @@ FILES TABLE_SCHEMA GLOBAL_STATUS VARIABLE_NAME GLOBAL_VARIABLES VARIABLE_NAME INDEX_STATISTICS TABLE_SCHEMA +KEY_CACHES KEY_CACHE_NAME KEY_COLUMN_USAGE CONSTRAINT_SCHEMA PARTITIONS TABLE_SCHEMA PLUGINS PLUGIN_NAME @@ -139,6 +141,7 @@ FILES TABLE_SCHEMA GLOBAL_STATUS VARIABLE_NAME GLOBAL_VARIABLES VARIABLE_NAME INDEX_STATISTICS TABLE_SCHEMA +KEY_CACHES KEY_CACHE_NAME KEY_COLUMN_USAGE CONSTRAINT_SCHEMA PARTITIONS TABLE_SCHEMA PLUGINS PLUGIN_NAME @@ -221,6 +224,7 @@ INNODB_LOCK_WAITS information_schema.INN INNODB_RSEG information_schema.INNODB_RSEG 1 INNODB_TABLE_STATS information_schema.INNODB_TABLE_STATS 1 INNODB_TRX information_schema.INNODB_TRX 1 +KEY_CACHES information_schema.KEY_CACHES 1 KEY_COLUMN_USAGE information_schema.KEY_COLUMN_USAGE 1 PARTITIONS information_schema.PARTITIONS 1 PBXT_STATISTICS information_schema.PBXT_STATISTICS 1 @@ -259,6 +263,7 @@ Database: information_schema | GLOBAL_STATUS | | GLOBAL_VARIABLES | | INDEX_STATISTICS | +| KEY_CACHES | | KEY_COLUMN_USAGE | | PARTITIONS | | PLUGINS | @@ -312,6 +317,7 @@ Database: INFORMATION_SCHEMA | GLOBAL_STATUS | | GLOBAL_VARIABLES | | INDEX_STATISTICS | +| KEY_CACHES | | KEY_COLUMN_USAGE | | PARTITIONS | | PLUGINS | @@ -357,5 +363,5 @@ Wildcard: inf_rmation_schema +--------------------+ SELECT table_schema, count(*) FROM information_schema.TABLES WHERE table_schema IN ('mysql', 'INFORMATION_SCHEMA', 'test', 'mysqltest') AND table_name<>'ndb_binlog_index' AND table_name<>'ndb_apply_status' GROUP BY TABLE_SCHEMA; table_schema count(*) -information_schema 48 +information_schema 49 mysql 22 === modified file 'mysql-test/r/key_cache.result' --- a/mysql-test/r/key_cache.result 2009-03-16 19:54:50 +0000 +++ b/mysql-test/r/key_cache.result 2010-04-01 21:42:40 +0000 @@ -1,5 +1,7 @@ drop table if exists t1, t2, t3; -SET @save_key_buffer=@@key_buffer_size; +SET @save_key_buffer_size=@@key_buffer_size; +SET @save_key_cache_block_size=@@key_cache_block_size; +SET @save_key_cache_partitions=@@key_cache_partitions; SELECT @@key_buffer_size, @@small.key_buffer_size; @@key_buffer_size @@small.key_buffer_size 2097152 131072 @@ -37,7 +39,7 @@ SELECT @@small.key_buffer_size; SELECT @@medium.key_buffer_size; @@medium.key_buffer_size 0 -SET @@global.key_buffer_size=@save_key_buffer; +SET @@global.key_buffer_size=@save_key_buffer_size; SELECT @@default.key_buffer_size; ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'default.key_buffer_size' at line 1 SELECT @@skr.storage_engine="test"; @@ -366,3 +368,432 @@ Variable_name Value key_cache_block_size 1536 SET GLOBAL key_cache_block_size= @bug28478_key_cache_block_size; DROP TABLE t1; +set global key_buffer_size=@save_key_buffer_size; +set global key_cache_block_size=@save_key_cache_block_size; +select @@key_buffer_size; +@@key_buffer_size +2097152 +select @@key_cache_block_size; +@@key_cache_block_size +1024 +select @@key_cache_partitions; +@@key_cache_partitions +0 +create table t1 ( +p int not null auto_increment primary key, +a char(10)); +create table t2 ( +p int not null auto_increment primary key, +i int, a char(10), key k1(i), key k2(a)); +select @@key_cache_partitions; +@@key_cache_partitions +0 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default NULL NULL 2097152 1024 0 # 0 0 0 0 0 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), +(3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +p a +1 qqqq +2 yyyy +select * from t2; +p i a +1 1 qqqq +2 1 pppp +3 1 yyyy +4 3 zzzz +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; +show status like 'key_%'; +Variable_name Value +Key_blocks_not_flushed 0 +Key_blocks_unused KEY_BLOCKS_UNUSED +Key_blocks_used 4 +Key_read_requests 22 +Key_reads 0 +Key_write_requests 26 +Key_writes 6 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default NULL NULL 2097152 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +delete from t2 where a='zzzz'; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default NULL NULL 2097152 1024 4 # 0 29 0 32 9 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +delete from t1; +delete from t2; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default NULL NULL 2097152 1024 4 # 0 29 0 32 9 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +set global key_cache_partitions=2; +select @@key_cache_partitions; +@@key_cache_partitions +2 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 2097152 1024 0 # 0 0 0 0 0 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), +(3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +p a +1 qqqq +2 yyyy +select * from t2; +p i a +1 1 qqqq +2 1 pppp +3 1 yyyy +4 3 zzzz +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; +show status like 'key_%'; +Variable_name Value +Key_blocks_not_flushed 0 +Key_blocks_unused KEY_BLOCKS_UNUSED +Key_blocks_used 4 +Key_read_requests 22 +Key_reads 0 +Key_write_requests 26 +Key_writes 6 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 2097152 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +delete from t1; +delete from t2; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 2097152 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +set global key_cache_partitions=1; +select @@key_cache_partitions; +@@key_cache_partitions +1 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 1 NULL 2097152 1024 0 # 0 0 0 0 0 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), +(3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +p a +1 qqqq +2 yyyy +select * from t2; +p i a +1 1 qqqq +2 1 pppp +3 1 yyyy +4 3 zzzz +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; +show status like 'key_%'; +Variable_name Value +Key_blocks_not_flushed 0 +Key_blocks_unused KEY_BLOCKS_UNUSED +Key_blocks_used 4 +Key_read_requests 22 +Key_reads 0 +Key_write_requests 26 +Key_writes 6 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 1 NULL 2097152 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +delete from t1; +delete from t2; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 1 NULL 2097152 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 1 0 2 1 +flush tables; +flush status; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 1 NULL 2097152 1024 4 # 0 0 0 0 0 +small NULL NULL 1048576 1024 1 # 0 0 0 0 0 +set global key_buffer_size=32*1024; +select @@key_buffer_size; +@@key_buffer_size +32768 +set global key_cache_partitions=2; +select @@key_cache_partitions; +@@key_cache_partitions +2 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 0 # 0 0 0 0 0 +small NULL NULL 1048576 1024 1 # 0 0 0 0 0 +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), +(3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +p a +1 qqqq +2 yyyy +select * from t2; +p i a +1 1 qqqq +2 1 pppp +3 1 yyyy +4 3 zzzz +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 4 # 0 22 0 26 6 +small NULL NULL 1048576 1024 1 # 0 0 0 0 0 +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 6733 # 3684 103 +small NULL NULL 1048576 1024 # # 0 0 # 0 0 +select * from t1 where p between 1010 and 1020 ; +p a +select * from t2 where p between 1010 and 1020 ; +p i a +1010 2 pppp +1011 2 yyyy +1012 3 zzzz +1013 2 qqqq +1014 2 pppp +1015 2 yyyy +1016 3 zzzz +1017 2 qqqq +1018 2 pppp +1019 2 yyyy +1020 3 zzzz +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 6750 # 3684 103 +small NULL NULL 1048576 1024 # # 0 0 # 0 0 +flush tables; +flush status; +update t1 set a='zzzz' where a='qqqq'; +update t2 set i=1 where i=2; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3076 18 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +set global keycache1.key_buffer_size=256*1024; +select @@keycache1.key_buffer_size; +@@keycache1.key_buffer_size +262144 +set global keycache1.key_cache_partitions=7; +select @@keycache1.key_cache_partitions; +@@keycache1.key_cache_partitions +7 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3076 18 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 7 NULL 262143 2048 # # 0 0 0 0 0 +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +keycache1 7 NULL 262143 2048 0 # 0 0 0 0 0 +cache index t1 key (`primary`) in keycache1; +Table Op Msg_type Msg_text +test.t1 assign_to_keycache status OK +explain select p from t1 where p between 1010 and 1020; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY PRIMARY 4 NULL 1 Using where; Using index +select p from t1 where p between 1010 and 1020; +p +explain select i from t2 where p between 1010 and 1020; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range PRIMARY PRIMARY 4 NULL 28 Using where +select i from t2 where p between 1010 and 1020; +i +1 +1 +3 +1 +1 +1 +3 +1 +1 +1 +3 +explain select count(*) from t1, t2 where t1.p = t2.i; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 index k1 k1 5 NULL 1024 Using index +1 SIMPLE t1 eq_ref PRIMARY PRIMARY 4 test.t2.i 1 Using index +select count(*) from t1, t2 where t1.p = t2.i; +count(*) +256 +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 7 NULL 262143 2048 # # 0 14 3 0 0 +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +keycache1 7 NULL 262143 2048 3 # 0 14 3 0 0 +cache index t2 in keycache1; +Table Op Msg_type Msg_text +test.t2 assign_to_keycache status OK +update t2 set p=p+3000, i=2 where a='qqqq'; +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +keycache1 7 NULL 262143 2048 25 # 0 2082 25 1071 19 +set global keycache2.key_buffer_size=1024*1024; +cache index t2 in keycache2; +Table Op Msg_type Msg_text +test.t2 assign_to_keycache status OK +insert into t2 values (2000, 3, 'yyyy'); +select * from information_schema.key_caches where key_cache_name like "keycache2" + and partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +keycache2 NULL NULL 1048576 1024 6 # 0 6 6 3 3 +select * from information_schema.key_caches where key_cache_name like "key%" +and partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +keycache1 7 NULL 262143 2048 25 # 0 2082 25 1071 19 +keycache2 NULL NULL 1048576 1024 6 # 0 6 6 3 3 +cache index t2 in keycache1; +Table Op Msg_type Msg_text +test.t2 assign_to_keycache status OK +update t2 set p=p+5000 where a='zzzz'; +select * from t2 where p between 1010 and 1020; +p i a +1010 1 pppp +1011 1 yyyy +1014 1 pppp +1015 1 yyyy +1018 1 pppp +1019 1 yyyy +explain select p from t2 where p between 1010 and 1020; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range PRIMARY PRIMARY 4 NULL 7 Using where; Using index +select p from t2 where p between 1010 and 1020; +p +1010 +1011 +1014 +1015 +1018 +1019 +explain select i from t2 where a='yyyy' and i=3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref k1,k2 k1 5 const 188 Using where +select i from t2 where a='yyyy' and i=3; +i +3 +explain select a from t2 where a='yyyy' and i=3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref k1,k2 k1 5 const 188 Using where +select a from t2 where a='yyyy' and i=3 ; +a +yyyy +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 7 NULL 262143 2048 # # 0 3201 43 1594 30 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=2*1024; +insert into t2 values (7000, 3, 'yyyy'); +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 7 NULL 262143 2048 # # 0 6 6 3 3 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=8*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 3 NULL 262143 8192 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +insert into t2 values (8000, 3, 'yyyy'); +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 3 NULL 262143 8192 # # 0 6 5 3 3 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_buffer_size=64*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=2*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 3 NULL 65535 2048 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=8*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_buffer_size=0; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=8*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_buffer_size=0; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_buffer_size=128*1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 1 NULL 131072 8192 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +set global keycache1.key_cache_block_size=1024; +select * from information_schema.key_caches where partition_number is null; +KEY_CACHE_NAME PARTITIONS PARTITION_NUMBER FULL_SIZE BLOCK_SIZE USED_BLOCKS UNUSED_BLOCKS DIRTY_BLOCKS READ_REQUESTS READS WRITE_REQUESTS WRITES +default 2 NULL 32768 1024 # # 0 3172 24 1552 18 +small NULL NULL 1048576 1024 # # 0 0 0 0 0 +keycache1 7 NULL 131068 1024 # # 0 0 0 0 0 +keycache2 NULL NULL 1048576 1024 # # 0 6 6 3 3 +drop table t1,t2; +set global keycache1.key_buffer_size=0; +set global keycache2.key_buffer_size=0; +set global key_buffer_size=@save_key_buffer_size; +set global key_cache_partitions=@save_key_cache_partitions; === modified file 'mysql-test/r/plugin.result' --- a/mysql-test/r/plugin.result 2010-03-03 14:44:14 +0000 +++ b/mysql-test/r/plugin.result 2010-04-08 17:19:01 +0000 @@ -62,4 +62,47 @@ select @@global.example_ulong_var; set session sql_mode=@old_sql_mode; set session old=bla; ERROR HY000: Variable 'old' is a read only variable +#legal values +CREATE TABLE t1 ( a int complex='c,f,f,f' ) ENGINE=example ULL=10000 STR='dskj' one_or_two='one' YESNO=0; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `complex`='c,f,f,f' +) ENGINE=EXAMPLE DEFAULT CHARSET=latin1 `ULL`=10000 `STR`='dskj' `one_or_two`='one' `YESNO`=0 +drop table t1; +SET @OLD_SQL_MODE=@@SQL_MODE; +SET SQL_MODE='IGNORE_BAD_TABLE_OPTIONS'; +#illegal value fixed +CREATE TABLE t1 (a int) ENGINE=example ULL=10000000000000000000 one_or_two='ttt' YESNO=SSS; +Warnings: +Warning 1651 Incorrect value '10000000000000000000' for option 'ULL' +Warning 1651 Incorrect value 'ttt' for option 'one_or_two' +Warning 1651 Incorrect value 'SSS' for option 'YESNO' +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL +) ENGINE=EXAMPLE DEFAULT CHARSET=latin1 `ULL`=10000000000000000000 `one_or_two`='ttt' `YESNO`=SSS +#alter table +alter table t1 ULL=10000000; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL +) ENGINE=EXAMPLE DEFAULT CHARSET=latin1 `one_or_two`='ttt' `YESNO`=SSS `ULL`=10000000 +alter table t1 change a a int complex='c,c,c'; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `a` int(11) DEFAULT NULL `complex`='c,c,c' +) ENGINE=EXAMPLE DEFAULT CHARSET=latin1 `one_or_two`='ttt' `YESNO`=SSS `ULL`=10000000 +drop table t1; +#illegal value error +SET SQL_MODE=''; +CREATE TABLE t1 (a int) ENGINE=example ULL=10000000000000000000 one_or_two='ttt' YESNO=SSS; +ERROR HY000: Incorrect value '10000000000000000000' for option 'ULL' +SET @@SQL_MODE=@OLD_SQL_MODE; +select 1; +1 +1 UNINSTALL PLUGIN example; === added file 'mysql-test/r/select_pkeycache.result' --- a/mysql-test/r/select_pkeycache.result 1970-01-01 00:00:00 +0000 +++ b/mysql-test/r/select_pkeycache.result 2010-03-29 21:16:12 +0000 @@ -0,0 +1,4784 @@ +drop table if exists t1,t2,t3,t4,t11; +drop table if exists t1_1,t1_2,t9_1,t9_2,t1aa,t2aa; +drop view if exists v1; +CREATE TABLE t1 ( +Period smallint(4) unsigned zerofill DEFAULT '0000' NOT NULL, +Varor_period smallint(4) unsigned DEFAULT '0' NOT NULL +); +INSERT INTO t1 VALUES (9410,9412); +select period from t1; +period +9410 +select * from t1; +Period Varor_period +9410 9412 +select t1.* from t1; +Period Varor_period +9410 9412 +CREATE TABLE t2 ( +auto int not null auto_increment, +fld1 int(6) unsigned zerofill DEFAULT '000000' NOT NULL, +companynr tinyint(2) unsigned zerofill DEFAULT '00' NOT NULL, +fld3 char(30) DEFAULT '' NOT NULL, +fld4 char(35) DEFAULT '' NOT NULL, +fld5 char(35) DEFAULT '' NOT NULL, +fld6 char(4) DEFAULT '' NOT NULL, +UNIQUE fld1 (fld1), +KEY fld3 (fld3), +PRIMARY KEY (auto) +); +select t2.fld3 from t2 where companynr = 58 and fld3 like "%imaginable%"; +fld3 +imaginable +select fld3 from t2 where fld3 like "%cultivation" ; +fld3 +cultivation +select t2.fld3,companynr from t2 where companynr = 57+1 order by fld3; +fld3 companynr +concoct 58 +druggists 58 +engrossing 58 +Eurydice 58 +exclaimers 58 +ferociousness 58 +hopelessness 58 +Huey 58 +imaginable 58 +judges 58 +merging 58 +ostrich 58 +peering 58 +Phelps 58 +presumes 58 +Ruth 58 +sentences 58 +Shylock 58 +straggled 58 +synergy 58 +thanking 58 +tying 58 +unlocks 58 +select fld3,companynr from t2 where companynr = 58 order by fld3; +fld3 companynr +concoct 58 +druggists 58 +engrossing 58 +Eurydice 58 +exclaimers 58 +ferociousness 58 +hopelessness 58 +Huey 58 +imaginable 58 +judges 58 +merging 58 +ostrich 58 +peering 58 +Phelps 58 +presumes 58 +Ruth 58 +sentences 58 +Shylock 58 +straggled 58 +synergy 58 +thanking 58 +tying 58 +unlocks 58 +select fld3 from t2 order by fld3 desc limit 10; +fld3 +youthfulness +yelped +Wotan +workers +Witt +witchcraft +Winsett +Willy +willed +wildcats +select fld3 from t2 order by fld3 desc limit 5; +fld3 +youthfulness +yelped +Wotan +workers +Witt +select fld3 from t2 order by fld3 desc limit 5,5; +fld3 +witchcraft +Winsett +Willy +willed +wildcats +select t2.fld3 from t2 where fld3 = 'honeysuckle'; +fld3 +honeysuckle +select t2.fld3 from t2 where fld3 LIKE 'honeysuckl_'; +fld3 +honeysuckle +select t2.fld3 from t2 where fld3 LIKE 'hon_ysuckl_'; +fld3 +honeysuckle +select t2.fld3 from t2 where fld3 LIKE 'honeysuckle%'; +fld3 +honeysuckle +select t2.fld3 from t2 where fld3 LIKE 'h%le'; +fld3 +honeysuckle +select t2.fld3 from t2 where fld3 LIKE 'honeysuckle_'; +fld3 +select t2.fld3 from t2 where fld3 LIKE 'don_t_find_me_please%'; +fld3 +explain select t2.fld3 from t2 where fld3 = 'honeysuckle'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref fld3 fld3 30 const 1 Using where; Using index +explain select fld3 from t2 ignore index (fld3) where fld3 = 'honeysuckle'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +explain select fld3 from t2 use index (fld1) where fld3 = 'honeysuckle'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +explain select fld3 from t2 use index (fld3) where fld3 = 'honeysuckle'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref fld3 fld3 30 const 1 Using where; Using index +explain select fld3 from t2 use index (fld1,fld3) where fld3 = 'honeysuckle'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref fld3 fld3 30 const 1 Using where; Using index +explain select fld3 from t2 ignore index (fld3,not_used); +ERROR 42000: Key 'not_used' doesn't exist in table 't2' +explain select fld3 from t2 use index (not_used); +ERROR 42000: Key 'not_used' doesn't exist in table 't2' +select t2.fld3 from t2 where fld3 >= 'honeysuckle' and fld3 <= 'honoring' order by fld3; +fld3 +honeysuckle +honoring +explain select t2.fld3 from t2 where fld3 >= 'honeysuckle' and fld3 <= 'honoring' order by fld3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range fld3 fld3 30 NULL 2 Using where; Using index +select fld1,fld3 from t2 where fld3="Colombo" or fld3 = "nondecreasing" order by fld3; +fld1 fld3 +148504 Colombo +068305 Colombo +000000 nondecreasing +select fld1,fld3 from t2 where companynr = 37 and fld3 = 'appendixes'; +fld1 fld3 +232605 appendixes +1232605 appendixes +1232606 appendixes +1232607 appendixes +1232608 appendixes +1232609 appendixes +select fld1 from t2 where fld1=250501 or fld1="250502"; +fld1 +250501 +250502 +explain select fld1 from t2 where fld1=250501 or fld1="250502"; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range fld1 fld1 4 NULL 2 Using where; Using index +select fld1 from t2 where fld1=250501 or fld1=250502 or fld1 >= 250505 and fld1 <= 250601 or fld1 between 250501 and 250502; +fld1 +250501 +250502 +250505 +250601 +explain select fld1 from t2 where fld1=250501 or fld1=250502 or fld1 >= 250505 and fld1 <= 250601 or fld1 between 250501 and 250502; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range fld1 fld1 4 NULL 4 Using where; Using index +select fld1,fld3 from t2 where companynr = 37 and fld3 like 'f%'; +fld1 fld3 +218401 faithful +018007 fanatic +228311 fated +018017 featherweight +218022 feed +088303 feminine +058004 Fenton +038017 fetched +018054 fetters +208101 fiftieth +238007 filial +013606 fingerings +218008 finishers +038205 firearm +188505 fitting +202301 Fitzpatrick +238008 fixedly +012001 flanking +018103 flint +018104 flopping +188007 flurried +013602 foldout +226205 foothill +232102 forgivably +228306 forthcoming +186002 freakish +208113 freest +231315 freezes +036002 funereal +226209 furnishings +198006 furthermore +select fld3 from t2 where fld3 like "L%" and fld3 = "ok"; +fld3 +select fld3 from t2 where (fld3 like "C%" and fld3 = "Chantilly"); +fld3 +Chantilly +select fld1,fld3 from t2 where fld1 like "25050%"; +fld1 fld3 +250501 poisoning +250502 Iraqis +250503 heaving +250504 population +250505 bomb +select fld1,fld3 from t2 where fld1 like "25050_"; +fld1 fld3 +250501 poisoning +250502 Iraqis +250503 heaving +250504 population +250505 bomb +select distinct companynr from t2; +companynr +00 +37 +36 +50 +58 +29 +40 +53 +65 +41 +34 +68 +select distinct companynr from t2 order by companynr; +companynr +00 +29 +34 +36 +37 +40 +41 +50 +53 +58 +65 +68 +select distinct companynr from t2 order by companynr desc; +companynr +68 +65 +58 +53 +50 +41 +40 +37 +36 +34 +29 +00 +select distinct t2.fld3,period from t2,t1 where companynr=37 and fld3 like "O%"; +fld3 period +obliterates 9410 +offload 9410 +opaquely 9410 +organizer 9410 +overestimating 9410 +overlay 9410 +select distinct fld3 from t2 where companynr = 34 order by fld3; +fld3 +absentee +accessed +ahead +alphabetic +Asiaticizations +attitude +aye +bankruptcies +belays +Blythe +bomb +boulevard +bulldozes +cannot +caressing +charcoal +checksumming +chess +clubroom +colorful +cosy +creator +crying +Darius +diffusing +duality +Eiffel +Epiphany +Ernestine +explorers +exterminated +famine +forked +Gershwins +heaving +Hodges +Iraqis +Italianization +Lagos +landslide +libretto +Majorca +mastering +narrowed +occurred +offerers +Palestine +Peruvianizes +pharmaceutic +poisoning +population +Pygmalion +rats +realest +recording +regimented +retransmitting +reviver +rouses +scars +sicker +sleepwalk +stopped +sugars +translatable +uncles +unexpected +uprisings +versatility +vest +select distinct fld3 from t2 limit 10; +fld3 +abates +abiding +Abraham +abrogating +absentee +abut +accessed +accruing +accumulating +accuracies +select distinct fld3 from t2 having fld3 like "A%" limit 10; +fld3 +abates +abiding +Abraham +abrogating +absentee +abut +accessed +accruing +accumulating +accuracies +select distinct substring(fld3,1,3) from t2 where fld3 like "A%"; +substring(fld3,1,3) +aba +abi +Abr +abs +abu +acc +acq +acu +Ade +adj +Adl +adm +Ado +ads +adv +aer +aff +afi +afl +afo +agi +ahe +aim +air +Ald +alg +ali +all +alp +alr +ama +ame +amm +ana +and +ane +Ang +ani +Ann +Ant +api +app +aqu +Ara +arc +Arm +arr +Art +Asi +ask +asp +ass +ast +att +aud +Aug +aut +ave +avo +awe +aye +Azt +select distinct substring(fld3,1,3) as a from t2 having a like "A%" order by a limit 10; +a +aba +abi +Abr +abs +abu +acc +acq +acu +Ade +adj +select distinct substring(fld3,1,3) from t2 where fld3 like "A%" limit 10; +substring(fld3,1,3) +aba +abi +Abr +abs +abu +acc +acq +acu +Ade +adj +select distinct substring(fld3,1,3) as a from t2 having a like "A%" limit 10; +a +aba +abi +Abr +abs +abu +acc +acq +acu +Ade +adj +create table t3 ( +period int not null, +name char(32) not null, +companynr int not null, +price double(11,0), +price2 double(11,0), +key (period), +key (name) +); +create temporary table tmp engine = myisam select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +insert into tmp select * from t3; +insert into t3 select * from tmp; +alter table t3 add t2nr int not null auto_increment primary key first; +drop table tmp; +SET SQL_BIG_TABLES=1; +select distinct concat(fld3," ",fld3) as namn from t2,t3 where t2.fld1=t3.t2nr order by namn limit 10; +namn +Abraham Abraham +abrogating abrogating +admonishing admonishing +Adolph Adolph +afield afield +aging aging +ammonium ammonium +analyzable analyzable +animals animals +animized animized +SET SQL_BIG_TABLES=0; +select distinct concat(fld3," ",fld3) from t2,t3 where t2.fld1=t3.t2nr order by fld3 limit 10; +concat(fld3," ",fld3) +Abraham Abraham +abrogating abrogating +admonishing admonishing +Adolph Adolph +afield afield +aging aging +ammonium ammonium +analyzable analyzable +animals animals +animized animized +select distinct fld5 from t2 limit 10; +fld5 +neat +Steinberg +jarring +tinily +balled +persist +attainments +fanatic +measures +rightfulness +select distinct fld3,count(*) from t2 group by companynr,fld3 limit 10; +fld3 count(*) +affixed 1 +and 1 +annoyers 1 +Anthony 1 +assayed 1 +assurers 1 +attendants 1 +bedlam 1 +bedpost 1 +boasted 1 +SET SQL_BIG_TABLES=1; +select distinct fld3,count(*) from t2 group by companynr,fld3 limit 10; +fld3 count(*) +affixed 1 +and 1 +annoyers 1 +Anthony 1 +assayed 1 +assurers 1 +attendants 1 +bedlam 1 +bedpost 1 +boasted 1 +SET SQL_BIG_TABLES=0; +select distinct fld3,repeat("a",length(fld3)),count(*) from t2 group by companynr,fld3 limit 100,10; +fld3 repeat("a",length(fld3)) count(*) +circus aaaaaa 1 +cited aaaaa 1 +Colombo aaaaaaa 1 +congresswoman aaaaaaaaaaaaa 1 +contrition aaaaaaaaaa 1 +corny aaaaa 1 +cultivation aaaaaaaaaaa 1 +definiteness aaaaaaaaaaaa 1 +demultiplex aaaaaaaaaaa 1 +disappointing aaaaaaaaaaaaa 1 +select distinct companynr,rtrim(space(512+companynr)) from t3 order by 1,2; +companynr rtrim(space(512+companynr)) +37 +78 +101 +154 +311 +447 +512 +select distinct fld3 from t2,t3 where t2.companynr = 34 and t2.fld1=t3.t2nr order by fld3; +fld3 +explain select t3.t2nr,fld3 from t2,t3 where t2.companynr = 34 and t2.fld1=t3.t2nr order by t3.t2nr,fld3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL fld1 NULL NULL NULL 1199 Using where; Using temporary; Using filesort +1 SIMPLE t3 eq_ref PRIMARY PRIMARY 4 test.t2.fld1 1 Using where; Using index +explain select * from t3 as t1,t3 where t1.period=t3.period order by t3.period; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL period NULL NULL NULL 41810 Using temporary; Using filesort +1 SIMPLE t3 ref period period 4 test.t1.period 4181 +explain select * from t3 as t1,t3 where t1.period=t3.period order by t3.period limit 10; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t3 index period period 4 NULL 1 +1 SIMPLE t1 ref period period 4 test.t3.period 4181 +explain select * from t3 as t1,t3 where t1.period=t3.period order by t1.period limit 10; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 index period period 4 NULL 1 +1 SIMPLE t3 ref period period 4 test.t1.period 4181 +select period from t1; +period +9410 +select period from t1 where period=1900; +period +select fld3,period from t1,t2 where fld1 = 011401 order by period; +fld3 period +breaking 9410 +select fld3,period from t2,t3 where t2.fld1 = 011401 and t2.fld1=t3.t2nr and t3.period=1001; +fld3 period +breaking 1001 +explain select fld3,period from t2,t3 where t2.fld1 = 011401 and t3.t2nr=t2.fld1 and 1001 = t3.period; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 const fld1 fld1 4 const 1 +1 SIMPLE t3 const PRIMARY,period PRIMARY 4 const 1 +select fld3,period from t2,t1 where companynr*10 = 37*10; +fld3 period +breaking 9410 +Romans 9410 +intercepted 9410 +bewilderingly 9410 +astound 9410 +admonishing 9410 +sumac 9410 +flanking 9410 +combed 9410 +subjective 9410 +scatterbrain 9410 +Eulerian 9410 +Kane 9410 +overlay 9410 +perturb 9410 +goblins 9410 +annihilates 9410 +Wotan 9410 +snatching 9410 +concludes 9410 +laterally 9410 +yelped 9410 +grazing 9410 +Baird 9410 +celery 9410 +misunderstander 9410 +handgun 9410 +foldout 9410 +mystic 9410 +succumbed 9410 +Nabisco 9410 +fingerings 9410 +aging 9410 +afield 9410 +ammonium 9410 +boat 9410 +intelligibility 9410 +Augustine 9410 +teethe 9410 +dreaded 9410 +scholastics 9410 +audiology 9410 +wallet 9410 +parters 9410 +eschew 9410 +quitter 9410 +neat 9410 +Steinberg 9410 +jarring 9410 +tinily 9410 +balled 9410 +persist 9410 +attainments 9410 +fanatic 9410 +measures 9410 +rightfulness 9410 +capably 9410 +impulsive 9410 +starlet 9410 +terminators 9410 +untying 9410 +announces 9410 +featherweight 9410 +pessimist 9410 +daughter 9410 +decliner 9410 +lawgiver 9410 +stated 9410 +readable 9410 +attrition 9410 +cascade 9410 +motors 9410 +interrogate 9410 +pests 9410 +stairway 9410 +dopers 9410 +testicle 9410 +Parsifal 9410 +leavings 9410 +postulation 9410 +squeaking 9410 +contrasted 9410 +leftover 9410 +whiteners 9410 +erases 9410 +Punjab 9410 +Merritt 9410 +Quixotism 9410 +sweetish 9410 +dogging 9410 +scornfully 9410 +bellow 9410 +bills 9410 +cupboard 9410 +sureties 9410 +puddings 9410 +fetters 9410 +bivalves 9410 +incurring 9410 +Adolph 9410 +pithed 9410 +Miles 9410 +trimmings 9410 +tragedies 9410 +skulking 9410 +flint 9410 +flopping 9410 +relaxing 9410 +offload 9410 +suites 9410 +lists 9410 +animized 9410 +multilayer 9410 +standardizes 9410 +Judas 9410 +vacuuming 9410 +dentally 9410 +humanness 9410 +inch 9410 +Weissmuller 9410 +irresponsibly 9410 +luckily 9410 +culled 9410 +medical 9410 +bloodbath 9410 +subschema 9410 +animals 9410 +Micronesia 9410 +repetitions 9410 +Antares 9410 +ventilate 9410 +pityingly 9410 +interdependent 9410 +Graves 9410 +neonatal 9410 +chafe 9410 +honoring 9410 +realtor 9410 +elite 9410 +funereal 9410 +abrogating 9410 +sorters 9410 +Conley 9410 +lectured 9410 +Abraham 9410 +Hawaii 9410 +cage 9410 +hushes 9410 +Simla 9410 +reporters 9410 +Dutchman 9410 +descendants 9410 +groupings 9410 +dissociate 9410 +coexist 9410 +Beebe 9410 +Taoism 9410 +Connally 9410 +fetched 9410 +checkpoints 9410 +rusting 9410 +galling 9410 +obliterates 9410 +traitor 9410 +resumes 9410 +analyzable 9410 +terminator 9410 +gritty 9410 +firearm 9410 +minima 9410 +Selfridge 9410 +disable 9410 +witchcraft 9410 +betroth 9410 +Manhattanize 9410 +imprint 9410 +peeked 9410 +swelling 9410 +interrelationships 9410 +riser 9410 +Gandhian 9410 +peacock 9410 +bee 9410 +kanji 9410 +dental 9410 +scarf 9410 +chasm 9410 +insolence 9410 +syndicate 9410 +alike 9410 +imperial 9410 +convulsion 9410 +railway 9410 +validate 9410 +normalizes 9410 +comprehensive 9410 +chewing 9410 +denizen 9410 +schemer 9410 +chronicle 9410 +Kline 9410 +Anatole 9410 +partridges 9410 +brunch 9410 +recruited 9410 +dimensions 9410 +Chicana 9410 +announced 9410 +praised 9410 +employing 9410 +linear 9410 +quagmire 9410 +western 9410 +relishing 9410 +serving 9410 +scheduling 9410 +lore 9410 +eventful 9410 +arteriole 9410 +disentangle 9410 +cured 9410 +Fenton 9410 +avoidable 9410 +drains 9410 +detectably 9410 +husky 9410 +impelling 9410 +undoes 9410 +evened 9410 +squeezes 9410 +destroyer 9410 +rudeness 9410 +beaner 9410 +boorish 9410 +Everhart 9410 +encompass 9410 +mushrooms 9410 +Alison 9410 +externally 9410 +pellagra 9410 +cult 9410 +creek 9410 +Huffman 9410 +Majorca 9410 +governing 9410 +gadfly 9410 +reassigned 9410 +intentness 9410 +craziness 9410 +psychic 9410 +squabbled 9410 +burlesque 9410 +capped 9410 +extracted 9410 +DiMaggio 9410 +exclamation 9410 +subdirectory 9410 +Gothicism 9410 +feminine 9410 +metaphysically 9410 +sanding 9410 +Miltonism 9410 +freakish 9410 +index 9410 +straight 9410 +flurried 9410 +denotative 9410 +coming 9410 +commencements 9410 +gentleman 9410 +gifted 9410 +Shanghais 9410 +sportswriting 9410 +sloping 9410 +navies 9410 +leaflet 9410 +shooter 9410 +Joplin 9410 +babies 9410 +assails 9410 +admiring 9410 +swaying 9410 +Goldstine 9410 +fitting 9410 +Norwalk 9410 +analogy 9410 +deludes 9410 +cokes 9410 +Clayton 9410 +exhausts 9410 +causality 9410 +sating 9410 +icon 9410 +throttles 9410 +communicants 9410 +dehydrate 9410 +priceless 9410 +publicly 9410 +incidentals 9410 +commonplace 9410 +mumbles 9410 +furthermore 9410 +cautioned 9410 +parametrized 9410 +registration 9410 +sadly 9410 +positioning 9410 +babysitting 9410 +eternal 9410 +hoarder 9410 +congregates 9410 +rains 9410 +workers 9410 +sags 9410 +unplug 9410 +garage 9410 +boulder 9410 +specifics 9410 +Teresa 9410 +Winsett 9410 +convenient 9410 +buckboards 9410 +amenities 9410 +resplendent 9410 +sews 9410 +participated 9410 +Simon 9410 +certificates 9410 +Fitzpatrick 9410 +Evanston 9410 +misted 9410 +textures 9410 +save 9410 +count 9410 +rightful 9410 +chaperone 9410 +Lizzy 9410 +clenched 9410 +effortlessly 9410 +accessed 9410 +beaters 9410 +Hornblower 9410 +vests 9410 +indulgences 9410 +infallibly 9410 +unwilling 9410 +excrete 9410 +spools 9410 +crunches 9410 +overestimating 9410 +ineffective 9410 +humiliation 9410 +sophomore 9410 +star 9410 +rifles 9410 +dialysis 9410 +arriving 9410 +indulge 9410 +clockers 9410 +languages 9410 +Antarctica 9410 +percentage 9410 +ceiling 9410 +specification 9410 +regimented 9410 +ciphers 9410 +pictures 9410 +serpents 9410 +allot 9410 +realized 9410 +mayoral 9410 +opaquely 9410 +hostess 9410 +fiftieth 9410 +incorrectly 9410 +decomposition 9410 +stranglings 9410 +mixture 9410 +electroencephalography 9410 +similarities 9410 +charges 9410 +freest 9410 +Greenberg 9410 +tinting 9410 +expelled 9410 +warm 9410 +smoothed 9410 +deductions 9410 +Romano 9410 +bitterroot 9410 +corset 9410 +securing 9410 +environing 9410 +cute 9410 +Crays 9410 +heiress 9410 +inform 9410 +avenge 9410 +universals 9410 +Kinsey 9410 +ravines 9410 +bestseller 9410 +equilibrium 9410 +extents 9410 +relatively 9410 +pressure 9410 +critiques 9410 +befouled 9410 +rightfully 9410 +mechanizing 9410 +Latinizes 9410 +timesharing 9410 +Aden 9410 +embassies 9410 +males 9410 +shapelessly 9410 +mastering 9410 +Newtonian 9410 +finishers 9410 +abates 9410 +teem 9410 +kiting 9410 +stodgy 9410 +feed 9410 +guitars 9410 +airships 9410 +store 9410 +denounces 9410 +Pyle 9410 +Saxony 9410 +serializations 9410 +Peruvian 9410 +taxonomically 9410 +kingdom 9410 +stint 9410 +Sault 9410 +faithful 9410 +Ganymede 9410 +tidiness 9410 +gainful 9410 +contrary 9410 +Tipperary 9410 +tropics 9410 +theorizers 9410 +renew 9410 +already 9410 +terminal 9410 +Hegelian 9410 +hypothesizer 9410 +warningly 9410 +journalizing 9410 +nested 9410 +Lars 9410 +saplings 9410 +foothill 9410 +labeled 9410 +imperiously 9410 +reporters 9410 +furnishings 9410 +precipitable 9410 +discounts 9410 +excises 9410 +Stalin 9410 +despot 9410 +ripeness 9410 +Arabia 9410 +unruly 9410 +mournfulness 9410 +boom 9410 +slaughter 9410 +Sabine 9410 +handy 9410 +rural 9410 +organizer 9410 +shipyard 9410 +civics 9410 +inaccuracy 9410 +rules 9410 +juveniles 9410 +comprised 9410 +investigations 9410 +stabilizes 9410 +seminaries 9410 +Hunter 9410 +sporty 9410 +test 9410 +weasels 9410 +CERN 9410 +tempering 9410 +afore 9410 +Galatean 9410 +techniques 9410 +error 9410 +veranda 9410 +severely 9410 +Cassites 9410 +forthcoming 9410 +guides 9410 +vanish 9410 +lied 9410 +sawtooth 9410 +fated 9410 +gradually 9410 +widens 9410 +preclude 9410 +evenhandedly 9410 +percentage 9410 +disobedience 9410 +humility 9410 +gleaning 9410 +petted 9410 +bloater 9410 +minion 9410 +marginal 9410 +apiary 9410 +measures 9410 +precaution 9410 +repelled 9410 +primary 9410 +coverings 9410 +Artemia 9410 +navigate 9410 +spatial 9410 +Gurkha 9410 +meanwhile 9410 +Melinda 9410 +Butterfield 9410 +Aldrich 9410 +previewing 9410 +glut 9410 +unaffected 9410 +inmate 9410 +mineral 9410 +impending 9410 +meditation 9410 +ideas 9410 +miniaturizes 9410 +lewdly 9410 +title 9410 +youthfulness 9410 +creak 9410 +Chippewa 9410 +clamored 9410 +freezes 9410 +forgivably 9410 +reduce 9410 +McGovern 9410 +Nazis 9410 +epistle 9410 +socializes 9410 +conceptions 9410 +Kevin 9410 +uncovering 9410 +chews 9410 +appendixes 9410 +appendixes 9410 +appendixes 9410 +appendixes 9410 +appendixes 9410 +appendixes 9410 +raining 9410 +infest 9410 +compartment 9410 +minting 9410 +ducks 9410 +roped 9410 +waltz 9410 +Lillian 9410 +repressions 9410 +chillingly 9410 +noncritical 9410 +lithograph 9410 +spongers 9410 +parenthood 9410 +posed 9410 +instruments 9410 +filial 9410 +fixedly 9410 +relives 9410 +Pandora 9410 +watering 9410 +ungrateful 9410 +secures 9410 +poison 9410 +dusted 9410 +encompasses 9410 +presentation 9410 +Kantian 9410 +select fld3,period,price,price2 from t2,t3 where t2.fld1=t3.t2nr and period >= 1001 and period <= 1002 and t2.companynr = 37 order by fld3,period, price; +fld3 period price price2 +admonishing 1002 28357832 8723648 +analyzable 1002 28357832 8723648 +annihilates 1001 5987435 234724 +Antares 1002 28357832 8723648 +astound 1001 5987435 234724 +audiology 1001 5987435 234724 +Augustine 1002 28357832 8723648 +Baird 1002 28357832 8723648 +bewilderingly 1001 5987435 234724 +breaking 1001 5987435 234724 +Conley 1001 5987435 234724 +dentally 1002 28357832 8723648 +dissociate 1002 28357832 8723648 +elite 1001 5987435 234724 +eschew 1001 5987435 234724 +Eulerian 1001 5987435 234724 +flanking 1001 5987435 234724 +foldout 1002 28357832 8723648 +funereal 1002 28357832 8723648 +galling 1002 28357832 8723648 +Graves 1001 5987435 234724 +grazing 1001 5987435 234724 +groupings 1001 5987435 234724 +handgun 1001 5987435 234724 +humility 1002 28357832 8723648 +impulsive 1002 28357832 8723648 +inch 1001 5987435 234724 +intelligibility 1001 5987435 234724 +jarring 1001 5987435 234724 +lawgiver 1001 5987435 234724 +lectured 1002 28357832 8723648 +Merritt 1002 28357832 8723648 +neonatal 1001 5987435 234724 +offload 1002 28357832 8723648 +parters 1002 28357832 8723648 +pityingly 1002 28357832 8723648 +puddings 1002 28357832 8723648 +Punjab 1001 5987435 234724 +quitter 1002 28357832 8723648 +realtor 1001 5987435 234724 +relaxing 1001 5987435 234724 +repetitions 1001 5987435 234724 +resumes 1001 5987435 234724 +Romans 1002 28357832 8723648 +rusting 1001 5987435 234724 +scholastics 1001 5987435 234724 +skulking 1002 28357832 8723648 +stated 1002 28357832 8723648 +suites 1002 28357832 8723648 +sureties 1001 5987435 234724 +testicle 1002 28357832 8723648 +tinily 1002 28357832 8723648 +tragedies 1001 5987435 234724 +trimmings 1001 5987435 234724 +vacuuming 1001 5987435 234724 +ventilate 1001 5987435 234724 +wallet 1001 5987435 234724 +Weissmuller 1002 28357832 8723648 +Wotan 1002 28357832 8723648 +select t2.fld1,fld3,period,price,price2 from t2,t3 where t2.fld1>= 18201 and t2.fld1 <= 18811 and t2.fld1=t3.t2nr and period = 1001 and t2.companynr = 37; +fld1 fld3 period price price2 +018201 relaxing 1001 5987435 234724 +018601 vacuuming 1001 5987435 234724 +018801 inch 1001 5987435 234724 +018811 repetitions 1001 5987435 234724 +create table t4 ( +companynr tinyint(2) unsigned zerofill NOT NULL default '00', +companyname char(30) NOT NULL default '', +PRIMARY KEY (companynr), +UNIQUE KEY companyname(companyname) +) ENGINE=MyISAM MAX_ROWS=50 PACK_KEYS=1 COMMENT='companynames'; +select STRAIGHT_JOIN t2.companynr,companyname from t4,t2 where t2.companynr=t4.companynr group by t2.companynr; +companynr companyname +00 Unknown +29 company 1 +34 company 2 +36 company 3 +37 company 4 +40 company 5 +41 company 6 +50 company 11 +53 company 7 +58 company 8 +65 company 9 +68 company 10 +select SQL_SMALL_RESULT t2.companynr,companyname from t4,t2 where t2.companynr=t4.companynr group by t2.companynr; +companynr companyname +00 Unknown +29 company 1 +34 company 2 +36 company 3 +37 company 4 +40 company 5 +41 company 6 +50 company 11 +53 company 7 +58 company 8 +65 company 9 +68 company 10 +select * from t1,t1 t12; +Period Varor_period Period Varor_period +9410 9412 9410 9412 +select t2.fld1,t22.fld1 from t2,t2 t22 where t2.fld1 >= 250501 and t2.fld1 <= 250505 and t22.fld1 >= 250501 and t22.fld1 <= 250505; +fld1 fld1 +250501 250501 +250502 250501 +250503 250501 +250504 250501 +250505 250501 +250501 250502 +250502 250502 +250503 250502 +250504 250502 +250505 250502 +250501 250503 +250502 250503 +250503 250503 +250504 250503 +250505 250503 +250501 250504 +250502 250504 +250503 250504 +250504 250504 +250505 250504 +250501 250505 +250502 250505 +250503 250505 +250504 250505 +250505 250505 +insert into t2 (fld1, companynr) values (999999,99); +select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null; +companynr companyname +99 NULL +select count(*) from t2 left join t4 using (companynr) where t4.companynr is not null; +count(*) +1199 +explain select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1200 +1 SIMPLE t4 eq_ref PRIMARY PRIMARY 1 test.t2.companynr 1 Using where; Not exists +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL NULL NULL NULL NULL 12 +1 SIMPLE t2 ALL NULL NULL NULL NULL 1200 Using where; Not exists +select companynr,companyname from t2 left join t4 using (companynr) where companynr is null; +companynr companyname +select count(*) from t2 left join t4 using (companynr) where companynr is not null; +count(*) +1200 +explain select companynr,companyname from t2 left join t4 using (companynr) where companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Impossible WHERE +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Impossible WHERE +delete from t2 where fld1=999999; +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +1 SIMPLE t4 eq_ref PRIMARY PRIMARY 1 test.t2.companynr 1 +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +1 SIMPLE t4 eq_ref PRIMARY PRIMARY 1 test.t2.companynr 1 +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 and t4.companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +1 SIMPLE t4 eq_ref PRIMARY PRIMARY 1 test.t2.companynr 1 +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr > 0 or companynr < 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr > 0 and companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL NULL NULL NULL NULL 12 +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0 or t4.companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +explain select t2.companynr,companyname from t4 left join t2 using (companynr) where ifnull(t2.companynr,1)>0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL NULL NULL NULL NULL 12 +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr > 0 or companynr is null; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +explain select companynr,companyname from t4 left join t2 using (companynr) where companynr > 0 or companynr < 0 or companynr > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL PRIMARY NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +explain select companynr,companyname from t4 left join t2 using (companynr) where ifnull(companynr,1)>0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 ALL NULL NULL NULL NULL 12 Using where +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +select distinct t2.companynr,t4.companynr from t2,t4 where t2.companynr=t4.companynr+1; +companynr companynr +37 36 +41 40 +explain select distinct t2.companynr,t4.companynr from t2,t4 where t2.companynr=t4.companynr+1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t4 index NULL PRIMARY 1 NULL 12 Using index; Using temporary +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 Using where; Using join buffer +select t2.fld1,t2.companynr,fld3,period from t3,t2 where t2.fld1 = 38208 and t2.fld1=t3.t2nr and period = 1008 or t2.fld1 = 38008 and t2.fld1 =t3.t2nr and period = 1008; +fld1 companynr fld3 period +038008 37 reporters 1008 +038208 37 Selfridge 1008 +select t2.fld1,t2.companynr,fld3,period from t3,t2 where (t2.fld1 = 38208 or t2.fld1 = 38008) and t2.fld1=t3.t2nr and period>=1008 and period<=1009; +fld1 companynr fld3 period +038008 37 reporters 1008 +038208 37 Selfridge 1008 +select t2.fld1,t2.companynr,fld3,period from t3,t2 where (t3.t2nr = 38208 or t3.t2nr = 38008) and t2.fld1=t3.t2nr and period>=1008 and period<=1009; +fld1 companynr fld3 period +038008 37 reporters 1008 +038208 37 Selfridge 1008 +select period from t1 where (((period > 0) or period < 10000 or (period = 1900)) and (period=1900 and period <= 1901) or (period=1903 and (period=1903)) and period>=1902) or ((period=1904 or period=1905) or (period=1906 or period>1907)) or (period=1908 and period = 1909); +period +9410 +select period from t1 where ((period > 0 and period < 1) or (((period > 0 and period < 100) and (period > 10)) or (period > 10)) or (period > 0 and (period > 5 or period > 6))); +period +9410 +select a.fld1 from t2 as a,t2 b where ((a.fld1 = 250501 and a.fld1=b.fld1) or a.fld1=250502 or a.fld1=250503 or (a.fld1=250505 and a.fld1<=b.fld1 and b.fld1>=a.fld1)) and a.fld1=b.fld1; +fld1 +250501 +250502 +250503 +250505 +select fld1 from t2 where fld1 in (250502,98005,98006,250503,250605,250606) and fld1 >=250502 and fld1 not in (250605,250606); +fld1 +250502 +250503 +select fld1 from t2 where fld1 between 250502 and 250504; +fld1 +250502 +250503 +250504 +select fld3 from t2 where (((fld3 like "_%L%" ) or (fld3 like "%ok%")) and ( fld3 like "L%" or fld3 like "G%")) and fld3 like "L%" ; +fld3 +label +labeled +labeled +landslide +laterally +leaflet +lewdly +Lillian +luckily +select count(*) from t1; +count(*) +1 +select companynr,count(*),sum(fld1) from t2 group by companynr; +companynr count(*) sum(fld1) +00 82 10355753 +29 95 14473298 +34 70 17788966 +36 215 22786296 +37 588 83602098 +40 37 6618386 +41 52 12816335 +50 11 1595438 +53 4 793210 +58 23 2254293 +65 10 2284055 +68 12 3097288 +select companynr,count(*) from t2 group by companynr order by companynr desc limit 5; +companynr count(*) +68 12 +65 10 +58 23 +53 4 +50 11 +select count(*),min(fld4),max(fld4),sum(fld1),avg(fld1),std(fld1),variance(fld1) from t2 where companynr = 34 and fld4<>""; +count(*) min(fld4) max(fld4) sum(fld1) avg(fld1) std(fld1) variance(fld1) +70 absentee vest 17788966 254128.0857 3272.5940 10709871.3069 +explain extended select count(*),min(fld4),max(fld4),sum(fld1),avg(fld1),std(fld1),variance(fld1) from t2 where companynr = 34 and fld4<>""; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 100.00 Using where +Warnings: +Note 1003 select count(0) AS `count(*)`,min(`test`.`t2`.`fld4`) AS `min(fld4)`,max(`test`.`t2`.`fld4`) AS `max(fld4)`,sum(`test`.`t2`.`fld1`) AS `sum(fld1)`,avg(`test`.`t2`.`fld1`) AS `avg(fld1)`,std(`test`.`t2`.`fld1`) AS `std(fld1)`,variance(`test`.`t2`.`fld1`) AS `variance(fld1)` from `test`.`t2` where ((`test`.`t2`.`companynr` = 34) and (`test`.`t2`.`fld4` <> '')) +select companynr,count(*),min(fld4),max(fld4),sum(fld1),avg(fld1),std(fld1),variance(fld1) from t2 group by companynr limit 3; +companynr count(*) min(fld4) max(fld4) sum(fld1) avg(fld1) std(fld1) variance(fld1) +00 82 Anthony windmills 10355753 126289.6707 115550.9757 13352027981.7087 +29 95 abut wetness 14473298 152350.5053 8368.5480 70032594.9026 +34 70 absentee vest 17788966 254128.0857 3272.5940 10709871.3069 +select companynr,t2nr,count(price),sum(price),min(price),max(price),avg(price) from t3 where companynr = 37 group by companynr,t2nr limit 10; +companynr t2nr count(price) sum(price) min(price) max(price) avg(price) +37 1 1 5987435 5987435 5987435 5987435.0000 +37 2 1 28357832 28357832 28357832 28357832.0000 +37 3 1 39654943 39654943 39654943 39654943.0000 +37 11 1 5987435 5987435 5987435 5987435.0000 +37 12 1 28357832 28357832 28357832 28357832.0000 +37 13 1 39654943 39654943 39654943 39654943.0000 +37 21 1 5987435 5987435 5987435 5987435.0000 +37 22 1 28357832 28357832 28357832 28357832.0000 +37 23 1 39654943 39654943 39654943 39654943.0000 +37 31 1 5987435 5987435 5987435 5987435.0000 +select /*! SQL_SMALL_RESULT */ companynr,t2nr,count(price),sum(price),min(price),max(price),avg(price) from t3 where companynr = 37 group by companynr,t2nr limit 10; +companynr t2nr count(price) sum(price) min(price) max(price) avg(price) +37 1 1 5987435 5987435 5987435 5987435.0000 +37 2 1 28357832 28357832 28357832 28357832.0000 +37 3 1 39654943 39654943 39654943 39654943.0000 +37 11 1 5987435 5987435 5987435 5987435.0000 +37 12 1 28357832 28357832 28357832 28357832.0000 +37 13 1 39654943 39654943 39654943 39654943.0000 +37 21 1 5987435 5987435 5987435 5987435.0000 +37 22 1 28357832 28357832 28357832 28357832.0000 +37 23 1 39654943 39654943 39654943 39654943.0000 +37 31 1 5987435 5987435 5987435 5987435.0000 +select companynr,count(price),sum(price),min(price),max(price),avg(price) from t3 group by companynr ; +companynr count(price) sum(price) min(price) max(price) avg(price) +37 12543 309394878010 5987435 39654943 24666736.6667 +78 8362 414611089292 726498 98439034 49582766.0000 +101 4181 3489454238 834598 834598 834598.0000 +154 4181 4112197254950 983543950 983543950 983543950.0000 +311 4181 979599938 234298 234298 234298.0000 +447 4181 9929180954 2374834 2374834 2374834.0000 +512 4181 3288532102 786542 786542 786542.0000 +select distinct mod(companynr,10) from t4 group by companynr; +mod(companynr,10) +0 +9 +4 +6 +7 +1 +3 +8 +5 +select distinct 1 from t4 group by companynr; +1 +1 +select count(distinct fld1) from t2; +count(distinct fld1) +1199 +select companynr,count(distinct fld1) from t2 group by companynr; +companynr count(distinct fld1) +00 82 +29 95 +34 70 +36 215 +37 588 +40 37 +41 52 +50 11 +53 4 +58 23 +65 10 +68 12 +select companynr,count(*) from t2 group by companynr; +companynr count(*) +00 82 +29 95 +34 70 +36 215 +37 588 +40 37 +41 52 +50 11 +53 4 +58 23 +65 10 +68 12 +select companynr,count(distinct concat(fld1,repeat(65,1000))) from t2 group by companynr; +companynr count(distinct concat(fld1,repeat(65,1000))) +00 82 +29 95 +34 70 +36 215 +37 588 +40 37 +41 52 +50 11 +53 4 +58 23 +65 10 +68 12 +select companynr,count(distinct concat(fld1,repeat(65,200))) from t2 group by companynr; +companynr count(distinct concat(fld1,repeat(65,200))) +00 82 +29 95 +34 70 +36 215 +37 588 +40 37 +41 52 +50 11 +53 4 +58 23 +65 10 +68 12 +select companynr,count(distinct floor(fld1/100)) from t2 group by companynr; +companynr count(distinct floor(fld1/100)) +00 47 +29 35 +34 14 +36 69 +37 108 +40 16 +41 11 +50 9 +53 1 +58 1 +65 1 +68 1 +select companynr,count(distinct concat(repeat(65,1000),floor(fld1/100))) from t2 group by companynr; +companynr count(distinct concat(repeat(65,1000),floor(fld1/100))) +00 47 +29 35 +34 14 +36 69 +37 108 +40 16 +41 11 +50 9 +53 1 +58 1 +65 1 +68 1 +select sum(fld1),fld3 from t2 where fld3="Romans" group by fld1 limit 10; +sum(fld1) fld3 +11402 Romans +select name,count(*) from t3 where name='cloakroom' group by name; +name count(*) +cloakroom 4181 +select name,count(*) from t3 where name='cloakroom' and price>10 group by name; +name count(*) +cloakroom 4181 +select count(*) from t3 where name='cloakroom' and price2=823742; +count(*) +4181 +select name,count(*) from t3 where name='cloakroom' and price2=823742 group by name; +name count(*) +cloakroom 4181 +select name,count(*) from t3 where name >= "extramarital" and price <= 39654943 group by name; +name count(*) +extramarital 4181 +gazer 4181 +gems 4181 +Iranizes 4181 +spates 4181 +tucked 4181 +violinist 4181 +select t2.fld3,count(*) from t2,t3 where t2.fld1=158402 and t3.name=t2.fld3 group by t3.name; +fld3 count(*) +spates 4181 +select companynr|0,companyname from t4 group by 1; +companynr|0 companyname +0 Unknown +29 company 1 +34 company 2 +36 company 3 +37 company 4 +40 company 5 +41 company 6 +50 company 11 +53 company 7 +58 company 8 +65 company 9 +68 company 10 +select t2.companynr,companyname,count(*) from t2,t4 where t2.companynr=t4.companynr group by t2.companynr order by companyname; +companynr companyname count(*) +29 company 1 95 +68 company 10 12 +50 company 11 11 +34 company 2 70 +36 company 3 215 +37 company 4 588 +40 company 5 37 +41 company 6 52 +53 company 7 4 +58 company 8 23 +65 company 9 10 +00 Unknown 82 +select t2.fld1,count(*) from t2,t3 where t2.fld1=158402 and t3.name=t2.fld3 group by t3.name; +fld1 count(*) +158402 4181 +select sum(Period)/count(*) from t1; +sum(Period)/count(*) +9410.0000 +select companynr,count(price) as "count",sum(price) as "sum" ,abs(sum(price)/count(price)-avg(price)) as "diff",(0+count(price))*companynr as func from t3 group by companynr; +companynr count sum diff func +37 12543 309394878010 0.0000 464091 +78 8362 414611089292 0.0000 652236 +101 4181 3489454238 0.0000 422281 +154 4181 4112197254950 0.0000 643874 +311 4181 979599938 0.0000 1300291 +447 4181 9929180954 0.0000 1868907 +512 4181 3288532102 0.0000 2140672 +select companynr,sum(price)/count(price) as avg from t3 group by companynr having avg > 70000000 order by avg; +companynr avg +154 983543950.0000 +select companynr,count(*) from t2 group by companynr order by 2 desc; +companynr count(*) +37 588 +36 215 +29 95 +00 82 +34 70 +41 52 +40 37 +58 23 +68 12 +50 11 +65 10 +53 4 +select companynr,count(*) from t2 where companynr > 40 group by companynr order by 2 desc; +companynr count(*) +41 52 +58 23 +68 12 +50 11 +65 10 +53 4 +select t2.fld4,t2.fld1,count(price),sum(price),min(price),max(price),avg(price) from t3,t2 where t3.companynr = 37 and t2.fld1 = t3.t2nr group by fld1,t2.fld4; +fld4 fld1 count(price) sum(price) min(price) max(price) avg(price) +teethe 000001 1 5987435 5987435 5987435 5987435.0000 +dreaded 011401 1 5987435 5987435 5987435 5987435.0000 +scholastics 011402 1 28357832 28357832 28357832 28357832.0000 +audiology 011403 1 39654943 39654943 39654943 39654943.0000 +wallet 011501 1 5987435 5987435 5987435 5987435.0000 +parters 011701 1 5987435 5987435 5987435 5987435.0000 +eschew 011702 1 28357832 28357832 28357832 28357832.0000 +quitter 011703 1 39654943 39654943 39654943 39654943.0000 +neat 012001 1 5987435 5987435 5987435 5987435.0000 +Steinberg 012003 1 39654943 39654943 39654943 39654943.0000 +balled 012301 1 5987435 5987435 5987435 5987435.0000 +persist 012302 1 28357832 28357832 28357832 28357832.0000 +attainments 012303 1 39654943 39654943 39654943 39654943.0000 +capably 012501 1 5987435 5987435 5987435 5987435.0000 +impulsive 012602 1 28357832 28357832 28357832 28357832.0000 +starlet 012603 1 39654943 39654943 39654943 39654943.0000 +featherweight 012701 1 5987435 5987435 5987435 5987435.0000 +pessimist 012702 1 28357832 28357832 28357832 28357832.0000 +daughter 012703 1 39654943 39654943 39654943 39654943.0000 +lawgiver 013601 1 5987435 5987435 5987435 5987435.0000 +stated 013602 1 28357832 28357832 28357832 28357832.0000 +readable 013603 1 39654943 39654943 39654943 39654943.0000 +testicle 013801 1 5987435 5987435 5987435 5987435.0000 +Parsifal 013802 1 28357832 28357832 28357832 28357832.0000 +leavings 013803 1 39654943 39654943 39654943 39654943.0000 +squeaking 013901 1 5987435 5987435 5987435 5987435.0000 +contrasted 016001 1 5987435 5987435 5987435 5987435.0000 +leftover 016201 1 5987435 5987435 5987435 5987435.0000 +whiteners 016202 1 28357832 28357832 28357832 28357832.0000 +erases 016301 1 5987435 5987435 5987435 5987435.0000 +Punjab 016302 1 28357832 28357832 28357832 28357832.0000 +Merritt 016303 1 39654943 39654943 39654943 39654943.0000 +sweetish 018001 1 5987435 5987435 5987435 5987435.0000 +dogging 018002 1 28357832 28357832 28357832 28357832.0000 +scornfully 018003 1 39654943 39654943 39654943 39654943.0000 +fetters 018012 1 28357832 28357832 28357832 28357832.0000 +bivalves 018013 1 39654943 39654943 39654943 39654943.0000 +skulking 018021 1 5987435 5987435 5987435 5987435.0000 +flint 018022 1 28357832 28357832 28357832 28357832.0000 +flopping 018023 1 39654943 39654943 39654943 39654943.0000 +Judas 018032 1 28357832 28357832 28357832 28357832.0000 +vacuuming 018033 1 39654943 39654943 39654943 39654943.0000 +medical 018041 1 5987435 5987435 5987435 5987435.0000 +bloodbath 018042 1 28357832 28357832 28357832 28357832.0000 +subschema 018043 1 39654943 39654943 39654943 39654943.0000 +interdependent 018051 1 5987435 5987435 5987435 5987435.0000 +Graves 018052 1 28357832 28357832 28357832 28357832.0000 +neonatal 018053 1 39654943 39654943 39654943 39654943.0000 +sorters 018061 1 5987435 5987435 5987435 5987435.0000 +epistle 018062 1 28357832 28357832 28357832 28357832.0000 +Conley 018101 1 5987435 5987435 5987435 5987435.0000 +lectured 018102 1 28357832 28357832 28357832 28357832.0000 +Abraham 018103 1 39654943 39654943 39654943 39654943.0000 +cage 018201 1 5987435 5987435 5987435 5987435.0000 +hushes 018202 1 28357832 28357832 28357832 28357832.0000 +Simla 018402 1 28357832 28357832 28357832 28357832.0000 +reporters 018403 1 39654943 39654943 39654943 39654943.0000 +coexist 018601 1 5987435 5987435 5987435 5987435.0000 +Beebe 018602 1 28357832 28357832 28357832 28357832.0000 +Taoism 018603 1 39654943 39654943 39654943 39654943.0000 +Connally 018801 1 5987435 5987435 5987435 5987435.0000 +fetched 018802 1 28357832 28357832 28357832 28357832.0000 +checkpoints 018803 1 39654943 39654943 39654943 39654943.0000 +gritty 018811 1 5987435 5987435 5987435 5987435.0000 +firearm 018812 1 28357832 28357832 28357832 28357832.0000 +minima 019101 1 5987435 5987435 5987435 5987435.0000 +Selfridge 019102 1 28357832 28357832 28357832 28357832.0000 +disable 019103 1 39654943 39654943 39654943 39654943.0000 +witchcraft 019201 1 5987435 5987435 5987435 5987435.0000 +betroth 030501 1 5987435 5987435 5987435 5987435.0000 +Manhattanize 030502 1 28357832 28357832 28357832 28357832.0000 +imprint 030503 1 39654943 39654943 39654943 39654943.0000 +swelling 031901 1 5987435 5987435 5987435 5987435.0000 +interrelationships 036001 1 5987435 5987435 5987435 5987435.0000 +riser 036002 1 28357832 28357832 28357832 28357832.0000 +bee 038001 1 5987435 5987435 5987435 5987435.0000 +kanji 038002 1 28357832 28357832 28357832 28357832.0000 +dental 038003 1 39654943 39654943 39654943 39654943.0000 +railway 038011 1 5987435 5987435 5987435 5987435.0000 +validate 038012 1 28357832 28357832 28357832 28357832.0000 +normalizes 038013 1 39654943 39654943 39654943 39654943.0000 +Kline 038101 1 5987435 5987435 5987435 5987435.0000 +Anatole 038102 1 28357832 28357832 28357832 28357832.0000 +partridges 038103 1 39654943 39654943 39654943 39654943.0000 +recruited 038201 1 5987435 5987435 5987435 5987435.0000 +dimensions 038202 1 28357832 28357832 28357832 28357832.0000 +Chicana 038203 1 39654943 39654943 39654943 39654943.0000 +select t3.companynr,fld3,sum(price) from t3,t2 where t2.fld1 = t3.t2nr and t3.companynr = 512 group by companynr,fld3; +companynr fld3 sum(price) +512 boat 786542 +512 capably 786542 +512 cupboard 786542 +512 decliner 786542 +512 descendants 786542 +512 dopers 786542 +512 erases 786542 +512 Micronesia 786542 +512 Miles 786542 +512 skies 786542 +select t2.companynr,count(*),min(fld3),max(fld3),sum(price),avg(price) from t2,t3 where t3.companynr >= 30 and t3.companynr <= 58 and t3.t2nr = t2.fld1 and 1+1=2 group by t2.companynr; +companynr count(*) min(fld3) max(fld3) sum(price) avg(price) +00 1 Omaha Omaha 5987435 5987435.0000 +36 1 dubbed dubbed 28357832 28357832.0000 +37 83 Abraham Wotan 1908978016 22999735.1325 +50 2 scribbled tapestry 68012775 34006387.5000 +select t3.companynr+0,t3.t2nr,fld3,sum(price) from t3,t2 where t2.fld1 = t3.t2nr and t3.companynr = 37 group by 1,t3.t2nr,fld3,fld3,fld3,fld3,fld3 order by fld1; +t3.companynr+0 t2nr fld3 sum(price) +37 1 Omaha 5987435 +37 11401 breaking 5987435 +37 11402 Romans 28357832 +37 11403 intercepted 39654943 +37 11501 bewilderingly 5987435 +37 11701 astound 5987435 +37 11702 admonishing 28357832 +37 11703 sumac 39654943 +37 12001 flanking 5987435 +37 12003 combed 39654943 +37 12301 Eulerian 5987435 +37 12302 dubbed 28357832 +37 12303 Kane 39654943 +37 12501 annihilates 5987435 +37 12602 Wotan 28357832 +37 12603 snatching 39654943 +37 12701 grazing 5987435 +37 12702 Baird 28357832 +37 12703 celery 39654943 +37 13601 handgun 5987435 +37 13602 foldout 28357832 +37 13603 mystic 39654943 +37 13801 intelligibility 5987435 +37 13802 Augustine 28357832 +37 13803 teethe 39654943 +37 13901 scholastics 5987435 +37 16001 audiology 5987435 +37 16201 wallet 5987435 +37 16202 parters 28357832 +37 16301 eschew 5987435 +37 16302 quitter 28357832 +37 16303 neat 39654943 +37 18001 jarring 5987435 +37 18002 tinily 28357832 +37 18003 balled 39654943 +37 18012 impulsive 28357832 +37 18013 starlet 39654943 +37 18021 lawgiver 5987435 +37 18022 stated 28357832 +37 18023 readable 39654943 +37 18032 testicle 28357832 +37 18033 Parsifal 39654943 +37 18041 Punjab 5987435 +37 18042 Merritt 28357832 +37 18043 Quixotism 39654943 +37 18051 sureties 5987435 +37 18052 puddings 28357832 +37 18053 tapestry 39654943 +37 18061 trimmings 5987435 +37 18062 humility 28357832 +37 18101 tragedies 5987435 +37 18102 skulking 28357832 +37 18103 flint 39654943 +37 18201 relaxing 5987435 +37 18202 offload 28357832 +37 18402 suites 28357832 +37 18403 lists 39654943 +37 18601 vacuuming 5987435 +37 18602 dentally 28357832 +37 18603 humanness 39654943 +37 18801 inch 5987435 +37 18802 Weissmuller 28357832 +37 18803 irresponsibly 39654943 +37 18811 repetitions 5987435 +37 18812 Antares 28357832 +37 19101 ventilate 5987435 +37 19102 pityingly 28357832 +37 19103 interdependent 39654943 +37 19201 Graves 5987435 +37 30501 neonatal 5987435 +37 30502 scribbled 28357832 +37 30503 chafe 39654943 +37 31901 realtor 5987435 +37 36001 elite 5987435 +37 36002 funereal 28357832 +37 38001 Conley 5987435 +37 38002 lectured 28357832 +37 38003 Abraham 39654943 +37 38011 groupings 5987435 +37 38012 dissociate 28357832 +37 38013 coexist 39654943 +37 38101 rusting 5987435 +37 38102 galling 28357832 +37 38103 obliterates 39654943 +37 38201 resumes 5987435 +37 38202 analyzable 28357832 +37 38203 terminator 39654943 +select sum(price) from t3,t2 where t2.fld1 = t3.t2nr and t3.companynr = 512 and t3.t2nr = 38008 and t2.fld1 = 38008 or t2.fld1= t3.t2nr and t3.t2nr = 38008 and t2.fld1 = 38008; +sum(price) +234298 +select t2.fld1,sum(price) from t3,t2 where t2.fld1 = t3.t2nr and t3.companynr = 512 and t3.t2nr = 38008 and t2.fld1 = 38008 or t2.fld1 = t3.t2nr and t3.t2nr = 38008 and t2.fld1 = 38008 or t3.t2nr = t2.fld1 and t2.fld1 = 38008 group by t2.fld1; +fld1 sum(price) +038008 234298 +explain select fld3 from t2 where 1>2 or 2>3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Impossible WHERE +explain select fld3 from t2 where fld1=fld1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 1199 +select companynr,fld1 from t2 HAVING fld1=250501 or fld1=250502; +companynr fld1 +34 250501 +34 250502 +select companynr,fld1 from t2 WHERE fld1>=250501 HAVING fld1<=250502; +companynr fld1 +34 250501 +34 250502 +select companynr,count(*) as count,sum(fld1) as sum from t2 group by companynr having count > 40 and sum/count >= 120000; +companynr count sum +00 82 10355753 +29 95 14473298 +34 70 17788966 +37 588 83602098 +41 52 12816335 +select companynr from t2 group by companynr having count(*) > 40 and sum(fld1)/count(*) >= 120000 ; +companynr +00 +29 +34 +37 +41 +select t2.companynr,companyname,count(*) from t2,t4 where t2.companynr=t4.companynr group by companyname having t2.companynr >= 40; +companynr companyname count(*) +68 company 10 12 +50 company 11 11 +40 company 5 37 +41 company 6 52 +53 company 7 4 +58 company 8 23 +65 company 9 10 +select count(*) from t2; +count(*) +1199 +select count(*) from t2 where fld1 < 098024; +count(*) +387 +select min(fld1) from t2 where fld1>= 098024; +min(fld1) +98024 +select max(fld1) from t2 where fld1>= 098024; +max(fld1) +1232609 +select count(*) from t3 where price2=76234234; +count(*) +4181 +select count(*) from t3 where companynr=512 and price2=76234234; +count(*) +4181 +explain select min(fld1),max(fld1),count(*) from t2; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +select min(fld1),max(fld1),count(*) from t2; +min(fld1) max(fld1) count(*) +0 1232609 1199 +select min(t2nr),max(t2nr) from t3 where t2nr=2115 and price2=823742; +min(t2nr) max(t2nr) +2115 2115 +select count(*),min(t2nr),max(t2nr) from t3 where name='spates' and companynr=78; +count(*) min(t2nr) max(t2nr) +4181 4 41804 +select t2nr,count(*) from t3 where name='gems' group by t2nr limit 20; +t2nr count(*) +9 1 +19 1 +29 1 +39 1 +49 1 +59 1 +69 1 +79 1 +89 1 +99 1 +109 1 +119 1 +129 1 +139 1 +149 1 +159 1 +169 1 +179 1 +189 1 +199 1 +select max(t2nr) from t3 where price=983543950; +max(t2nr) +41807 +select t1.period from t3 = t1 limit 1; +period +1001 +select t1.period from t1 as t1 limit 1; +period +9410 +select t1.period as "Nuvarande period" from t1 as t1 limit 1; +Nuvarande period +9410 +select period as ok_period from t1 limit 1; +ok_period +9410 +select period as ok_period from t1 group by ok_period limit 1; +ok_period +9410 +select 1+1 as summa from t1 group by summa limit 1; +summa +2 +select period as "Nuvarande period" from t1 group by "Nuvarande period" limit 1; +Nuvarande period +9410 +show tables; +Tables_in_test +t1 +t2 +t3 +t4 +show tables from test like "s%"; +Tables_in_test (s%) +show tables from test like "t?"; +Tables_in_test (t?) +show full columns from t2; +Field Type Collation Null Key Default Extra Privileges Comment +auto int(11) NULL NO PRI NULL auto_increment # +fld1 int(6) unsigned zerofill NULL NO UNI 000000 # +companynr tinyint(2) unsigned zerofill NULL NO 00 # +fld3 char(30) latin1_swedish_ci NO MUL # +fld4 char(35) latin1_swedish_ci NO # +fld5 char(35) latin1_swedish_ci NO # +fld6 char(4) latin1_swedish_ci NO # +show full columns from t2 from test like 'f%'; +Field Type Collation Null Key Default Extra Privileges Comment +fld1 int(6) unsigned zerofill NULL NO UNI 000000 # +fld3 char(30) latin1_swedish_ci NO MUL # +fld4 char(35) latin1_swedish_ci NO # +fld5 char(35) latin1_swedish_ci NO # +fld6 char(4) latin1_swedish_ci NO # +show full columns from t2 from test like 's%'; +Field Type Collation Null Key Default Extra Privileges Comment +show keys from t2; +Table Non_unique Key_name Seq_in_index Column_name Collation Cardinality Sub_part Packed Null Index_type Comment +t2 0 PRIMARY 1 auto A 1199 NULL NULL BTREE +t2 0 fld1 1 fld1 A 1199 NULL NULL BTREE +t2 1 fld3 1 fld3 A NULL NULL NULL BTREE +drop table t4, t3, t2, t1; +DO 1; +DO benchmark(100,1+1),1,1; +do default; +ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 1 +do foobar; +ERROR 42S22: Unknown column 'foobar' in 'field list' +CREATE TABLE t1 ( +id mediumint(8) unsigned NOT NULL auto_increment, +pseudo varchar(35) NOT NULL default '', +PRIMARY KEY (id), +UNIQUE KEY pseudo (pseudo) +); +INSERT INTO t1 (pseudo) VALUES ('test'); +INSERT INTO t1 (pseudo) VALUES ('test1'); +SELECT 1 as rnd1 from t1 where rand() > 2; +rnd1 +DROP TABLE t1; +CREATE TABLE t1 (gvid int(10) unsigned default NULL, hmid int(10) unsigned default NULL, volid int(10) unsigned default NULL, mmid int(10) unsigned default NULL, hdid int(10) unsigned default NULL, fsid int(10) unsigned default NULL, ctid int(10) unsigned default NULL, dtid int(10) unsigned default NULL, cost int(10) unsigned default NULL, performance int(10) unsigned default NULL, serialnumber bigint(20) unsigned default NULL, monitored tinyint(3) unsigned default '1', removed tinyint(3) unsigned default '0', target tinyint(3) unsigned default '0', dt_modified timestamp NOT NULL, name varchar(255) binary default NULL, description varchar(255) default NULL, UNIQUE KEY hmid (hmid,volid)) ENGINE=MyISAM; +INSERT INTO t1 VALUES (200001,2,1,1,100,1,1,1,0,0,0,1,0,1,20020425060057,'\\\\ARKIVIO-TESTPDC\\E$',''),(200002,2,2,1,101,1,1,1,0,0,0,1,0,1,20020425060057,'\\\\ARKIVIO-TESTPDC\\C$',''),(200003,1,3,2,NULL,NULL,NULL,NULL,NULL,NULL,NULL,1,0,1,20020425060427,'c:',NULL); +CREATE TABLE t2 ( hmid int(10) unsigned default NULL, volid int(10) unsigned default NULL, sampletid smallint(5) unsigned default NULL, sampletime datetime default NULL, samplevalue bigint(20) unsigned default NULL, KEY idx1 (hmid,volid,sampletid,sampletime)) ENGINE=MyISAM; +INSERT INTO t2 VALUES (1,3,10,'2002-06-01 08:00:00',35),(1,3,1010,'2002-06-01 12:00:01',35); +SELECT a.gvid, (SUM(CASE b.sampletid WHEN 140 THEN b.samplevalue ELSE 0 END)) as the_success,(SUM(CASE b.sampletid WHEN 141 THEN b.samplevalue ELSE 0 END)) as the_fail,(SUM(CASE b.sampletid WHEN 142 THEN b.samplevalue ELSE 0 END)) as the_size,(SUM(CASE b.sampletid WHEN 143 THEN b.samplevalue ELSE 0 END)) as the_time FROM t1 a, t2 b WHERE a.hmid = b.hmid AND a.volid = b.volid AND b.sampletime >= 'wrong-date-value' AND b.sampletime < 'wrong-date-value' AND b.sampletid IN (140, 141, 142, 143) GROUP BY a.gvid; +gvid the_success the_fail the_size the_time +Warnings: +Warning 1292 Incorrect datetime value: 'wrong-date-value' for column 'sampletime' at row 1 +Warning 1292 Incorrect datetime value: 'wrong-date-value' for column 'sampletime' at row 1 +SELECT a.gvid, (SUM(CASE b.sampletid WHEN 140 THEN b.samplevalue ELSE 0 END)) as the_success,(SUM(CASE b.sampletid WHEN 141 THEN b.samplevalue ELSE 0 END)) as the_fail,(SUM(CASE b.sampletid WHEN 142 THEN b.samplevalue ELSE 0 END)) as the_size,(SUM(CASE b.sampletid WHEN 143 THEN b.samplevalue ELSE 0 END)) as the_time FROM t1 a, t2 b WHERE a.hmid = b.hmid AND a.volid = b.volid AND b.sampletime >= NULL AND b.sampletime < NULL AND b.sampletid IN (140, 141, 142, 143) GROUP BY a.gvid; +gvid the_success the_fail the_size the_time +DROP TABLE t1,t2; +create table t1 ( A_Id bigint(20) NOT NULL default '0', A_UpdateBy char(10) NOT NULL default '', A_UpdateDate bigint(20) NOT NULL default '0', A_UpdateSerial int(11) NOT NULL default '0', other_types bigint(20) NOT NULL default '0', wss_type bigint(20) NOT NULL default '0'); +INSERT INTO t1 VALUES (102935998719055004,'brade',1029359987,2,102935229116544068,102935229216544093); +select wss_type from t1 where wss_type ='102935229216544106'; +wss_type +select wss_type from t1 where wss_type ='102935229216544105'; +wss_type +select wss_type from t1 where wss_type ='102935229216544104'; +wss_type +select wss_type from t1 where wss_type ='102935229216544093'; +wss_type +102935229216544093 +select wss_type from t1 where wss_type =102935229216544093; +wss_type +102935229216544093 +drop table t1; +select 1+2,"aaaa",3.13*2.0 into @a,@b,@c; +select @a; +@a +3 +select @b; +@b +aaaa +select @c; +@c +6.260 +create table t1 (a int not null auto_increment primary key); +insert into t1 values (); +insert into t1 values (); +insert into t1 values (); +select * from (t1 as t2 left join t1 as t3 using (a)), t1; +a a +1 1 +2 1 +3 1 +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from t1, (t1 as t2 left join t1 as t3 using (a)); +a a +1 1 +2 1 +3 1 +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from (t1 as t2 left join t1 as t3 using (a)) straight_join t1; +a a +1 1 +2 1 +3 1 +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from t1 straight_join (t1 as t2 left join t1 as t3 using (a)); +a a +1 1 +2 1 +3 1 +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from (t1 as t2 left join t1 as t3 using (a)) inner join t1 on t1.a>1; +a a +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from t1 inner join (t1 as t2 left join t1 as t3 using (a)) on t1.a>1; +a a +2 1 +3 1 +2 2 +3 2 +2 3 +3 3 +select * from (t1 as t2 left join t1 as t3 using (a)) inner join t1 using ( a ); +a +1 +2 +3 +select * from t1 inner join (t1 as t2 left join t1 as t3 using (a)) using ( a ); +a +1 +2 +3 +select * from (t1 as t2 left join t1 as t3 using (a)) left outer join t1 on t1.a>1; +a a +1 2 +1 3 +2 2 +2 3 +3 2 +3 3 +select * from t1 left outer join (t1 as t2 left join t1 as t3 using (a)) on t1.a>1; +a a +1 NULL +2 1 +2 2 +2 3 +3 1 +3 2 +3 3 +select * from (t1 as t2 left join t1 as t3 using (a)) left join t1 using ( a ); +a +1 +2 +3 +select * from t1 left join (t1 as t2 left join t1 as t3 using (a)) using ( a ); +a +1 +2 +3 +select * from (t1 as t2 left join t1 as t3 using (a)) natural left join t1; +a +1 +2 +3 +select * from t1 natural left join (t1 as t2 left join t1 as t3 using (a)); +a +1 +2 +3 +select * from (t1 as t2 left join t1 as t3 using (a)) right join t1 on t1.a>1; +a a +NULL 1 +1 2 +2 2 +3 2 +1 3 +2 3 +3 3 +select * from t1 right join (t1 as t2 left join t1 as t3 using (a)) on t1.a>1; +a a +2 1 +3 1 +2 2 +3 2 +2 3 +3 3 +select * from (t1 as t2 left join t1 as t3 using (a)) right outer join t1 using ( a ); +a +1 +2 +3 +select * from t1 right outer join (t1 as t2 left join t1 as t3 using (a)) using ( a ); +a +1 +2 +3 +select * from (t1 as t2 left join t1 as t3 using (a)) natural right join t1; +a +1 +2 +3 +select * from t1 natural right join (t1 as t2 left join t1 as t3 using (a)); +a +1 +2 +3 +select * from t1 natural join (t1 as t2 left join t1 as t3 using (a)); +a +1 +2 +3 +select * from (t1 as t2 left join t1 as t3 using (a)) natural join t1; +a +1 +2 +3 +drop table t1; +CREATE TABLE t1 ( aa char(2), id int(11) NOT NULL auto_increment, t2_id int(11) NOT NULL default '0', PRIMARY KEY (id), KEY replace_id (t2_id)) ENGINE=MyISAM; +INSERT INTO t1 VALUES ("1",8264,2506),("2",8299,2517),("3",8301,2518),("4",8302,2519),("5",8303,2520),("6",8304,2521),("7",8305,2522); +CREATE TABLE t2 ( id int(11) NOT NULL auto_increment, PRIMARY KEY (id)) ENGINE=MyISAM; +INSERT INTO t2 VALUES (2517), (2518), (2519), (2520), (2521), (2522); +select * from t1, t2 WHERE t1.t2_id = t2.id and t1.t2_id > 0 order by t1.id LIMIT 0, 5; +aa id t2_id id +2 8299 2517 2517 +3 8301 2518 2518 +4 8302 2519 2519 +5 8303 2520 2520 +6 8304 2521 2521 +drop table t1,t2; +create table t1 (id1 int NOT NULL); +create table t2 (id2 int NOT NULL); +create table t3 (id3 int NOT NULL); +create table t4 (id4 int NOT NULL, id44 int NOT NULL, KEY (id4)); +insert into t1 values (1); +insert into t1 values (2); +insert into t2 values (1); +insert into t4 values (1,1); +explain select * from t1 left join t2 on id1 = id2 left join t3 on id1 = id3 +left join t4 on id3 = id4 where id2 = 1 or id4 = 1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t3 system NULL NULL NULL NULL 0 const row not found +1 SIMPLE t4 const id4 NULL NULL NULL 1 +1 SIMPLE t1 ALL NULL NULL NULL NULL 2 +1 SIMPLE t2 ALL NULL NULL NULL NULL 1 Using where +select * from t1 left join t2 on id1 = id2 left join t3 on id1 = id3 +left join t4 on id3 = id4 where id2 = 1 or id4 = 1; +id1 id2 id3 id4 id44 +1 1 NULL NULL NULL +drop table t1,t2,t3,t4; +create table t1(s varchar(10) not null); +create table t2(s varchar(10) not null primary key); +create table t3(s varchar(10) not null primary key); +insert into t1 values ('one\t'), ('two\t'); +insert into t2 values ('one\r'), ('two\t'); +insert into t3 values ('one '), ('two\t'); +select * from t1 where s = 'one'; +s +select * from t2 where s = 'one'; +s +select * from t3 where s = 'one'; +s +one +select * from t1,t2 where t1.s = t2.s; +s s +two two +select * from t2,t3 where t2.s = t3.s; +s s +two two +drop table t1, t2, t3; +create table t1 (a integer, b integer, index(a), index(b)); +create table t2 (c integer, d integer, index(c), index(d)); +insert into t1 values (1,2), (2,2), (3,2), (4,2); +insert into t2 values (1,3), (2,3), (3,4), (4,4); +explain select * from t1 left join t2 on a=c where d in (4); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref c,d d 5 const 2 Using where +1 SIMPLE t1 ALL a NULL NULL NULL 4 Using where; Using join buffer +select * from t1 left join t2 on a=c where d in (4); +a b c d +3 2 3 4 +4 2 4 4 +explain select * from t1 left join t2 on a=c where d = 4; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref c,d d 5 const 2 Using where +1 SIMPLE t1 ALL a NULL NULL NULL 4 Using where; Using join buffer +select * from t1 left join t2 on a=c where d = 4; +a b c d +3 2 3 4 +4 2 4 4 +drop table t1, t2; +CREATE TABLE t1 ( +i int(11) NOT NULL default '0', +c char(10) NOT NULL default '', +PRIMARY KEY (i), +UNIQUE KEY c (c) +) ENGINE=MyISAM; +INSERT INTO t1 VALUES (1,'a'); +INSERT INTO t1 VALUES (2,'b'); +INSERT INTO t1 VALUES (3,'c'); +EXPLAIN SELECT i FROM t1 WHERE i=1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 Using index +DROP TABLE t1; +CREATE TABLE t1 ( a BLOB, INDEX (a(20)) ); +CREATE TABLE t2 ( a BLOB, INDEX (a(20)) ); +INSERT INTO t1 VALUES ('one'),('two'),('three'),('four'),('five'); +INSERT INTO t2 VALUES ('one'),('two'),('three'),('four'),('five'); +EXPLAIN SELECT * FROM t1 LEFT JOIN t2 USE INDEX (a) ON t1.a=t2.a; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 +1 SIMPLE t2 ref a a 23 test.t1.a 2 +EXPLAIN SELECT * FROM t1 LEFT JOIN t2 FORCE INDEX (a) ON t1.a=t2.a; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 +1 SIMPLE t2 ref a a 23 test.t1.a 2 +DROP TABLE t1, t2; +CREATE TABLE t1 ( city char(30) ); +INSERT INTO t1 VALUES ('London'); +INSERT INTO t1 VALUES ('Paris'); +SELECT * FROM t1 WHERE city='London'; +city +London +SELECT * FROM t1 WHERE city='london'; +city +London +EXPLAIN SELECT * FROM t1 WHERE city='London' AND city='london'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 2 Using where +SELECT * FROM t1 WHERE city='London' AND city='london'; +city +London +EXPLAIN SELECT * FROM t1 WHERE city LIKE '%london%' AND city='London'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 2 Using where +SELECT * FROM t1 WHERE city LIKE '%london%' AND city='London'; +city +London +DROP TABLE t1; +create table t1 (a int(11) unsigned, b int(11) unsigned); +insert into t1 values (1,0), (1,1), (1,2); +select a-b from t1 order by 1; +a-b +0 +1 +18446744073709551615 +select a-b , (a-b < 0) from t1 order by 1; +a-b (a-b < 0) +0 0 +1 0 +18446744073709551615 0 +select a-b as d, (a-b >= 0), b from t1 group by b having d >= 0; +d (a-b >= 0) b +1 1 0 +0 1 1 +18446744073709551615 1 2 +select cast((a - b) as unsigned) from t1 order by 1; +cast((a - b) as unsigned) +0 +1 +18446744073709551615 +drop table t1; +create table t1 (a int(11)); +select all all * from t1; +a +select distinct distinct * from t1; +a +select all distinct * from t1; +ERROR HY000: Incorrect usage of ALL and DISTINCT +select distinct all * from t1; +ERROR HY000: Incorrect usage of ALL and DISTINCT +drop table t1; +CREATE TABLE t1 ( +kunde_intern_id int(10) unsigned NOT NULL default '0', +kunde_id int(10) unsigned NOT NULL default '0', +FK_firma_id int(10) unsigned NOT NULL default '0', +aktuell enum('Ja','Nein') NOT NULL default 'Ja', +vorname varchar(128) NOT NULL default '', +nachname varchar(128) NOT NULL default '', +geloescht enum('Ja','Nein') NOT NULL default 'Nein', +firma varchar(128) NOT NULL default '' +); +INSERT INTO t1 VALUES +(3964,3051,1,'Ja','Vorname1','1Nachname','Nein','Print Schau XXXX'), +(3965,3051111,1,'Ja','Vorname1111','1111Nachname','Nein','Print Schau XXXX'); +SELECT kunde_id ,FK_firma_id ,aktuell, vorname, nachname, geloescht FROM t1 +WHERE +( +( +( '' != '' AND firma LIKE CONCAT('%', '', '%')) +OR +(vorname LIKE CONCAT('%', 'Vorname1', '%') AND +nachname LIKE CONCAT('%', '1Nachname', '%') AND +'Vorname1' != '' AND 'xxxx' != '') +) +AND +( +aktuell = 'Ja' AND geloescht = 'Nein' AND FK_firma_id = 2 +) +) +; +kunde_id FK_firma_id aktuell vorname nachname geloescht +SELECT kunde_id ,FK_firma_id ,aktuell, vorname, nachname, +geloescht FROM t1 +WHERE +( +( +aktuell = 'Ja' AND geloescht = 'Nein' AND FK_firma_id = 2 +) +AND +( +( '' != '' AND firma LIKE CONCAT('%', '', '%') ) +OR +( vorname LIKE CONCAT('%', 'Vorname1', '%') AND +nachname LIKE CONCAT('%', '1Nachname', '%') AND 'Vorname1' != '' AND +'xxxx' != '') +) +) +; +kunde_id FK_firma_id aktuell vorname nachname geloescht +SELECT COUNT(*) FROM t1 WHERE +( 0 OR (vorname LIKE '%Vorname1%' AND nachname LIKE '%1Nachname%' AND 1)) +AND FK_firma_id = 2; +COUNT(*) +0 +drop table t1; +CREATE TABLE t1 (b BIGINT(20) UNSIGNED NOT NULL, PRIMARY KEY (b)); +INSERT INTO t1 VALUES (0x8000000000000000); +SELECT b FROM t1 WHERE b=0x8000000000000000; +b +9223372036854775808 +DROP TABLE t1; +CREATE TABLE `t1` ( `gid` int(11) default NULL, `uid` int(11) default NULL); +CREATE TABLE `t2` ( `ident` int(11) default NULL, `level` char(16) default NULL); +INSERT INTO `t2` VALUES (0,'READ'); +CREATE TABLE `t3` ( `id` int(11) default NULL, `name` char(16) default NULL); +INSERT INTO `t3` VALUES (1,'fs'); +select * from t3 left join t1 on t3.id = t1.uid, t2 where t2.ident in (0, t1.gid, t3.id, 0); +id name gid uid ident level +1 fs NULL NULL 0 READ +drop table t1,t2,t3; +CREATE TABLE t1 ( +acct_id int(11) NOT NULL default '0', +profile_id smallint(6) default NULL, +UNIQUE KEY t1$acct_id (acct_id), +KEY t1$profile_id (profile_id) +); +INSERT INTO t1 VALUES (132,17),(133,18); +CREATE TABLE t2 ( +profile_id smallint(6) default NULL, +queue_id int(11) default NULL, +seq int(11) default NULL, +KEY t2$queue_id (queue_id) +); +INSERT INTO t2 VALUES (17,31,4),(17,30,3),(17,36,2),(17,37,1); +CREATE TABLE t3 ( +id int(11) NOT NULL default '0', +qtype int(11) default NULL, +seq int(11) default NULL, +warn_lvl int(11) default NULL, +crit_lvl int(11) default NULL, +rr1 tinyint(4) NOT NULL default '0', +rr2 int(11) default NULL, +default_queue tinyint(4) NOT NULL default '0', +KEY t3$qtype (qtype), +KEY t3$id (id) +); +INSERT INTO t3 VALUES (30,1,29,NULL,NULL,0,NULL,0),(31,1,28,NULL,NULL,0,NULL,0), +(36,1,34,NULL,NULL,0,NULL,0),(37,1,35,NULL,NULL,0,121,0); +SELECT COUNT(*) FROM t1 a STRAIGHT_JOIN t2 pq STRAIGHT_JOIN t3 q +WHERE +(pq.profile_id = a.profile_id) AND (a.acct_id = 132) AND +(pq.queue_id = q.id) AND (q.rr1 <> 1); +COUNT(*) +4 +drop table t1,t2,t3; +create table t1 (f1 int); +insert into t1 values (1),(NULL); +create table t2 (f2 int, f3 int, f4 int); +create index idx1 on t2 (f4); +insert into t2 values (1,2,3),(2,4,6); +select A.f2 from t1 left join t2 A on A.f2 = f1 where A.f3=(select min(f3) +from t2 C where A.f4 = C.f4) or A.f3 IS NULL; +f2 +1 +NULL +drop table t1,t2; +create table t2 (a tinyint unsigned); +create index t2i on t2(a); +insert into t2 values (0), (254), (255); +explain select * from t2 where a > -1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 index t2i t2i 2 NULL 3 Using where; Using index +select * from t2 where a > -1; +a +0 +254 +255 +drop table t2; +CREATE TABLE t1 (a int, b int, c int); +INSERT INTO t1 +SELECT 50, 3, 3 FROM DUAL +WHERE NOT EXISTS +(SELECT * FROM t1 WHERE a = 50 AND b = 3); +SELECT * FROM t1; +a b c +50 3 3 +INSERT INTO t1 +SELECT 50, 3, 3 FROM DUAL +WHERE NOT EXISTS +(SELECT * FROM t1 WHERE a = 50 AND b = 3); +select found_rows(); +found_rows() +0 +SELECT * FROM t1; +a b c +50 3 3 +select count(*) from t1; +count(*) +1 +select found_rows(); +found_rows() +1 +select count(*) from t1 limit 2,3; +count(*) +select found_rows(); +found_rows() +0 +select SQL_CALC_FOUND_ROWS count(*) from t1 limit 2,3; +count(*) +select found_rows(); +found_rows() +1 +DROP TABLE t1; +CREATE TABLE t1 (a INT, b INT); +(SELECT a, b AS c FROM t1) ORDER BY c+1; +a c +(SELECT a, b AS c FROM t1) ORDER BY b+1; +a c +SELECT a, b AS c FROM t1 ORDER BY c+1; +a c +SELECT a, b AS c FROM t1 ORDER BY b+1; +a c +drop table t1; +create table t1(f1 int, f2 int); +create table t2(f3 int); +select f1 from t1,t2 where f1=f2 and (f1,f2) = ((1,1)); +f1 +select f1 from t1,t2 where f1=f2 and (f1,NULL) = ((1,1)); +f1 +select f1 from t1,t2 where f1=f2 and (f1,f2) = ((1,NULL)); +f1 +insert into t1 values(1,1),(2,null); +insert into t2 values(2); +select * from t1,t2 where f1=f3 and (f1,f2) = (2,null); +f1 f2 f3 +select * from t1,t2 where f1=f3 and (f1,f2) <=> (2,null); +f1 f2 f3 +2 NULL 2 +drop table t1,t2; +create table t1 (f1 int not null auto_increment primary key, f2 varchar(10)); +create table t11 like t1; +insert into t1 values(1,""),(2,""); +show table status like 't1%'; +Name Engine Version Row_format Rows Avg_row_length Data_length Max_data_length Index_length Data_free Auto_increment Create_time Update_time Check_time Collation Checksum Create_options Comment +t1 MyISAM 10 Dynamic 2 20 X X X X X X X X latin1_swedish_ci NULL +t11 MyISAM 10 Dynamic 0 0 X X X X X X X X latin1_swedish_ci NULL +select 123 as a from t1 where f1 is null; +a +drop table t1,t11; +CREATE TABLE t1 ( a INT NOT NULL, b INT NOT NULL, UNIQUE idx (a,b) ); +INSERT INTO t1 VALUES (1,1),(1,2),(1,3),(1,4); +CREATE TABLE t2 ( a INT NOT NULL, b INT NOT NULL, e INT ); +INSERT INTO t2 VALUES ( 1,10,1), (1,10,2), (1,11,1), (1,11,2), (1,2,1), (1,2,2),(1,2,3); +SELECT t2.a, t2.b, IF(t1.b IS NULL,'',e) AS c, COUNT(*) AS d FROM t2 LEFT JOIN +t1 ON t2.a = t1.a AND t2.b = t1.b GROUP BY a, b, c; +a b c d +1 2 1 1 +1 2 2 1 +1 2 3 1 +1 10 2 +1 11 2 +SELECT t2.a, t2.b, IF(t1.b IS NULL,'',e) AS c, COUNT(*) AS d FROM t2 LEFT JOIN +t1 ON t2.a = t1.a AND t2.b = t1.b GROUP BY t1.a, t1.b, c; +a b c d +1 10 4 +1 2 1 1 +1 2 2 1 +1 2 3 1 +SELECT t2.a, t2.b, IF(t1.b IS NULL,'',e) AS c, COUNT(*) AS d FROM t2 LEFT JOIN +t1 ON t2.a = t1.a AND t2.b = t1.b GROUP BY t2.a, t2.b, c; +a b c d +1 2 1 1 +1 2 2 1 +1 2 3 1 +1 10 2 +1 11 2 +SELECT t2.a, t2.b, IF(t1.b IS NULL,'',e) AS c, COUNT(*) AS d FROM t2,t1 +WHERE t2.a = t1.a AND t2.b = t1.b GROUP BY a, b, c; +a b c d +1 2 1 1 +1 2 2 1 +1 2 3 1 +DROP TABLE IF EXISTS t1, t2; +create table t1 (f1 int primary key, f2 int); +create table t2 (f3 int, f4 int, primary key(f3,f4)); +insert into t1 values (1,1); +insert into t2 values (1,1),(1,2); +select distinct count(f2) >0 from t1 left join t2 on f1=f3 group by f1; +count(f2) >0 +1 +drop table t1,t2; +create table t1 (f1 int,f2 int); +insert into t1 values(1,1); +create table t2 (f3 int, f4 int, primary key(f3,f4)); +insert into t2 values(1,1); +select * from t1 where f1 in (select f3 from t2 where (f3,f4)= (select f3,f4 from t2)); +f1 f2 +1 1 +drop table t1,t2; +CREATE TABLE t1(a int, b int, c int, KEY b(b), KEY c(c)); +insert into t1 values (1,0,0),(2,0,0); +CREATE TABLE t2 (a int, b varchar(2), c varchar(2), PRIMARY KEY(a)); +insert into t2 values (1,'',''), (2,'',''); +CREATE TABLE t3 (a int, b int, PRIMARY KEY (a,b), KEY a (a), KEY b (b)); +insert into t3 values (1,1),(1,2); +explain select straight_join DISTINCT t2.a,t2.b, t1.c from t1, t3, t2 +where (t1.c=t2.a or (t1.c=t3.a and t2.a=t3.b)) and t1.b=556476786 and +t2.b like '%%' order by t2.b limit 0,1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref b,c b 5 const 1 Using where; Using temporary; Using filesort +1 SIMPLE t3 index PRIMARY,a,b PRIMARY 8 NULL 2 Using index; Using join buffer +1 SIMPLE t2 ALL PRIMARY NULL NULL NULL 2 Range checked for each record (index map: 0x1) +DROP TABLE t1,t2,t3; +CREATE TABLE t1 (a int, INDEX idx(a)); +INSERT INTO t1 VALUES (2), (3), (1); +EXPLAIN SELECT * FROM t1 IGNORE INDEX (idx); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 3 +EXPLAIN SELECT * FROM t1 IGNORE INDEX (a); +ERROR 42000: Key 'a' doesn't exist in table 't1' +EXPLAIN SELECT * FROM t1 FORCE INDEX (a); +ERROR 42000: Key 'a' doesn't exist in table 't1' +DROP TABLE t1; +CREATE TABLE t1 (a int, b int); +INSERT INTO t1 VALUES (1,1), (2,1), (4,10); +CREATE TABLE t2 (a int PRIMARY KEY, b int, KEY b (b)); +INSERT INTO t2 VALUES (1,NULL), (2,10); +ALTER TABLE t1 ENABLE KEYS; +EXPLAIN SELECT STRAIGHT_JOIN SQL_NO_CACHE COUNT(*) FROM t2, t1 WHERE t1.b = t2.b OR t2.b IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 index b b 5 NULL 2 Using index +1 SIMPLE t1 ALL NULL NULL NULL NULL 3 Using where; Using join buffer +SELECT STRAIGHT_JOIN SQL_NO_CACHE * FROM t2, t1 WHERE t1.b = t2.b OR t2.b IS NULL; +a b a b +1 NULL 1 1 +1 NULL 2 1 +1 NULL 4 10 +2 10 4 10 +EXPLAIN SELECT STRAIGHT_JOIN SQL_NO_CACHE COUNT(*) FROM t2, t1 WHERE t1.b = t2.b OR t2.b IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 index b b 5 NULL 2 Using index +1 SIMPLE t1 ALL NULL NULL NULL NULL 3 Using where; Using join buffer +SELECT STRAIGHT_JOIN SQL_NO_CACHE * FROM t2, t1 WHERE t1.b = t2.b OR t2.b IS NULL; +a b a b +1 NULL 1 1 +1 NULL 2 1 +1 NULL 4 10 +2 10 4 10 +DROP TABLE IF EXISTS t1,t2; +CREATE TABLE t1 (key1 float default NULL, UNIQUE KEY key1 (key1)); +CREATE TABLE t2 (key2 float default NULL, UNIQUE KEY key2 (key2)); +INSERT INTO t1 VALUES (0.3762),(0.3845),(0.6158),(0.7941); +INSERT INTO t2 VALUES (1.3762),(1.3845),(1.6158),(1.7941); +explain select max(key1) from t1 where key1 <= 0.6158; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select max(key2) from t2 where key2 <= 1.6158; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select min(key1) from t1 where key1 >= 0.3762; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select min(key2) from t2 where key2 >= 1.3762; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select max(key1), min(key2) from t1, t2 +where key1 <= 0.6158 and key2 >= 1.3762; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select max(key1) from t1 where key1 <= 0.6158 and rand() + 0.5 >= 0.5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +explain select min(key1) from t1 where key1 >= 0.3762 and rand() + 0.5 >= 0.5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL Select tables optimized away +select max(key1) from t1 where key1 <= 0.6158; +max(key1) +0.615800023078918 +select max(key2) from t2 where key2 <= 1.6158; +max(key2) +1.61580002307892 +select min(key1) from t1 where key1 >= 0.3762; +min(key1) +0.376199990510941 +select min(key2) from t2 where key2 >= 1.3762; +min(key2) +1.37619996070862 +select max(key1), min(key2) from t1, t2 +where key1 <= 0.6158 and key2 >= 1.3762; +max(key1) min(key2) +0.615800023078918 1.37619996070862 +select max(key1) from t1 where key1 <= 0.6158 and rand() + 0.5 >= 0.5; +max(key1) +0.615800023078918 +select min(key1) from t1 where key1 >= 0.3762 and rand() + 0.5 >= 0.5; +min(key1) +0.376199990510941 +DROP TABLE t1,t2; +CREATE TABLE t1 (i BIGINT UNSIGNED NOT NULL); +INSERT INTO t1 VALUES (10); +SELECT i='1e+01',i=1e+01, i in (1e+01,1e+01), i in ('1e+01','1e+01') FROM t1; +i='1e+01' i=1e+01 i in (1e+01,1e+01) i in ('1e+01','1e+01') +1 1 1 1 +DROP TABLE t1; +create table t1(a bigint unsigned, b bigint); +insert into t1 values (0xfffffffffffffffff, 0xfffffffffffffffff), +(0x10000000000000000, 0x10000000000000000), +(0x8fffffffffffffff, 0x8fffffffffffffff); +Warnings: +Warning 1264 Out of range value for column 'a' at row 1 +Warning 1264 Out of range value for column 'b' at row 1 +Warning 1264 Out of range value for column 'a' at row 2 +Warning 1264 Out of range value for column 'b' at row 2 +Warning 1264 Out of range value for column 'b' at row 3 +select hex(a), hex(b) from t1; +hex(a) hex(b) +FFFFFFFFFFFFFFFF 7FFFFFFFFFFFFFFF +FFFFFFFFFFFFFFFF 7FFFFFFFFFFFFFFF +8FFFFFFFFFFFFFFF 7FFFFFFFFFFFFFFF +drop table t1; +CREATE TABLE t1 (c0 int); +CREATE TABLE t2 (c0 int); +INSERT INTO t1 VALUES(@@connect_timeout); +INSERT INTO t2 VALUES(@@connect_timeout); +SELECT * FROM t1 JOIN t2 ON t1.c0 = t2.c0 WHERE (t1.c0 <=> @@connect_timeout); +c0 c0 +X X +DROP TABLE t1, t2; +End of 4.1 tests +CREATE TABLE t1 ( +K2C4 varchar(4) character set latin1 collate latin1_bin NOT NULL default '', +K4N4 varchar(4) character set latin1 collate latin1_bin NOT NULL default '0000', +F2I4 int(11) NOT NULL default '0' +) ENGINE=MyISAM DEFAULT CHARSET=latin1; +INSERT INTO t1 VALUES +('W%RT', '0100', 1), +('W-RT', '0100', 1), +('WART', '0100', 1), +('WART', '0200', 1), +('WERT', '0100', 2), +('WORT','0200', 2), +('WT', '0100', 2), +('W_RT', '0100', 2), +('WaRT', '0100', 3), +('WART', '0300', 3), +('WRT' , '0400', 3), +('WURM', '0500', 3), +('W%T', '0600', 4), +('WA%T', '0700', 4), +('WA_T', '0800', 4); +SELECT K2C4, K4N4, F2I4 FROM t1 +WHERE K2C4 = 'WART' AND +(F2I4 = 2 AND K2C4 = 'WART' OR (F2I4 = 2 OR K4N4 = '0200')); +K2C4 K4N4 F2I4 +WART 0200 1 +SELECT K2C4, K4N4, F2I4 FROM t1 +WHERE K2C4 = 'WART' AND (K2C4 = 'WART' OR K4N4 = '0200'); +K2C4 K4N4 F2I4 +WART 0100 1 +WART 0200 1 +WART 0300 3 +DROP TABLE t1; +create table t1 (a int, b int); +create table t2 like t1; +select t1.a from (t1 inner join t2 on t1.a=t2.a) where t2.a=1; +a +select t1.a from ((t1 inner join t2 on t1.a=t2.a)) where t2.a=1; +a +select x.a, y.a, z.a from ( (t1 x inner join t2 y on x.a=y.a) inner join t2 z on y.a=z.a) WHERE x.a=1; +a a a +drop table t1,t2; +create table t1 (s1 varchar(5)); +insert into t1 values ('Wall'); +select min(s1) from t1 group by s1 with rollup; +min(s1) +Wall +Wall +drop table t1; +create table t1 (s1 int) engine=myisam; +insert into t1 values (0); +select avg(distinct s1) from t1 group by s1 with rollup; +avg(distinct s1) +0.0000 +0.0000 +drop table t1; +create table t1 (s1 int); +insert into t1 values (null),(1); +select distinct avg(s1) as x from t1 group by s1 with rollup; +x +NULL +1.0000 +drop table t1; +CREATE TABLE t1 (a int); +CREATE TABLE t2 (a int); +INSERT INTO t1 VALUES (1), (2), (3), (4), (5); +INSERT INTO t2 VALUES (2), (4), (6); +SELECT t1.a FROM t1 STRAIGHT_JOIN t2 ON t1.a=t2.a; +a +2 +4 +EXPLAIN SELECT t1.a FROM t1 STRAIGHT_JOIN t2 ON t1.a=t2.a; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 +1 SIMPLE t2 ALL NULL NULL NULL NULL 3 Using where; Using join buffer +EXPLAIN SELECT t1.a FROM t1 INNER JOIN t2 ON t1.a=t2.a; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ALL NULL NULL NULL NULL 3 +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 Using where; Using join buffer +DROP TABLE t1,t2; +select x'10' + 0, X'10' + 0, b'10' + 0, B'10' + 0; +x'10' + 0 X'10' + 0 b'10' + 0 B'10' + 0 +16 16 2 2 +create table t1 (f1 varchar(6) default NULL, f2 int(6) primary key not null); +create table t2 (f3 varchar(5) not null, f4 varchar(5) not null, UNIQUE KEY UKEY (f3,f4)); +insert into t1 values (" 2", 2); +insert into t2 values (" 2", " one "),(" 2", " two "); +select * from t1 left join t2 on f1 = f3; +f1 f2 f3 f4 + 2 2 2 one + 2 2 2 two +drop table t1,t2; +create table t1 (empnum smallint, grp int); +create table t2 (empnum int, name char(5)); +insert into t1 values(1,1); +insert into t2 values(1,'bob'); +create view v1 as select * from t2 inner join t1 using (empnum); +select * from v1; +empnum name grp +1 bob 1 +drop table t1,t2; +drop view v1; +create table t1 (pk int primary key, b int); +create table t2 (pk int primary key, c int); +select pk from t1 inner join t2 using (pk); +pk +drop table t1,t2; +create table t1 (s1 int, s2 char(5), s3 decimal(10)); +create view v1 as select s1, s2, 'x' as s3 from t1; +select * from t1 natural join v1; +s1 s2 s3 +insert into t1 values (1,'x',5); +select * from t1 natural join v1; +s1 s2 s3 +Warnings: +Warning 1292 Truncated incorrect DOUBLE value: 'x' +drop table t1; +drop view v1; +create table t1(a1 int); +create table t2(a2 int); +insert into t1 values(1),(2); +insert into t2 values(1),(2); +create view v2 (c) as select a1 from t1; +select * from t1 natural left join t2; +a1 a2 +1 1 +1 2 +2 1 +2 2 +select * from t1 natural right join t2; +a2 a1 +1 1 +1 2 +2 1 +2 2 +select * from v2 natural left join t2; +c a2 +1 1 +1 2 +2 1 +2 2 +select * from v2 natural right join t2; +a2 c +1 1 +1 2 +2 1 +2 2 +drop table t1, t2; +drop view v2; +create table t1 (a int(10), t1_val int(10)); +create table t2 (b int(10), t2_val int(10)); +create table t3 (a int(10), b int(10)); +insert into t1 values (1,1),(2,2); +insert into t2 values (1,1),(2,2),(3,3); +insert into t3 values (1,1),(2,1),(3,1),(4,1); +select * from t1 natural join t2 natural join t3; +a b t1_val t2_val +1 1 1 1 +2 1 2 1 +select * from t1 natural join t3 natural join t2; +b a t1_val t2_val +1 1 1 1 +1 2 2 1 +drop table t1, t2, t3; +DO IFNULL(NULL, NULL); +SELECT CAST(IFNULL(NULL, NULL) AS DECIMAL); +CAST(IFNULL(NULL, NULL) AS DECIMAL) +NULL +SELECT ABS(IFNULL(NULL, NULL)); +ABS(IFNULL(NULL, NULL)) +NULL +SELECT IFNULL(NULL, NULL); +IFNULL(NULL, NULL) +NULL +SET @OLD_SQL_MODE12595=@@SQL_MODE, @@SQL_MODE=''; +SHOW LOCAL VARIABLES LIKE 'SQL_MODE'; +Variable_name Value +sql_mode +CREATE TABLE BUG_12595(a varchar(100)); +INSERT INTO BUG_12595 VALUES ('hakan%'), ('hakank'), ("ha%an"); +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan\%'; +a +hakan% +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan*%' ESCAPE '*'; +a +hakan% +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan**%' ESCAPE '**'; +ERROR HY000: Incorrect arguments to ESCAPE +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan%' ESCAPE ''; +a +hakan% +hakank +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan\%' ESCAPE ''; +a +SELECT * FROM BUG_12595 WHERE a LIKE 'ha\%an' ESCAPE 0x5c; +a +ha%an +SELECT * FROM BUG_12595 WHERE a LIKE 'ha%%an' ESCAPE '%'; +a +ha%an +SELECT * FROM BUG_12595 WHERE a LIKE 'ha\%an' ESCAPE '\\'; +a +ha%an +SELECT * FROM BUG_12595 WHERE a LIKE 'ha|%an' ESCAPE '|'; +a +ha%an +SET @@SQL_MODE='NO_BACKSLASH_ESCAPES'; +SHOW LOCAL VARIABLES LIKE 'SQL_MODE'; +Variable_name Value +sql_mode NO_BACKSLASH_ESCAPES +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan\%'; +a +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan*%' ESCAPE '*'; +a +hakan% +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan**%' ESCAPE '**'; +ERROR HY000: Incorrect arguments to ESCAPE +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan\%' ESCAPE '\\'; +ERROR HY000: Incorrect arguments to ESCAPE +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan%' ESCAPE ''; +ERROR HY000: Incorrect arguments to ESCAPE +SELECT * FROM BUG_12595 WHERE a LIKE 'ha\%an' ESCAPE 0x5c; +a +ha%an +SELECT * FROM BUG_12595 WHERE a LIKE 'ha|%an' ESCAPE '|'; +a +ha%an +SELECT * FROM BUG_12595 WHERE a LIKE 'hakan\n%' ESCAPE '\n'; +ERROR HY000: Incorrect arguments to ESCAPE +SET @@SQL_MODE=@OLD_SQL_MODE12595; +DROP TABLE BUG_12595; +create table t1 (a char(1)); +create table t2 (a char(1)); +insert into t1 values ('a'),('b'),('c'); +insert into t2 values ('b'),('c'),('d'); +select a from t1 natural join t2; +a +b +c +select * from t1 natural join t2 where a = 'b'; +a +b +drop table t1, t2; +CREATE TABLE t1 (`id` TINYINT); +CREATE TABLE t2 (`id` TINYINT); +CREATE TABLE t3 (`id` TINYINT); +INSERT INTO t1 VALUES (1),(2),(3); +INSERT INTO t2 VALUES (2); +INSERT INTO t3 VALUES (3); +SELECT t1.id,t3.id FROM t1 JOIN t2 ON (t2.id=t1.id) LEFT JOIN t3 USING (id); +ERROR 23000: Column 'id' in from clause is ambiguous +SELECT t1.id,t3.id FROM t1 JOIN t2 ON (t2.notacolumn=t1.id) LEFT JOIN t3 USING (id); +ERROR 23000: Column 'id' in from clause is ambiguous +SELECT id,t3.id FROM t1 JOIN t2 ON (t2.id=t1.id) LEFT JOIN t3 USING (id); +ERROR 23000: Column 'id' in from clause is ambiguous +SELECT id,t3.id FROM (t1 JOIN t2 ON (t2.id=t1.id)) LEFT JOIN t3 USING (id); +ERROR 23000: Column 'id' in from clause is ambiguous +drop table t1, t2, t3; +create table t1 (a int(10),b int(10)); +create table t2 (a int(10),b int(10)); +insert into t1 values (1,10),(2,20),(3,30); +insert into t2 values (1,10); +select * from t1 inner join t2 using (A); +a b b +1 10 10 +select * from t1 inner join t2 using (a); +a b b +1 10 10 +drop table t1, t2; +create table t1 (a int, c int); +create table t2 (b int); +create table t3 (b int, a int); +create table t4 (c int); +insert into t1 values (1,1); +insert into t2 values (1); +insert into t3 values (1,1); +insert into t4 values (1); +select * from t1 join t2 join t3 on (t2.b = t3.b and t1.a = t3.a); +a c b b a +1 1 1 1 1 +select * from t1, t2 join t3 on (t2.b = t3.b and t1.a = t3.a); +ERROR 42S22: Unknown column 't1.a' in 'on clause' +select * from t1 join t2 join t3 join t4 on (t1.a = t4.c and t2.b = t4.c); +a c b b a c +1 1 1 1 1 1 +select * from t1 join t2 join t4 using (c); +c a b +1 1 1 +drop table t1, t2, t3, t4; +create table t1(x int, y int); +create table t2(x int, y int); +create table t3(x int, primary key(x)); +insert into t1 values (1, 1), (2, 1), (3, 1), (4, 3), (5, 6), (6, 6); +insert into t2 values (1, 1), (2, 1), (3, 3), (4, 6), (5, 6); +insert into t3 values (1), (2), (3), (4), (5); +select t1.x, t3.x from t1, t2, t3 where t1.x = t2.x and t3.x >= t1.y and t3.x <= t2.y; +x x +1 1 +2 1 +3 1 +3 2 +3 3 +4 3 +4 4 +4 5 +drop table t1,t2,t3; +create table t1 (id char(16) not null default '', primary key (id)); +insert into t1 values ('100'),('101'),('102'); +create table t2 (id char(16) default null); +insert into t2 values (1); +create view v1 as select t1.id from t1; +create view v2 as select t2.id from t2; +create view v3 as select (t1.id+2) as id from t1 natural left join t2; +select t1.id from t1 left join v2 using (id); +id +100 +101 +102 +select t1.id from v2 right join t1 using (id); +id +100 +101 +102 +select t1.id from t1 left join v3 using (id); +id +100 +101 +102 +select * from t1 left join v2 using (id); +id +100 +101 +102 +select * from v2 right join t1 using (id); +id +100 +101 +102 +select * from t1 left join v3 using (id); +id +100 +101 +102 +select v1.id from v1 left join v2 using (id); +id +100 +101 +102 +select v1.id from v2 right join v1 using (id); +id +100 +101 +102 +select v1.id from v1 left join v3 using (id); +id +100 +101 +102 +select * from v1 left join v2 using (id); +id +100 +101 +102 +select * from v2 right join v1 using (id); +id +100 +101 +102 +select * from v1 left join v3 using (id); +id +100 +101 +102 +drop table t1, t2; +drop view v1, v2, v3; +create table t1 (id int(11) not null default '0'); +insert into t1 values (123),(191),(192); +create table t2 (id char(16) character set utf8 not null); +insert into t2 values ('58013'),('58014'),('58015'),('58016'); +create table t3 (a_id int(11) not null, b_id char(16) character set utf8); +insert into t3 values (123,null),(123,null),(123,null),(123,null),(123,null),(123,'58013'); +select count(*) +from t1 inner join (t3 left join t2 on t2.id = t3.b_id) on t1.id = t3.a_id; +count(*) +6 +select count(*) +from t1 inner join (t2 right join t3 on t2.id = t3.b_id) on t1.id = t3.a_id; +count(*) +6 +drop table t1,t2,t3; +create table t1 (a int); +create table t2 (b int); +create table t3 (c int); +select * from t1 join t2 join t3 on (t1.a=t3.c); +a b c +select * from t1 join t2 left join t3 on (t1.a=t3.c); +a b c +select * from t1 join t2 right join t3 on (t1.a=t3.c); +a b c +select * from t1 join t2 straight_join t3 on (t1.a=t3.c); +a b c +drop table t1, t2 ,t3; +create table t1(f1 int, f2 date); +insert into t1 values(1,'2005-01-01'),(2,'2005-09-01'),(3,'2005-09-30'), +(4,'2005-10-01'),(5,'2005-12-30'); +select * from t1 where f2 >= 0 order by f2; +f1 f2 +1 2005-01-01 +2 2005-09-01 +3 2005-09-30 +4 2005-10-01 +5 2005-12-30 +select * from t1 where f2 >= '0000-00-00' order by f2; +f1 f2 +1 2005-01-01 +2 2005-09-01 +3 2005-09-30 +4 2005-10-01 +5 2005-12-30 +select * from t1 where f2 >= '2005-09-31' order by f2; +f1 f2 +4 2005-10-01 +5 2005-12-30 +select * from t1 where f2 >= '2005-09-3a' order by f2; +f1 f2 +3 2005-09-30 +4 2005-10-01 +5 2005-12-30 +Warnings: +Warning 1292 Incorrect date value: '2005-09-3a' for column 'f2' at row 1 +select * from t1 where f2 <= '2005-09-31' order by f2; +f1 f2 +1 2005-01-01 +2 2005-09-01 +3 2005-09-30 +select * from t1 where f2 <= '2005-09-3a' order by f2; +f1 f2 +1 2005-01-01 +2 2005-09-01 +Warnings: +Warning 1292 Incorrect date value: '2005-09-3a' for column 'f2' at row 1 +drop table t1; +create table t1 (f1 int, f2 int); +insert into t1 values (1, 30), (2, 20), (3, 10); +create algorithm=merge view v1 as select f1, f2 from t1; +create algorithm=merge view v2 (f2, f1) as select f1, f2 from t1; +create algorithm=merge view v3 as select t1.f1 as f2, t1.f2 as f1 from t1; +select t1.f1 as x1, f1 from t1 order by t1.f1; +x1 f1 +1 1 +2 2 +3 3 +select v1.f1 as x1, f1 from v1 order by v1.f1; +x1 f1 +1 1 +2 2 +3 3 +select v2.f1 as x1, f1 from v2 order by v2.f1; +x1 f1 +10 10 +20 20 +30 30 +select v3.f1 as x1, f1 from v3 order by v3.f1; +x1 f1 +10 10 +20 20 +30 30 +select f1, f2, v1.f1 as x1 from v1 order by v1.f1; +f1 f2 x1 +1 30 1 +2 20 2 +3 10 3 +select f1, f2, v2.f1 as x1 from v2 order by v2.f1; +f1 f2 x1 +10 3 10 +20 2 20 +30 1 30 +select f1, f2, v3.f1 as x1 from v3 order by v3.f1; +f1 f2 x1 +10 3 10 +20 2 20 +30 1 30 +drop table t1; +drop view v1, v2, v3; +CREATE TABLE t1(key_a int4 NOT NULL, optimus varchar(32), PRIMARY KEY(key_a)); +CREATE TABLE t2(key_a int4 NOT NULL, prime varchar(32), PRIMARY KEY(key_a)); +CREATE table t3(key_a int4 NOT NULL, key_b int4 NOT NULL, foo varchar(32), +PRIMARY KEY(key_a,key_b)); +INSERT INTO t1 VALUES (0,''); +INSERT INTO t1 VALUES (1,'i'); +INSERT INTO t1 VALUES (2,'j'); +INSERT INTO t1 VALUES (3,'k'); +INSERT INTO t2 VALUES (1,'r'); +INSERT INTO t2 VALUES (2,'s'); +INSERT INTO t2 VALUES (3,'t'); +INSERT INTO t3 VALUES (1,5,'x'); +INSERT INTO t3 VALUES (1,6,'y'); +INSERT INTO t3 VALUES (2,5,'xx'); +INSERT INTO t3 VALUES (2,6,'yy'); +INSERT INTO t3 VALUES (2,7,'zz'); +INSERT INTO t3 VALUES (3,5,'xxx'); +SELECT t2.key_a,foo +FROM t1 INNER JOIN t2 ON t1.key_a = t2.key_a +INNER JOIN t3 ON t1.key_a = t3.key_a +WHERE t2.key_a=2 and key_b=5; +key_a foo +2 xx +EXPLAIN SELECT t2.key_a,foo +FROM t1 INNER JOIN t2 ON t1.key_a = t2.key_a +INNER JOIN t3 ON t1.key_a = t3.key_a +WHERE t2.key_a=2 and key_b=5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 Using index +1 SIMPLE t2 const PRIMARY PRIMARY 4 const 1 Using index +1 SIMPLE t3 const PRIMARY PRIMARY 8 const,const 1 +SELECT t2.key_a,foo +FROM t1 INNER JOIN t2 ON t2.key_a = t1.key_a +INNER JOIN t3 ON t1.key_a = t3.key_a +WHERE t2.key_a=2 and key_b=5; +key_a foo +2 xx +EXPLAIN SELECT t2.key_a,foo +FROM t1 INNER JOIN t2 ON t2.key_a = t1.key_a +INNER JOIN t3 ON t1.key_a = t3.key_a +WHERE t2.key_a=2 and key_b=5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 Using index +1 SIMPLE t2 const PRIMARY PRIMARY 4 const 1 Using index +1 SIMPLE t3 const PRIMARY PRIMARY 8 const,const 1 +DROP TABLE t1,t2,t3; +create table t1 (f1 int); +insert into t1 values(1),(2); +create table t2 (f2 int, f3 int, key(f2)); +insert into t2 values(1,1),(2,2); +create table t3 (f4 int not null); +insert into t3 values (2),(2),(2); +select f1,(select count(*) from t2,t3 where f2=f1 and f3=f4) as count from t1; +f1 count +1 0 +2 3 +drop table t1,t2,t3; +create table t1 (f1 int unique); +create table t2 (f2 int unique); +create table t3 (f3 int unique); +insert into t1 values(1),(2); +insert into t2 values(1),(2); +insert into t3 values(1),(NULL); +select * from t3 where f3 is null; +f3 +NULL +select t2.f2 from t1 left join t2 on f1=f2 join t3 on f1=f3 where f1=1; +f2 +1 +drop table t1,t2,t3; +create table t1(f1 char, f2 char not null); +insert into t1 values(null,'a'); +create table t2 (f2 char not null); +insert into t2 values('b'); +select * from t1 left join t2 on f1=t2.f2 where t1.f2='a'; +f1 f2 f2 +NULL a NULL +drop table t1,t2; +select * from (select * left join t on f1=f2) tt; +ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'on f1=f2) tt' at line 1 +CREATE TABLE t1 (sku int PRIMARY KEY, pr int); +CREATE TABLE t2 (sku int PRIMARY KEY, sppr int, name varchar(255)); +INSERT INTO t1 VALUES +(10, 10), (20, 10), (30, 20), (40, 30), (50, 10), (60, 10); +INSERT INTO t2 VALUES +(10, 10, 'aaa'), (20, 10, 'bbb'), (30, 10, 'ccc'), (40, 20, 'ddd'), +(50, 10, 'eee'), (60, 20, 'fff'), (70, 20, 'ggg'), (80, 30, 'hhh'); +SELECT t2.sku, t2.sppr, t2.name, t1.sku, t1.pr +FROM t2, t1 WHERE t2.sku=20 AND (t2.sku=t1.sku OR t2.sppr=t1.sku); +sku sppr name sku pr +20 10 bbb 10 10 +20 10 bbb 20 10 +EXPLAIN +SELECT t2.sku, t2.sppr, t2.name, t1.sku, t1.pr +FROM t2, t1 WHERE t2.sku=20 AND (t2.sku=t1.sku OR t2.sppr=t1.sku); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t1 range PRIMARY PRIMARY 4 NULL 2 Using where +DROP TABLE t1,t2; +CREATE TABLE t1 (i TINYINT UNSIGNED NOT NULL); +INSERT t1 SET i = 0; +UPDATE t1 SET i = -1; +Warnings: +Warning 1264 Out of range value for column 'i' at row 1 +SELECT * FROM t1; +i +0 +UPDATE t1 SET i = CAST(i - 1 AS SIGNED); +Warnings: +Warning 1264 Out of range value for column 'i' at row 1 +SELECT * FROM t1; +i +0 +UPDATE t1 SET i = i - 1; +Warnings: +Warning 1264 Out of range value for column 'i' at row 1 +SELECT * FROM t1; +i +255 +DROP TABLE t1; +create table t1 (a int); +insert into t1 values (0),(1),(2),(3),(4),(5),(6),(7),(8),(9); +create table t2 (a int, b int, c int, e int, primary key(a,b,c)); +insert into t2 select A.a, B.a, C.a, C.a from t1 A, t1 B, t1 C; +analyze table t2; +Table Op Msg_type Msg_text +test.t2 analyze status OK +select 'In next EXPLAIN, B.rows must be exactly 10:' Z; +Z +In next EXPLAIN, B.rows must be exactly 10: +explain select * from t2 A, t2 B where A.a=5 and A.b=5 and A.C<5 +and B.a=5 and B.b=A.e and (B.b =1 or B.b = 3 or B.b=5); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE A range PRIMARY PRIMARY 12 NULL 4 Using where +1 SIMPLE B ref PRIMARY PRIMARY 8 const,test.A.e 10 +drop table t1, t2; +CREATE TABLE t1 (a int PRIMARY KEY, b int, INDEX(b)); +INSERT INTO t1 VALUES (1, 3), (9,4), (7,5), (4,5), (6,2), +(3,1), (5,1), (8,9), (2,2), (0,9); +CREATE TABLE t2 (c int, d int, f int, INDEX(c,f)); +INSERT INTO t2 VALUES +(1,0,0), (1,0,1), (2,0,0), (2,0,1), (3,0,0), (4,0,1), +(5,0,0), (5,0,1), (6,0,0), (0,0,1), (7,0,0), (7,0,1), +(0,0,0), (0,0,1), (8,0,0), (8,0,1), (9,0,0), (9,0,1); +EXPLAIN +SELECT a, c, d, f FROM t1,t2 WHERE a=c AND b BETWEEN 4 AND 6; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY,b b 5 NULL 3 Using where +1 SIMPLE t2 ref c c 5 test.t1.a 2 Using where +EXPLAIN +SELECT a, c, d, f FROM t1,t2 WHERE a=c AND b BETWEEN 4 AND 6 AND a > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY,b b 5 NULL 3 Using where +1 SIMPLE t2 ref c c 5 test.t1.a 2 Using where +DROP TABLE t1, t2; +create table t1 ( +a int unsigned not null auto_increment primary key, +b bit not null, +c bit not null +); +create table t2 ( +a int unsigned not null auto_increment primary key, +b bit not null, +c int unsigned not null, +d varchar(50) +); +insert into t1 (b,c) values (0,1), (0,1); +insert into t2 (b,c) values (0,1); +select t1.a, t1.b + 0, t1.c + 0, t2.a, t2.b + 0, t2.c, t2.d +from t1 left outer join t2 on t1.a = t2.c and t2.b <> 1 +where t1.b <> 1 order by t1.a; +a t1.b + 0 t1.c + 0 a t2.b + 0 c d +1 0 1 1 0 1 NULL +2 0 1 NULL NULL NULL NULL +drop table t1,t2; +SELECT 0.9888889889 * 1.011111411911; +0.9888889889 * 1.011111411911 +0.9998769417899202067879 +prepare stmt from 'select 1 as " a "'; +Warnings: +Warning 1466 Leading spaces are removed from name ' a ' +execute stmt; +a +1 +CREATE TABLE t1 (a int NOT NULL PRIMARY KEY, b int NOT NULL); +INSERT INTO t1 VALUES (1,1), (2,2), (3,3), (4,4); +CREATE TABLE t2 (c int NOT NULL, INDEX idx(c)); +INSERT INTO t2 VALUES +(1), (1), (1), (1), (1), (1), (1), (1), +(2), (2), (2), (2), +(3), (3), +(4); +EXPLAIN SELECT b FROM t1, t2 WHERE b=c AND a=1; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 ref idx idx 4 const 7 Using index +EXPLAIN SELECT b FROM t1, t2 WHERE b=c AND a=4; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 ref idx idx 4 const 1 Using index +DROP TABLE t1, t2; +CREATE TABLE t1 (id int NOT NULL PRIMARY KEY, a int); +INSERT INTO t1 VALUES (1,2), (2,NULL), (3,2); +CREATE TABLE t2 (b int, c INT, INDEX idx1(b)); +INSERT INTO t2 VALUES (2,1), (3,2); +CREATE TABLE t3 (d int, e int, INDEX idx1(d)); +INSERT INTO t3 VALUES (2,10), (2,20), (1,30), (2,40), (2,50); +EXPLAIN +SELECT * FROM t1 LEFT JOIN t2 ON t2.b=t1.a INNER JOIN t3 ON t3.d=t1.id +WHERE t1.id=2; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 const idx1 NULL NULL NULL 1 +1 SIMPLE t3 ref idx1 idx1 5 const 3 Using where +SELECT * FROM t1 LEFT JOIN t2 ON t2.b=t1.a INNER JOIN t3 ON t3.d=t1.id +WHERE t1.id=2; +id a b c d e +2 NULL NULL NULL 2 10 +2 NULL NULL NULL 2 20 +2 NULL NULL NULL 2 40 +2 NULL NULL NULL 2 50 +DROP TABLE t1,t2,t3; +create table t1 (c1 varchar(1), c2 int, c3 int, c4 int, c5 int, c6 int, +c7 int, c8 int, c9 int, fulltext key (`c1`)); +select distinct match (`c1`) against ('z') , c2, c3, c4,c5, c6,c7, c8 +from t1 where c9=1 order by c2, c2; +match (`c1`) against ('z') c2 c3 c4 c5 c6 c7 c8 +drop table t1; +CREATE TABLE t1 (pk varchar(10) PRIMARY KEY, fk varchar(16)); +CREATE TABLE t2 (pk varchar(16) PRIMARY KEY, fk varchar(10)); +INSERT INTO t1 VALUES +('d','dddd'), ('i','iii'), ('a','aa'), ('b','bb'), ('g','gg'), +('e','eee'), ('c','cccc'), ('h','hhh'), ('j','jjj'), ('f','fff'); +INSERT INTO t2 VALUES +('jjj', 'j'), ('cc','c'), ('ccc','c'), ('aaa', 'a'), ('jjjj','j'), +('hhh','h'), ('gg','g'), ('fff','f'), ('ee','e'), ('ffff','f'), +('bbb','b'), ('ff','f'), ('cccc','c'), ('dddd','d'), ('jj','j'), +('aaaa','a'), ('bb','b'), ('eeee','e'), ('aa','a'), ('hh','h'); +EXPLAIN SELECT t2.* +FROM t1 JOIN t2 ON t2.fk=t1.pk +WHERE t2.fk < 'c' AND t2.pk=t1.fk; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY PRIMARY 12 NULL 3 Using where +1 SIMPLE t2 eq_ref PRIMARY PRIMARY 18 test.t1.fk 1 Using where +EXPLAIN SELECT t2.* +FROM t1 JOIN t2 ON t2.fk=t1.pk +WHERE t2.fk BETWEEN 'a' AND 'b' AND t2.pk=t1.fk; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY PRIMARY 12 NULL 2 Using where +1 SIMPLE t2 eq_ref PRIMARY PRIMARY 18 test.t1.fk 1 Using where +EXPLAIN SELECT t2.* +FROM t1 JOIN t2 ON t2.fk=t1.pk +WHERE t2.fk IN ('a','b') AND t2.pk=t1.fk; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range PRIMARY PRIMARY 12 NULL 2 Using where +1 SIMPLE t2 eq_ref PRIMARY PRIMARY 18 test.t1.fk 1 Using where +DROP TABLE t1,t2; +CREATE TABLE t1 (a int, b varchar(20) NOT NULL, PRIMARY KEY(a)); +CREATE TABLE t2 (a int, b varchar(20) NOT NULL, +PRIMARY KEY (a), UNIQUE KEY (b)); +INSERT INTO t1 VALUES (1,'a'),(2,'b'),(3,'c'); +INSERT INTO t2 VALUES (1,'a'),(2,'b'),(3,'c'); +EXPLAIN SELECT t1.a FROM t1 LEFT JOIN t2 ON t2.b=t1.b WHERE t1.a=3; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +DROP TABLE t1,t2; +CREATE TABLE t1(id int PRIMARY KEY, b int, e int); +CREATE TABLE t2(i int, a int, INDEX si(i), INDEX ai(a)); +CREATE TABLE t3(a int PRIMARY KEY, c char(4), INDEX ci(c)); +INSERT INTO t1 VALUES +(1,10,19), (2,20,22), (4,41,42), (9,93,95), (7, 77,79), +(6,63,67), (5,55,58), (3,38,39), (8,81,89); +INSERT INTO t2 VALUES +(21,210), (41,410), (82,820), (83,830), (84,840), +(65,650), (51,510), (37,370), (94,940), (76,760), +(22,220), (33,330), (40,400), (95,950), (38,380), +(67,670), (88,880), (57,570), (96,960), (97,970); +INSERT INTO t3 VALUES +(210,'bb'), (950,'ii'), (400,'ab'), (500,'ee'), (220,'gg'), +(440,'gg'), (310,'eg'), (380,'ee'), (840,'bb'), (830,'ff'), +(230,'aa'), (960,'ii'), (410,'aa'), (510,'ee'), (290,'bb'), +(450,'gg'), (320,'dd'), (390,'hh'), (850,'jj'), (860,'ff'); +EXPLAIN +SELECT t3.a FROM t1,t2 FORCE INDEX (si),t3 +WHERE t1.id = 8 AND t2.i BETWEEN t1.b AND t1.e AND +t3.a=t2.a AND t3.c IN ('bb','ee'); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 range si si 5 NULL 4 Using where +1 SIMPLE t3 eq_ref PRIMARY,ci PRIMARY 4 test.t2.a 1 Using where +EXPLAIN +SELECT t3.a FROM t1,t2,t3 +WHERE t1.id = 8 AND t2.i BETWEEN t1.b AND t1.e AND +t3.a=t2.a AND t3.c IN ('bb','ee') ; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 range si,ai si 5 NULL 4 Using where +1 SIMPLE t3 eq_ref PRIMARY,ci PRIMARY 4 test.t2.a 1 Using where +EXPLAIN +SELECT t3.a FROM t1,t2 FORCE INDEX (si),t3 +WHERE t1.id = 8 AND (t2.i=t1.b OR t2.i=t1.e) AND t3.a=t2.a AND +t3.c IN ('bb','ee'); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 range si si 5 NULL 2 Using where +1 SIMPLE t3 eq_ref PRIMARY,ci PRIMARY 4 test.t2.a 1 Using where +EXPLAIN +SELECT t3.a FROM t1,t2,t3 +WHERE t1.id = 8 AND (t2.i=t1.b OR t2.i=t1.e) AND t3.a=t2.a AND +t3.c IN ('bb','ee'); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t2 range si,ai si 5 NULL 2 Using where +1 SIMPLE t3 eq_ref PRIMARY,ci PRIMARY 4 test.t2.a 1 Using where +DROP TABLE t1,t2,t3; +CREATE TABLE t1 ( f1 int primary key, f2 int, f3 int, f4 int, f5 int, f6 int, checked_out int); +CREATE TABLE t2 ( f11 int PRIMARY KEY ); +INSERT INTO t1 VALUES (1,1,1,0,0,0,0),(2,1,1,3,8,1,0),(3,1,1,4,12,1,0); +INSERT INTO t2 VALUES (62); +SELECT * FROM t1 LEFT JOIN t2 ON f11 = t1.checked_out GROUP BY f1 ORDER BY f2, f3, f4, f5 LIMIT 0, 1; +f1 f2 f3 f4 f5 f6 checked_out f11 +1 1 1 0 0 0 0 NULL +DROP TABLE t1, t2; +DROP TABLE IF EXISTS t1; +CREATE TABLE t1(a int); +INSERT into t1 values (1), (2), (3); +SELECT * FROM t1 LIMIT 2, -1; +ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '-1' at line 1 +DROP TABLE t1; +CREATE TABLE t1 ( +ID_with_null int NULL, +ID_better int NOT NULL, +INDEX idx1 (ID_with_null), +INDEX idx2 (ID_better) +); +INSERT INTO t1 VALUES (1,1), (2,1), (null,3), (null,3), (null,3), (null,3); +INSERT INTO t1 SELECT * FROM t1 WHERE ID_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID_with_null IS NULL; +SELECT COUNT(*) FROM t1 WHERE ID_with_null IS NULL; +COUNT(*) +128 +SELECT COUNT(*) FROM t1 WHERE ID_better=1; +COUNT(*) +2 +EXPLAIN SELECT * FROM t1 WHERE ID_better=1 AND ID_with_null IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +DROP INDEX idx1 ON t1; +CREATE UNIQUE INDEX idx1 ON t1(ID_with_null); +EXPLAIN SELECT * FROM t1 WHERE ID_better=1 AND ID_with_null IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +DROP TABLE t1; +CREATE TABLE t1 ( +ID1_with_null int NULL, +ID2_with_null int NULL, +ID_better int NOT NULL, +INDEX idx1 (ID1_with_null, ID2_with_null), +INDEX idx2 (ID_better) +); +INSERT INTO t1 VALUES (1,1,1), (2,2,1), (3,null,3), (null,3,3), (null,null,3), +(3,null,3), (null,3,3), (null,null,3), (3,null,3), (null,3,3), (null,null,3); +INSERT INTO t1 SELECT * FROM t1 WHERE ID1_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID2_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID1_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID2_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID1_with_null IS NULL; +INSERT INTO t1 SELECT * FROM t1 WHERE ID2_with_null IS NULL; +SELECT COUNT(*) FROM t1 WHERE ID1_with_null IS NULL AND ID2_with_null=3; +COUNT(*) +24 +SELECT COUNT(*) FROM t1 WHERE ID1_with_null=3 AND ID2_with_null IS NULL; +COUNT(*) +24 +SELECT COUNT(*) FROM t1 WHERE ID1_with_null IS NULL AND ID2_with_null IS NULL; +COUNT(*) +192 +SELECT COUNT(*) FROM t1 WHERE ID_better=1; +COUNT(*) +2 +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null IS NULL AND ID2_with_null=3 ; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null=3 AND ID2_with_null=3 IS NULL ; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null IS NULL AND ID2_with_null IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +DROP INDEX idx1 ON t1; +CREATE UNIQUE INDEX idx1 ON t1(ID1_with_null,ID2_with_null); +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null IS NULL AND ID2_with_null=3 ; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null=3 AND ID2_with_null IS NULL ; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null IS NULL AND ID2_with_null IS NULL; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +EXPLAIN SELECT * FROM t1 +WHERE ID_better=1 AND ID1_with_null IS NULL AND +(ID2_with_null=1 OR ID2_with_null=2); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ref idx1,idx2 idx2 4 const 1 Using where +DROP TABLE t1; +CREATE TABLE t1 (a INT, ts TIMESTAMP, KEY ts(ts)); +INSERT INTO t1 VALUES (30,"2006-01-03 23:00:00"), (31,"2006-01-03 23:00:00"); +ANALYZE TABLE t1; +Table Op Msg_type Msg_text +test.t1 analyze status OK +CREATE TABLE t2 (a INT, dt1 DATETIME, dt2 DATETIME, PRIMARY KEY (a)); +INSERT INTO t2 VALUES (30, "2006-01-01 00:00:00", "2999-12-31 00:00:00"); +INSERT INTO t2 SELECT a+1,dt1,dt2 FROM t2; +ANALYZE TABLE t2; +Table Op Msg_type Msg_text +test.t2 analyze status OK +EXPLAIN +SELECT * FROM t1 LEFT JOIN t2 ON (t1.a=t2.a) WHERE t1.a=30 +AND t1.ts BETWEEN t2.dt1 AND t2.dt2 +AND t1.ts BETWEEN "2006-01-01" AND "2006-12-31"; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 const PRIMARY PRIMARY 4 const 1 +1 SIMPLE t1 range ts ts 4 NULL 1 Using where +Warnings: +Warning 1292 Incorrect datetime value: '2999-12-31 00:00:00' for column 'ts' at row 1 +SELECT * FROM t1 LEFT JOIN t2 ON (t1.a=t2.a) WHERE t1.a=30 +AND t1.ts BETWEEN t2.dt1 AND t2.dt2 +AND t1.ts BETWEEN "2006-01-01" AND "2006-12-31"; +a ts a dt1 dt2 +30 2006-01-03 23:00:00 30 2006-01-01 00:00:00 2999-12-31 00:00:00 +Warnings: +Warning 1292 Incorrect datetime value: '2999-12-31 00:00:00' for column 'ts' at row 1 +DROP TABLE t1,t2; +create table t1 (a bigint unsigned); +insert into t1 values +(if(1, 9223372036854775808, 1)), +(case when 1 then 9223372036854775808 else 1 end), +(coalesce(9223372036854775808, 1)); +select * from t1; +a +9223372036854775808 +9223372036854775808 +9223372036854775808 +drop table t1; +create table t1 select +if(1, 9223372036854775808, 1) i, +case when 1 then 9223372036854775808 else 1 end c, +coalesce(9223372036854775808, 1) co; +show create table t1; +Table Create Table +t1 CREATE TABLE `t1` ( + `i` decimal(19,0) NOT NULL DEFAULT '0', + `c` decimal(19,0) NOT NULL DEFAULT '0', + `co` decimal(19,0) NOT NULL DEFAULT '0' +) ENGINE=MyISAM DEFAULT CHARSET=latin1 +drop table t1; +select +if(1, cast(1111111111111111111 as unsigned), 1) i, +case when 1 then cast(1111111111111111111 as unsigned) else 1 end c, +coalesce(cast(1111111111111111111 as unsigned), 1) co; +i c co +1111111111111111111 1111111111111111111 1111111111111111111 +CREATE TABLE t1 (name varchar(255)); +CREATE TABLE t2 (name varchar(255), n int, KEY (name(3))); +INSERT INTO t1 VALUES ('ccc'), ('bb'), ('cc '), ('aa '), ('aa'); +INSERT INTO t2 VALUES ('bb',1), ('aa',2), ('cc ',3); +INSERT INTO t2 VALUES (concat('cc ', 0x06), 4); +INSERT INTO t2 VALUES ('cc',5), ('bb ',6), ('cc ',7); +SELECT * FROM t2; +name n +bb 1 +aa 2 +cc 3 +cc 4 +cc 5 +bb 6 +cc 7 +SELECT * FROM t2 ORDER BY name; +name n +aa 2 +bb 1 +bb 6 +cc 4 +cc 3 +cc 5 +cc 7 +SELECT name, LENGTH(name), n FROM t2 ORDER BY name; +name LENGTH(name) n +aa 2 2 +bb 2 1 +bb 3 6 +cc 4 4 +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name, LENGTH(name), n FROM t2 WHERE name='cc '; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref name name 6 const 3 Using where +SELECT name, LENGTH(name), n FROM t2 WHERE name='cc '; +name LENGTH(name) n +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range name name 6 NULL 3 Using where +SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%'; +name LENGTH(name) n +cc 5 3 +cc 4 4 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%' ORDER BY name; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range name name 6 NULL 3 Using where; Using filesort +SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%' ORDER BY name; +name LENGTH(name) n +cc 4 4 +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT * FROM t1 LEFT JOIN t2 ON t1.name=t2.name; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 +1 SIMPLE t2 ref name name 6 test.t1.name 2 +SELECT * FROM t1 LEFT JOIN t2 ON t1.name=t2.name; +name name n +ccc NULL NULL +bb bb 1 +bb bb 6 +cc cc 3 +cc cc 5 +cc cc 7 +aa aa 2 +aa aa 2 +DROP TABLE t1,t2; +CREATE TABLE t1 (name text); +CREATE TABLE t2 (name text, n int, KEY (name(3))); +INSERT INTO t1 VALUES ('ccc'), ('bb'), ('cc '), ('aa '), ('aa'); +INSERT INTO t2 VALUES ('bb',1), ('aa',2), ('cc ',3); +INSERT INTO t2 VALUES (concat('cc ', 0x06), 4); +INSERT INTO t2 VALUES ('cc',5), ('bb ',6), ('cc ',7); +SELECT * FROM t2; +name n +bb 1 +aa 2 +cc 3 +cc 4 +cc 5 +bb 6 +cc 7 +SELECT * FROM t2 ORDER BY name; +name n +aa 2 +bb 1 +bb 6 +cc 4 +cc 3 +cc 5 +cc 7 +SELECT name, LENGTH(name), n FROM t2 ORDER BY name; +name LENGTH(name) n +aa 2 2 +bb 2 1 +bb 3 6 +cc 4 4 +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name, LENGTH(name), n FROM t2 WHERE name='cc '; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 ref name name 6 const 3 Using where +SELECT name, LENGTH(name), n FROM t2 WHERE name='cc '; +name LENGTH(name) n +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range name name 6 NULL 3 Using where +SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%'; +name LENGTH(name) n +cc 5 3 +cc 4 4 +cc 2 5 +cc 3 7 +EXPLAIN SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%' ORDER BY name; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t2 range name name 6 NULL 3 Using where; Using filesort +SELECT name , LENGTH(name), n FROM t2 WHERE name LIKE 'cc%' ORDER BY name; +name LENGTH(name) n +cc 4 4 +cc 5 3 +cc 2 5 +cc 3 7 +EXPLAIN SELECT * FROM t1 LEFT JOIN t2 ON t1.name=t2.name; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 5 +1 SIMPLE t2 ref name name 6 test.t1.name 2 +SELECT * FROM t1 LEFT JOIN t2 ON t1.name=t2.name; +name name n +ccc NULL NULL +bb bb 1 +bb bb 6 +cc cc 3 +cc cc 5 +cc cc 7 +aa aa 2 +aa aa 2 +DROP TABLE t1,t2; +CREATE TABLE t1 ( +access_id int NOT NULL default '0', +name varchar(20) default NULL, +rank int NOT NULL default '0', +KEY idx (access_id) +); +CREATE TABLE t2 ( +faq_group_id int NOT NULL default '0', +faq_id int NOT NULL default '0', +access_id int default NULL, +UNIQUE KEY idx1 (faq_id), +KEY idx2 (faq_group_id,faq_id) +); +INSERT INTO t1 VALUES +(1,'Everyone',2),(2,'Help',3),(3,'Technical Support',1),(4,'Chat User',4); +INSERT INTO t2 VALUES +(261,265,1),(490,494,1); +SELECT t2.faq_id +FROM t1 INNER JOIN t2 IGNORE INDEX (idx1) +ON (t1.access_id = t2.access_id) +LEFT JOIN t2 t +ON (t.faq_group_id = t2.faq_group_id AND +find_in_set(t.access_id, '1,4') < find_in_set(t2.access_id, '1,4')) +WHERE +t2.access_id IN (1,4) AND t.access_id IS NULL AND t2.faq_id in (265); +faq_id +265 +SELECT t2.faq_id +FROM t1 INNER JOIN t2 +ON (t1.access_id = t2.access_id) +LEFT JOIN t2 t +ON (t.faq_group_id = t2.faq_group_id AND +find_in_set(t.access_id, '1,4') < find_in_set(t2.access_id, '1,4')) +WHERE +t2.access_id IN (1,4) AND t.access_id IS NULL AND t2.faq_id in (265); +faq_id +265 +DROP TABLE t1,t2; +CREATE TABLE t1 (a INT, b INT, KEY inx (b,a)); +INSERT INTO t1 VALUES (1,1), (1,2), (1,3), (1,4), (1,5), (1, 6), (1,7); +EXPLAIN SELECT COUNT(*) FROM t1 f1 INNER JOIN t1 f2 +ON ( f1.b=f2.b AND f1.a<f2.a ) +WHERE 1 AND f1.b NOT IN (100,2232,3343,51111); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE f1 index inx inx 10 NULL 7 Using where; Using index +1 SIMPLE f2 ref inx inx 5 test.f1.b 1 Using where; Using index +DROP TABLE t1; +CREATE TABLE t1 (c1 INT, c2 INT); +INSERT INTO t1 VALUES (1,11), (2,22), (2,22); +EXPLAIN SELECT c1 FROM t1 WHERE (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT COUNT(c2)))))))))))))))))))))))))))))))) > 0; +id select_type table type possible_keys key key_len ref rows Extra +1 PRIMARY t1 ALL NULL NULL NULL NULL 3 Using where +31 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL No tables used +32 DEPENDENT SUBQUERY NULL NULL NULL NULL NULL NULL NULL No tables used +EXPLAIN SELECT c1 FROM t1 WHERE (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT (SELECT COUNT(c2))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) > 0; +ERROR HY000: Too high level of nesting for select +DROP TABLE t1; +CREATE TABLE t1 ( +c1 int(11) NOT NULL AUTO_INCREMENT, +c2 varchar(1000) DEFAULT NULL, +c3 bigint(20) DEFAULT NULL, +c4 bigint(20) DEFAULT NULL, +PRIMARY KEY (c1) +); +EXPLAIN EXTENDED +SELECT join_2.c1 +FROM +t1 AS join_0, +t1 AS join_1, +t1 AS join_2, +t1 AS join_3, +t1 AS join_4, +t1 AS join_5, +t1 AS join_6, +t1 AS join_7 +WHERE +join_0.c1=join_1.c1 AND +join_1.c1=join_2.c1 AND +join_2.c1=join_3.c1 AND +join_3.c1=join_4.c1 AND +join_4.c1=join_5.c1 AND +join_5.c1=join_6.c1 AND +join_6.c1=join_7.c1 +OR +join_0.c2 < '?' AND +join_1.c2 < '?' AND +join_2.c2 > '?' AND +join_2.c2 < '!' AND +join_3.c2 > '?' AND +join_4.c2 = '?' AND +join_5.c2 <> '?' AND +join_6.c2 <> '?' AND +join_7.c2 >= '?' AND +join_0.c1=join_1.c1 AND +join_1.c1=join_2.c1 AND +join_2.c1=join_3.c1 AND +join_3.c1=join_4.c1 AND +join_4.c1=join_5.c1 AND +join_5.c1=join_6.c1 AND +join_6.c1=join_7.c1 +GROUP BY +join_3.c1, +join_2.c1, +join_7.c1, +join_1.c1, +join_0.c1; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables +Warnings: +Note 1003 select '0' AS `c1` from `test`.`t1` `join_0` join `test`.`t1` `join_1` join `test`.`t1` `join_2` join `test`.`t1` `join_3` join `test`.`t1` `join_4` join `test`.`t1` `join_5` join `test`.`t1` `join_6` join `test`.`t1` `join_7` where 0 group by '0','0','0','0','0' +SHOW WARNINGS; +Level Code Message +Note 1003 select '0' AS `c1` from `test`.`t1` `join_0` join `test`.`t1` `join_1` join `test`.`t1` `join_2` join `test`.`t1` `join_3` join `test`.`t1` `join_4` join `test`.`t1` `join_5` join `test`.`t1` `join_6` join `test`.`t1` `join_7` where 0 group by '0','0','0','0','0' +DROP TABLE t1; +SELECT 1 AS ` `; + +1 +Warnings: +Warning 1474 Name ' ' has become '' +SELECT 1 AS ` `; + +1 +Warnings: +Warning 1474 Name ' ' has become '' +SELECT 1 AS ` x`; +x +1 +Warnings: +Warning 1466 Leading spaces are removed from name ' x' +CREATE VIEW v1 AS SELECT 1 AS ``; +ERROR 42000: Incorrect column name '' +CREATE VIEW v1 AS SELECT 1 AS ` `; +ERROR 42000: Incorrect column name ' ' +CREATE VIEW v1 AS SELECT 1 AS ` `; +ERROR 42000: Incorrect column name ' ' +CREATE VIEW v1 AS SELECT (SELECT 1 AS ` `); +ERROR 42000: Incorrect column name ' ' +CREATE VIEW v1 AS SELECT 1 AS ` x`; +Warnings: +Warning 1466 Leading spaces are removed from name ' x' +SELECT `x` FROM v1; +x +1 +ALTER VIEW v1 AS SELECT 1 AS ` `; +ERROR 42000: Incorrect column name ' ' +DROP VIEW v1; +select str_to_date('2007-10-09','%Y-%m-%d') between '2007/10/01 00:00:00 GMT' + and '2007/10/20 00:00:00 GMT'; +str_to_date('2007-10-09','%Y-%m-%d') between '2007/10/01 00:00:00 GMT' + and '2007/10/20 00:00:00 GMT' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007/10/01 00:00:00 GMT' +Warning 1292 Truncated incorrect datetime value: '2007/10/20 00:00:00 GMT' +select str_to_date('2007-10-09','%Y-%m-%d') > '2007/10/01 00:00:00 GMT-6'; +str_to_date('2007-10-09','%Y-%m-%d') > '2007/10/01 00:00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect date value: '2007/10/01 00:00:00 GMT-6' +select str_to_date('2007-10-09','%Y-%m-%d') <= '2007/10/2000:00:00 GMT-6'; +str_to_date('2007-10-09','%Y-%m-%d') <= '2007/10/2000:00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect date value: '2007/10/2000:00:00 GMT-6' +select str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-1 00:00:00 GMT-6'; +str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-1 00:00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect date value: '2007-10-1 00:00:00 GMT-6' +select str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-01 x00:00:00 GMT-6'; +str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-01 x00:00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect date value: '2007-10-01 x00:00:00 GMT-6' +select str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 00:00:00 GMT-6'; +str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 00:00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 00:00:00 GMT-6' +select str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 00:x00:00 GMT-6'; +str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 00:x00:00 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 00:x00:00 GMT-6' +select str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 x12:34:56 GMT-6'; +str_to_date('2007-10-01','%Y-%m-%d %H:%i:%s') = '2007-10-01 x12:34:56 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 x12:34:56 GMT-6' +select str_to_date('2007-10-01 12:34:00','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34x:56 GMT-6'; +str_to_date('2007-10-01 12:34:00','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34x:56 GMT-6' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 12:34x:56 GMT-6' +select str_to_date('2007-10-01 12:34:56','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34x:56 GMT-6'; +str_to_date('2007-10-01 12:34:56','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34x:56 GMT-6' +0 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 12:34x:56 GMT-6' +select str_to_date('2007-10-01 12:34:56','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34:56'; +str_to_date('2007-10-01 12:34:56','%Y-%m-%d %H:%i:%s') = '2007-10-01 12:34:56' +1 +select str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-01 12:00:00'; +str_to_date('2007-10-01','%Y-%m-%d') = '2007-10-01 12:00:00' +0 +select str_to_date('2007-10-01 12','%Y-%m-%d %H') = '2007-10-01 12:00:00'; +str_to_date('2007-10-01 12','%Y-%m-%d %H') = '2007-10-01 12:00:00' +1 +select str_to_date('2007-10-01 12:34','%Y-%m-%d %H') = '2007-10-01 12:00:00'; +str_to_date('2007-10-01 12:34','%Y-%m-%d %H') = '2007-10-01 12:00:00' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-01 12:34' +select str_to_date('2007-02-30 12:34','%Y-%m-%d %H:%i') = '2007-02-30 12:34'; +str_to_date('2007-02-30 12:34','%Y-%m-%d %H:%i') = '2007-02-30 12:34' +1 +select str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34'; +str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34' +1 +select str_to_date('2007-10-00','%Y-%m-%d') between '2007/09/01 00:00:00' + and '2007/10/20 00:00:00'; +str_to_date('2007-10-00','%Y-%m-%d') between '2007/09/01 00:00:00' + and '2007/10/20 00:00:00' +1 +set SQL_MODE=TRADITIONAL; +select str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34'; +str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34' +0 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-00 12:34' +select str_to_date('2007-10-01 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34'; +str_to_date('2007-10-01 12:34','%Y-%m-%d %H:%i') = '2007-10-00 12:34' +0 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-00 12:34' +select str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-01 12:34'; +str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '2007-10-01 12:34' +0 +Warnings: +Warning 1292 Truncated incorrect datetime value: '2007-10-00 12:34:00' +select str_to_date('2007-10-00','%Y-%m-%d') between '2007/09/01' + and '2007/10/20'; +str_to_date('2007-10-00','%Y-%m-%d') between '2007/09/01' + and '2007/10/20' +0 +Warnings: +Warning 1292 Incorrect datetime value: '2007-10-00' for column '2007/09/01' at row 1 +Warning 1292 Incorrect datetime value: '2007-10-00' for column '2007/10/20' at row 1 +set SQL_MODE=DEFAULT; +select str_to_date('2007-10-00','%Y-%m-%d') between '' and '2007/10/20'; +str_to_date('2007-10-00','%Y-%m-%d') between '' and '2007/10/20' +1 +Warnings: +Warning 1292 Truncated incorrect datetime value: '' +select str_to_date('','%Y-%m-%d') between '2007/10/01' and '2007/10/20'; +str_to_date('','%Y-%m-%d') between '2007/10/01' and '2007/10/20' +0 +select str_to_date('','%Y-%m-%d %H:%i') = '2007-10-01 12:34'; +str_to_date('','%Y-%m-%d %H:%i') = '2007-10-01 12:34' +0 +select str_to_date(NULL,'%Y-%m-%d %H:%i') = '2007-10-01 12:34'; +str_to_date(NULL,'%Y-%m-%d %H:%i') = '2007-10-01 12:34' +NULL +select str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = ''; +str_to_date('2007-10-00 12:34','%Y-%m-%d %H:%i') = '' +0 +Warnings: +Warning 1292 Truncated incorrect datetime value: '' +select str_to_date('1','%Y-%m-%d') = '1'; +str_to_date('1','%Y-%m-%d') = '1' +0 +Warnings: +Warning 1292 Truncated incorrect date value: '1' +select str_to_date('1','%Y-%m-%d') = '1'; +str_to_date('1','%Y-%m-%d') = '1' +0 +Warnings: +Warning 1292 Truncated incorrect date value: '1' +select str_to_date('','%Y-%m-%d') = ''; +str_to_date('','%Y-%m-%d') = '' +0 +Warnings: +Warning 1292 Truncated incorrect date value: '' +select str_to_date('1000-01-01','%Y-%m-%d') between '0000-00-00' and NULL; +str_to_date('1000-01-01','%Y-%m-%d') between '0000-00-00' and NULL +0 +select str_to_date('1000-01-01','%Y-%m-%d') between NULL and '2000-00-00'; +str_to_date('1000-01-01','%Y-%m-%d') between NULL and '2000-00-00' +0 +select str_to_date('1000-01-01','%Y-%m-%d') between NULL and NULL; +str_to_date('1000-01-01','%Y-%m-%d') between NULL and NULL +0 +CREATE TABLE t1 (c11 INT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY); +CREATE TABLE t2 (c21 INT UNSIGNED NOT NULL, +c22 INT DEFAULT NULL, +KEY(c21, c22)); +CREATE TABLE t3 (c31 INT UNSIGNED NOT NULL DEFAULT 0, +c32 INT DEFAULT NULL, +c33 INT NOT NULL, +c34 INT UNSIGNED DEFAULT 0, +KEY (c33, c34, c32)); +INSERT INTO t1 values (),(),(),(),(); +INSERT INTO t2 SELECT a.c11, b.c11 FROM t1 a, t1 b; +INSERT INTO t3 VALUES (1, 1, 1, 0), +(2, 2, 0, 0), +(3, 3, 1, 0), +(4, 4, 0, 0), +(5, 5, 1, 0); +SELECT c32 FROM t1, t2, t3 WHERE t1.c11 IN (1, 3, 5) AND +t3.c31 = t1.c11 AND t2.c21 = t1.c11 AND +t3.c33 = 1 AND t2.c22 in (1, 3) +ORDER BY c32; +c32 +1 +1 +3 +3 +5 +5 +SELECT c32 FROM t1, t2, t3 WHERE t1.c11 IN (1, 3, 5) AND +t3.c31 = t1.c11 AND t2.c21 = t1.c11 AND +t3.c33 = 1 AND t2.c22 in (1, 3) +ORDER BY c32 DESC; +c32 +5 +5 +3 +3 +1 +1 +DROP TABLE t1, t2, t3; + +# +# Bug#30736: Row Size Too Large Error Creating a Table and +# Inserting Data. +# +DROP TABLE IF EXISTS t1; +DROP TABLE IF EXISTS t2; + +CREATE TABLE t1( +c1 DECIMAL(10, 2), +c2 FLOAT); + +INSERT INTO t1 VALUES (0, 1), (2, 3), (4, 5); + +CREATE TABLE t2( +c3 DECIMAL(10, 2)) +SELECT +c1 * c2 AS c3 +FROM t1; + +SELECT * FROM t1; +c1 c2 +0.00 1 +2.00 3 +4.00 5 + +SELECT * FROM t2; +c3 +0.00 +6.00 +20.00 + +DROP TABLE t1; +DROP TABLE t2; + +CREATE TABLE t1 (c1 BIGINT NOT NULL); +INSERT INTO t1 (c1) VALUES (1); +SELECT * FROM t1 WHERE c1 > NULL + 1; +c1 +DROP TABLE t1; + +CREATE TABLE t1 (a VARCHAR(10) NOT NULL PRIMARY KEY); +INSERT INTO t1 (a) VALUES ('foo0'), ('bar0'), ('baz0'); +SELECT * FROM t1 WHERE a IN (CONCAT('foo', 0), 'bar'); +a +foo0 +DROP TABLE t1; +CREATE TABLE t1 (a INT, b INT); +CREATE TABLE t2 (a INT, c INT, KEY(a)); +INSERT INTO t1 VALUES (1, 1), (2, 2); +INSERT INTO t2 VALUES (1, 1), (1, 2), (1, 3), (1, 4), (1, 5), +(2, 1), (2, 2), (2, 3), (2, 4), (2, 5), +(3, 1), (3, 2), (3, 3), (3, 4), (3, 5), +(4, 1), (4, 2), (4, 3), (4, 4), (4, 5); +FLUSH STATUS; +SELECT DISTINCT b FROM t1 LEFT JOIN t2 USING(a) WHERE c <= 3; +b +1 +2 +SHOW STATUS LIKE 'Handler_read%'; +Variable_name Value +Handler_read_first 0 +Handler_read_key 2 +Handler_read_next 0 +Handler_read_prev 0 +Handler_read_rnd 0 +Handler_read_rnd_next 6 +DROP TABLE t1, t2; +CREATE TABLE t1 (f1 bigint(20) NOT NULL default '0', +f2 int(11) NOT NULL default '0', +f3 bigint(20) NOT NULL default '0', +f4 varchar(255) NOT NULL default '', +PRIMARY KEY (f1), +KEY key1 (f4), +KEY key2 (f2)); +CREATE TABLE t2 (f1 int(11) NOT NULL default '0', +f2 enum('A1','A2','A3') NOT NULL default 'A1', +f3 int(11) NOT NULL default '0', +PRIMARY KEY (f1), +KEY key1 (f3)); +CREATE TABLE t3 (f1 bigint(20) NOT NULL default '0', +f2 datetime NOT NULL default '1980-01-01 00:00:00', +PRIMARY KEY (f1)); +insert into t1 values (1, 1, 1, 'abc'); +insert into t1 values (2, 1, 2, 'def'); +insert into t1 values (3, 1, 2, 'def'); +insert into t2 values (1, 'A1', 1); +insert into t3 values (1, '1980-01-01'); +SELECT a.f3, cr.f4, count(*) count +FROM t2 a +STRAIGHT_JOIN t1 cr ON cr.f2 = a.f1 +LEFT JOIN +(t1 cr2 +JOIN t3 ae2 ON cr2.f3 = ae2.f1 +) ON a.f1 = cr2.f2 AND ae2.f2 < now() - INTERVAL 7 DAY AND +cr.f4 = cr2.f4 +GROUP BY a.f3, cr.f4; +f3 f4 count +1 abc 1 +1 def 2 +drop table t1, t2, t3; +CREATE TABLE t1 (a INT KEY, b INT); +INSERT INTO t1 VALUES (1,1), (2,2), (3,3), (4,4); +EXPLAIN EXTENDED SELECT a, b FROM t1 WHERE a > 1 AND a = b LIMIT 2; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 range PRIMARY PRIMARY 4 NULL 3 100.00 Using where +Warnings: +Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b` from `test`.`t1` where ((`test`.`t1`.`b` = `test`.`t1`.`a`) and (`test`.`t1`.`a` > 1)) limit 2 +EXPLAIN EXTENDED SELECT a, b FROM t1 WHERE a > 1 AND b = a LIMIT 2; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 range PRIMARY PRIMARY 4 NULL 3 100.00 Using where +Warnings: +Note 1003 select `test`.`t1`.`a` AS `a`,`test`.`t1`.`b` AS `b` from `test`.`t1` where ((`test`.`t1`.`a` = `test`.`t1`.`b`) and (`test`.`t1`.`a` > 1)) limit 2 +DROP TABLE t1; +# +# Bug#47019: Assertion failed: 0, file .\rt_mbr.c, line 138 when +# forcing a spatial index +# +CREATE TABLE t1(a LINESTRING NOT NULL, SPATIAL KEY(a)); +INSERT INTO t1 VALUES +(GEOMFROMTEXT('LINESTRING(-1 -1, 1 -1, -1 -1, -1 1, 1 1)')), +(GEOMFROMTEXT('LINESTRING(-1 -1, 1 -1, -1 -1, -1 1, 1 1)')); +EXPLAIN SELECT 1 FROM t1 NATURAL LEFT JOIN t1 AS t2; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 2 +1 SIMPLE t2 ALL a NULL NULL NULL 2 +SELECT 1 FROM t1 NATURAL LEFT JOIN t1 AS t2; +1 +1 +1 +1 +1 +EXPLAIN SELECT 1 FROM t1 NATURAL LEFT JOIN t1 AS t2 FORCE INDEX(a); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 2 +1 SIMPLE t2 ALL a NULL NULL NULL 2 +SELECT 1 FROM t1 NATURAL LEFT JOIN t1 AS t2 FORCE INDEX(a); +1 +1 +1 +1 +1 +DROP TABLE t1; +# +# Bug #48291 : crash with row() operator,select into @var, and +# subquery returning multiple rows +# +CREATE TABLE t1(a INT); +INSERT INTO t1 VALUES (2),(3); +# Should not crash +SELECT 1 FROM t1 WHERE a <> 1 AND NOT +ROW(a,a) <=> ROW((SELECT 1 FROM t1 WHERE 1=2),(SELECT 1 FROM t1)) +INTO @var0; +ERROR 21000: Subquery returns more than 1 row +DROP TABLE t1; +# +# Bug #48458: simple query tries to allocate enormous amount of +# memory +# +CREATE TABLE t1(a INT NOT NULL, b YEAR); +INSERT INTO t1 VALUES (); +Warnings: +Warning 1364 Field 'a' doesn't have a default value +CREATE TABLE t2(c INT); +# Should not err out because of out-of-memory +SELECT 1 FROM t2 JOIN t1 ON 1=1 +WHERE a != '1' AND NOT a >= b OR NOT ROW(b,a )<> ROW(a,a); +1 +DROP TABLE t1,t2; +# +# Bug #49199: Optimizer handles incorrectly: +# field='const1' AND field='const2' in some cases + +CREATE TABLE t1(a DATETIME NOT NULL); +INSERT INTO t1 VALUES('2001-01-01'); +SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +a +2001-01-01 00:00:00 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a` from `test`.`t1` where 1 +DROP TABLE t1; +CREATE TABLE t1(a DATE NOT NULL); +INSERT INTO t1 VALUES('2001-01-01'); +SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +a +2001-01-01 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01' AS `a` from `test`.`t1` where 1 +DROP TABLE t1; +CREATE TABLE t1(a TIMESTAMP NOT NULL); +INSERT INTO t1 VALUES('2001-01-01'); +SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +a +2001-01-01 00:00:00 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01' AND a='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a` from `test`.`t1` where 1 +DROP TABLE t1; +CREATE TABLE t1(a DATETIME NOT NULL, b DATE NOT NULL); +INSERT INTO t1 VALUES('2001-01-01', '2001-01-01'); +SELECT * FROM t1 WHERE a='2001-01-01' AND a=b AND b='2001-01-01 00:00:00'; +a b +2001-01-01 00:00:00 2001-01-01 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01' AND a=b AND b='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a`,'2001-01-01' AS `b` from `test`.`t1` where 1 +DROP TABLE t1; +CREATE TABLE t1(a DATETIME NOT NULL, b VARCHAR(20) NOT NULL); +INSERT INTO t1 VALUES('2001-01-01', '2001-01-01'); +SELECT * FROM t1 WHERE a='2001-01-01' AND a=b AND b='2001-01-01 00:00:00'; +a b +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01' AND a=b AND b='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE NULL NULL NULL NULL NULL NULL NULL NULL Impossible WHERE noticed after reading const tables +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a`,'2001-01-01' AS `b` from `test`.`t1` where 0 +SELECT * FROM t1 WHERE a='2001-01-01 00:00:00' AND a=b AND b='2001-01-01'; +a b +2001-01-01 00:00:00 2001-01-01 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE a='2001-01-01 00:00:00' AND a=b AND b='2001-01-01'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a`,'2001-01-01' AS `b` from `test`.`t1` where 1 +DROP TABLE t1; +CREATE TABLE t1(a DATETIME NOT NULL, b DATE NOT NULL); +INSERT INTO t1 VALUES('2001-01-01', '2001-01-01'); +SELECT x.a, y.a, z.a FROM t1 x +JOIN t1 y ON x.a=y.a +JOIN t1 z ON y.a=z.a +WHERE x.a='2001-01-01' AND z.a='2001-01-01 00:00:00'; +a a a +2001-01-01 00:00:00 2001-01-01 00:00:00 2001-01-01 00:00:00 +EXPLAIN EXTENDED SELECT x.a, y.a, z.a FROM t1 x +JOIN t1 y ON x.a=y.a +JOIN t1 z ON y.a=z.a +WHERE x.a='2001-01-01' AND z.a='2001-01-01 00:00:00'; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE x system NULL NULL NULL NULL 1 100.00 +1 SIMPLE y system NULL NULL NULL NULL 1 100.00 +1 SIMPLE z system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '2001-01-01 00:00:00' AS `a`,'2001-01-01 00:00:00' AS `a`,'2001-01-01 00:00:00' AS `a` from `test`.`t1` `x` join `test`.`t1` `y` join `test`.`t1` `z` where 1 +DROP TABLE t1; +# +# Bug #49897: crash in ptr_compare when char(0) NOT NULL +# column is used for ORDER BY +# +SET @old_sort_buffer_size= @@session.sort_buffer_size; +SET @@sort_buffer_size= 40000; +CREATE TABLE t1(a CHAR(0) NOT NULL); +INSERT INTO t1 VALUES (0), (0), (0); +INSERT INTO t1 SELECT t11.a FROM t1 t11, t1 t12; +INSERT INTO t1 SELECT t11.a FROM t1 t11, t1 t12; +INSERT INTO t1 SELECT t11.a FROM t1 t11, t1 t12; +EXPLAIN SELECT a FROM t1 ORDER BY a; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 24492 +SELECT a FROM t1 ORDER BY a; +DROP TABLE t1; +CREATE TABLE t1(a CHAR(0) NOT NULL, b CHAR(0) NOT NULL, c int); +INSERT INTO t1 VALUES (0, 0, 0), (0, 0, 2), (0, 0, 1); +INSERT INTO t1 SELECT t11.a, t11.b, t11.c FROM t1 t11, t1 t12; +INSERT INTO t1 SELECT t11.a, t11.b, t11.c FROM t1 t11, t1 t12; +INSERT INTO t1 SELECT t11.a, t11.b, t11.c FROM t1 t11, t1 t12; +EXPLAIN SELECT a FROM t1 ORDER BY a LIMIT 5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 24492 +SELECT a FROM t1 ORDER BY a LIMIT 5; +a + + + + + +EXPLAIN SELECT * FROM t1 ORDER BY a, b LIMIT 5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 24492 +SELECT * FROM t1 ORDER BY a, b LIMIT 5; +a b c + 0 + 2 + 1 + 0 + 2 +EXPLAIN SELECT * FROM t1 ORDER BY a, b, c LIMIT 5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 24492 Using filesort +SELECT * FROM t1 ORDER BY a, b, c LIMIT 5; +a b c + 0 + 0 + 0 + 0 + 0 +EXPLAIN SELECT * FROM t1 ORDER BY c, a LIMIT 5; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 ALL NULL NULL NULL NULL 24492 Using filesort +SELECT * FROM t1 ORDER BY c, a LIMIT 5; +a b c + 0 + 0 + 0 + 0 + 0 +SET @@sort_buffer_size= @old_sort_buffer_size; +DROP TABLE t1; +End of 5.0 tests +create table t1(a INT, KEY (a)); +INSERT INTO t1 VALUES (1),(2),(3),(4),(5); +SELECT a FROM t1 ORDER BY a LIMIT 2; +a +1 +2 +SELECT a FROM t1 ORDER BY a LIMIT 2,4294967296; +a +3 +4 +5 +SELECT a FROM t1 ORDER BY a LIMIT 2,4294967297; +a +3 +4 +5 +DROP TABLE t1; +CREATE TABLE A (date_key date); +CREATE TABLE C ( +pk int, +int_nokey int, +int_key int, +date_key date NOT NULL, +date_nokey date, +varchar_key varchar(1) +); +INSERT INTO C VALUES +(1,1,1,'0000-00-00',NULL,NULL), +(1,1,1,'0000-00-00',NULL,NULL); +SELECT 1 FROM C WHERE pk > ANY (SELECT 1 FROM C); +1 +SELECT COUNT(DISTINCT 1) FROM C +WHERE date_key = (SELECT 1 FROM A WHERE C.date_key IS NULL) GROUP BY pk; +COUNT(DISTINCT 1) +SELECT date_nokey FROM C +WHERE int_key IN (SELECT 1 FROM A) +HAVING date_nokey = '10:41:7' +ORDER BY date_key; +date_nokey +Warnings: +Warning 1292 Incorrect date value: '10:41:7' for column 'date_nokey' at row 1 +DROP TABLE A,C; +CREATE TABLE t1 (a INT NOT NULL, b INT); +INSERT INTO t1 VALUES (1, 1); +EXPLAIN EXTENDED SELECT * FROM t1 WHERE (a=a AND a=a) OR b > 2; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 1 100.00 +Warnings: +Note 1003 select '1' AS `a`,'1' AS `b` from `test`.`t1` where 1 +SELECT * FROM t1 WHERE (a=a AND a=a) OR b > 2; +a b +1 1 +DROP TABLE t1; +CREATE TABLE t1 (a INT NOT NULL, b INT NOT NULL, c INT NOT NULL); +EXPLAIN EXTENDED SELECT * FROM t1 WHERE (a=a AND b=b AND c=c) OR b > 20; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 0 0.00 const row not found +Warnings: +Note 1003 select '0' AS `a`,'0' AS `b`,'0' AS `c` from `test`.`t1` where 1 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE (a=a AND a=a AND b=b) OR b > 20; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 0 0.00 const row not found +Warnings: +Note 1003 select '0' AS `a`,'0' AS `b`,'0' AS `c` from `test`.`t1` where 1 +EXPLAIN EXTENDED SELECT * FROM t1 WHERE (a=a AND b=b AND a=a) OR b > 20; +id select_type table type possible_keys key key_len ref rows filtered Extra +1 SIMPLE t1 system NULL NULL NULL NULL 0 0.00 const row not found +Warnings: +Note 1003 select '0' AS `a`,'0' AS `b`,'0' AS `c` from `test`.`t1` where 1 +DROP TABLE t1; +# +# Bug#45266: Uninitialized variable lead to an empty result. +# +drop table if exists A,AA,B,BB; +CREATE TABLE `A` ( +`pk` int(11) NOT NULL AUTO_INCREMENT, +`date_key` date NOT NULL, +`date_nokey` date NOT NULL, +`datetime_key` datetime NOT NULL, +`int_nokey` int(11) NOT NULL, +`time_key` time NOT NULL, +`time_nokey` time NOT NULL, +PRIMARY KEY (`pk`), +KEY `date_key` (`date_key`), +KEY `time_key` (`time_key`), +KEY `datetime_key` (`datetime_key`) +); +CREATE TABLE `AA` ( +`pk` int(11) NOT NULL AUTO_INCREMENT, +`int_nokey` int(11) NOT NULL, +`time_key` time NOT NULL, +KEY `time_key` (`time_key`), +PRIMARY KEY (`pk`) +); +CREATE TABLE `B` ( +`date_nokey` date NOT NULL, +`date_key` date NOT NULL, +`time_key` time NOT NULL, +`datetime_nokey` datetime NOT NULL, +`varchar_key` varchar(1) NOT NULL, +KEY `date_key` (`date_key`), +KEY `time_key` (`time_key`), +KEY `varchar_key` (`varchar_key`) +); +INSERT INTO `B` VALUES ('2003-07-28','2003-07-28','15:13:38','0000-00-00 00:00:00','f'),('0000-00-00','0000-00-00','00:05:48','2004-07-02 14:34:13','x'); +CREATE TABLE `BB` ( +`pk` int(11) NOT NULL AUTO_INCREMENT, +`int_nokey` int(11) NOT NULL, +`date_key` date NOT NULL, +`varchar_nokey` varchar(1) NOT NULL, +`date_nokey` date NOT NULL, +PRIMARY KEY (`pk`), +KEY `date_key` (`date_key`) +); +INSERT INTO `BB` VALUES (10,8,'0000-00-00','i','0000-00-00'),(11,0,'2005-08-18','','2005-08-18'); +SELECT table1 . `pk` AS field1 +FROM +(BB AS table1 INNER JOIN +(AA AS table2 STRAIGHT_JOIN A AS table3 +ON ( table3 . `date_key` = table2 . `pk` )) +ON ( table3 . `datetime_key` = table2 . `int_nokey` )) +WHERE ( table3 . `date_key` <= 4 AND table2 . `pk` = table1 . `varchar_nokey`) +GROUP BY field1 ; +field1 +SELECT table3 .`date_key` field1 +FROM +B table1 LEFT JOIN B table3 JOIN +(BB table6 JOIN A table7 ON table6 .`varchar_nokey`) +ON table6 .`int_nokey` ON table6 .`date_key` + WHERE NOT ( table1 .`varchar_key` AND table7 .`pk`) GROUP BY field1; +field1 +NULL +SELECT table4 . `time_nokey` AS field1 FROM +(AA AS table1 CROSS JOIN +(AA AS table2 STRAIGHT_JOIN +(B AS table3 STRAIGHT_JOIN A AS table4 +ON ( table4 . `date_key` = table3 . `time_key` )) +ON ( table4 . `pk` = table3 . `date_nokey` )) +ON ( table4 . `time_key` = table3 . `datetime_nokey` )) +WHERE ( table4 . `time_key` < table1 . `time_key` AND +table1 . `int_nokey` != 'f') +GROUP BY field1 ORDER BY field1 , field1; +field1 +SELECT table1 .`time_key` field2 FROM B table1 LEFT JOIN BB JOIN A table5 ON table5 .`date_nokey` ON table5 .`int_nokey` GROUP BY field2; +field2 +00:05:48 +15:13:38 +drop table A,AA,B,BB; +#end of test for bug#45266 +# +# BUG#48052: Valgrind warning - uninitialized value in init_read_record() +# +CREATE TABLE t1 ( +pk int(11) NOT NULL, +i int(11) DEFAULT NULL, +v varchar(1) DEFAULT NULL, +PRIMARY KEY (pk) +); +INSERT INTO t1 VALUES (2,7,'m'); +INSERT INTO t1 VALUES (3,9,'m'); +SELECT v +FROM t1 +WHERE NOT pk > 0 +HAVING v <= 't' +ORDER BY pk; +v +DROP TABLE t1; +# +# Bug#49489 Uninitialized cache led to a wrong result. +# +CREATE TABLE t1(c1 DOUBLE(5,4)); +INSERT INTO t1 VALUES (9.1234); +SELECT * FROM t1 WHERE c1 < 9.12345; +c1 +9.1234 +DROP TABLE t1; +# End of test for bug#49489. +# +# Bug #49517: Inconsistent behavior while using +# NULLable BIGINT and INT columns in comparison +# +CREATE TABLE t1(a BIGINT UNSIGNED NOT NULL, b BIGINT NULL, c INT NULL); +INSERT INTO t1 VALUES(105, NULL, NULL); +SELECT * FROM t1 WHERE b < 102; +a b c +SELECT * FROM t1 WHERE c < 102; +a b c +SELECT * FROM t1 WHERE 102 < b; +a b c +SELECT * FROM t1 WHERE 102 < c; +a b c +DROP TABLE t1; +End of 5.1 tests === modified file 'mysql-test/r/sp.result' --- a/mysql-test/r/sp.result 2009-12-23 13:44:03 +0000 +++ b/mysql-test/r/sp.result 2010-04-08 12:10:05 +0000 @@ -6940,9 +6940,8 @@ create procedure p() begin end; call p(); select @@sql_mode; @@sql_mode -REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,?,ONLY_FULL_GROUP_BY,NO_UNSIGNED_SUBTRACTION,NO_DIR_IN_CREATE,POSTGRESQL,ORACLE,MSSQL,DB2,MAXDB,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,MYSQL323,MYSQL40,ANSI,NO_AUTO_VALUE_ON_ZERO,NO_BACKSLASH_ESCAPES,STRICT_TRANS_TABLES,STRICT_ALL_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ALLOW_INVALID_DATES,ERROR_FOR_DIVISION_BY_ZERO,TRADITIONAL,NO_AUTO_CREATE_USER,HIGH_NOT_PRECEDENCE,NO_ENGINE_SUBSTITUTION,PAD_CHAR_TO_FULL_LENGTH +REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,IGNORE_BAD_TABLE_OPTIONS,ONLY_FULL_GROUP_BY,NO_UNSIGNED_SUBTRACTION,NO_DIR_IN_CREATE,POSTGRESQL,ORACLE,MSSQL,DB2,MAXDB,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,MYSQL323,MYSQL40,ANSI,NO_AUTO_VALUE_ON_ZERO,NO_BACKSLASH_ESCAPES,STRICT_TRANS_TABLES,STRICT_ALL_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ALLOW_INVALID_DATES,ERROR_FOR_DIVISION_BY_ZERO,TRADITIONAL,NO_AUTO_CREATE_USER,HIGH_NOT_PRECEDENCE,NO_ENGINE_SUBSTITUTION,PAD_CHAR_TO_FULL_LENGTH set @@sql_mode= @old_mode; -select replace(@full_mode, '?', 'NOT_USED') into @full_mode; select replace(@full_mode, 'ALLOW_INVALID_DATES', 'INVALID_DATES') into @full_mode; select name from mysql.proc where name = 'p' and sql_mode = @full_mode; name === modified file 'mysql-test/r/status_user.result' --- a/mysql-test/r/status_user.result 2010-03-21 20:58:19 +0000 +++ b/mysql-test/r/status_user.result 2010-04-02 18:51:02 +0000 @@ -137,7 +137,7 @@ ROWS_INSERTED 8 ROWS_UPDATED 5 SELECT_COMMANDS 3 UPDATE_COMMANDS 11 -OTHER_COMMANDS 9 +OTHER_COMMANDS 7 COMMIT_TRANSACTIONS 19 ROLLBACK_TRANSACTIONS 2 DENIED_CONNECTIONS 0 @@ -154,7 +154,7 @@ ROWS_INSERTED 8 ROWS_UPDATED 5 SELECT_COMMANDS 3 UPDATE_COMMANDS 11 -OTHER_COMMANDS 9 +OTHER_COMMANDS 7 COMMIT_TRANSACTIONS 19 ROLLBACK_TRANSACTIONS 2 DENIED_CONNECTIONS 0 === modified file 'mysql-test/r/system_mysql_db.result' --- a/mysql-test/r/system_mysql_db.result 2010-03-29 15:13:53 +0000 +++ b/mysql-test/r/system_mysql_db.result 2010-04-08 12:10:05 +0000 @@ -202,7 +202,7 @@ proc CREATE TABLE `proc` ( `definer` char(77) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL DEFAULT '', `created` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP, `modified` timestamp NOT NULL DEFAULT '0000-00-00 00:00:00', - `sql_mode` set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NOT NULL DEFAULT '', + `sql_mode` set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NOT NULL DEFAULT '', `comment` char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL DEFAULT '', `character_set_client` char(32) CHARACTER SET utf8 COLLATE utf8_bin DEFAULT NULL, `collation_connection` char(32) CHARACTER SET utf8 COLLATE utf8_bin DEFAULT NULL, @@ -227,7 +227,7 @@ event CREATE TABLE `event` ( `ends` datetime DEFAULT NULL, `status` enum('ENABLED','DISABLED','SLAVESIDE_DISABLED') NOT NULL DEFAULT 'ENABLED', `on_completion` enum('DROP','PRESERVE') NOT NULL DEFAULT 'DROP', - `sql_mode` set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NOT NULL DEFAULT '', + `sql_mode` set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NOT NULL DEFAULT '', `comment` char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL DEFAULT '', `originator` int(10) unsigned NOT NULL, `time_zone` char(64) CHARACTER SET latin1 NOT NULL DEFAULT 'SYSTEM', === modified file 'mysql-test/suite/binlog/r/binlog_index.result' --- a/mysql-test/suite/binlog/r/binlog_index.result 2010-03-20 14:12:53 +0000 +++ b/mysql-test/suite/binlog/r/binlog_index.result 2010-04-08 09:52:53 +0000 @@ -3,6 +3,7 @@ call mtr.add_suppression('MSYQL_BIN_LOG: call mtr.add_suppression('MSYQL_BIN_LOG::open failed to sync the index file'); call mtr.add_suppression('Turning logging off for the whole duration of the MySQL server process.'); call mtr.add_suppression('MSYQL_BIN_LOG::purge_logs failed to clean registers before purging logs.'); +flush tables; flush logs; flush logs; flush logs; === added file 'mysql-test/suite/binlog/t/binlog_index-master.opt' --- a/mysql-test/suite/binlog/t/binlog_index-master.opt 1970-01-01 00:00:00 +0000 +++ b/mysql-test/suite/binlog/t/binlog_index-master.opt 2010-04-06 22:47:08 +0000 @@ -0,0 +1 @@ +--skip-stack-trace --test-expect-abort --log-warnings=0 === modified file 'mysql-test/suite/binlog/t/binlog_index.test' --- a/mysql-test/suite/binlog/t/binlog_index.test 2010-03-20 14:12:53 +0000 +++ b/mysql-test/suite/binlog/t/binlog_index.test 2010-04-08 09:52:53 +0000 @@ -11,6 +11,8 @@ call mtr.add_suppression('MSYQL_BIN_LOG: call mtr.add_suppression('MSYQL_BIN_LOG::open failed to sync the index file'); call mtr.add_suppression('Turning logging off for the whole duration of the MySQL server process.'); call mtr.add_suppression('MSYQL_BIN_LOG::purge_logs failed to clean registers before purging logs.'); +flush tables; + let $old=`select @@debug`; let $MYSQLD_DATADIR= `select @@datadir`; === modified file 'mysql-test/suite/funcs_1/datadict/processlist_priv.inc' --- a/mysql-test/suite/funcs_1/datadict/processlist_priv.inc 2009-10-10 09:59:06 +0000 +++ b/mysql-test/suite/funcs_1/datadict/processlist_priv.inc 2010-04-07 15:26:30 +0000 @@ -89,18 +89,38 @@ USE information_schema; --echo 1 Prepare test. --echo connection default (user=root) --echo #################################################################################### -if (`SELECT COUNT(*) <> 1 FROM processlist`) + +# Check that we have only one connection around. +# If there is more, it may be the check() connection that has not yet ended, +# so we wait for it for up to 10 seconds. + +let $counter=100; +while ($counter) { - --echo This test expects one connection to the server. - --echo Expectation: USER HOST DB COMMAND STATE INFO - --echo Expectation: root localhost information_schema Query executing SELECT USER,HOST,DB,COMMAND,STATE,INFO FROM processlist ORDER BY ID - --echo But we found in the moment: - SELECT USER,HOST,DB,COMMAND,STATE,INFO FROM processlist ORDER BY ID; - --echo Maybe - --echo - the base configuration (no of parallel auxiliary sessions) of the server has changed - --echo - a parallel test intended for another server accidently connected to our current one - --echo We cannot proceed in this situation. Abort - exit; + dec $counter; + let $res= `SELECT COUNT(*) <> 1 FROM processlist`; + if (!$res) + { + # Success; Abort while loop + let $counter=0; + } + if ($res) + { + --sleep 0.1 + if (!$counter) + { + --echo This test expects one connection to the server. + --echo Expectation: USER HOST DB COMMAND STATE INFO + --echo Expectation: root localhost information_schema Query executing SELECT USER,HOST,DB,COMMAND,STATE,INFO FROM processlist ORDER BY ID + --echo But we found in the moment: + SELECT USER,HOST,DB,COMMAND,STATE,INFO FROM processlist ORDER BY ID; + --echo Maybe + --echo - the base configuration (no of parallel auxiliary sessions) of the server has changed + --echo - a parallel test intended for another server accidently connected to our current one + --echo We cannot proceed in this situation. Abort + exit; + } + } } --echo #################################################################################### === modified file 'mysql-test/suite/funcs_1/r/is_columns_is.result' --- a/mysql-test/suite/funcs_1/r/is_columns_is.result 2010-02-01 06:14:12 +0000 +++ b/mysql-test/suite/funcs_1/r/is_columns_is.result 2010-04-06 02:08:10 +0000 @@ -230,6 +230,18 @@ NULL information_schema INNODB_TRX trx_s NULL information_schema INNODB_TRX trx_state 2 NO varchar 13 39 NULL NULL utf8 utf8_general_ci varchar(13) select NULL information_schema INNODB_TRX trx_wait_started 5 NULL YES datetime NULL NULL NULL NULL NULL NULL datetime select NULL information_schema INNODB_TRX trx_weight 6 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES BLOCK_SIZE 5 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES DIRTY_BLOCKS 8 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES FULL_SIZE 4 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES KEY_CACHE_NAME 1 NO varchar 192 576 NULL NULL utf8 utf8_general_ci varchar(192) select +NULL information_schema KEY_CACHES PARTITIONS 2 NULL YES int NULL NULL 10 0 NULL NULL int(3) unsigned select +NULL information_schema KEY_CACHES PARTITION_NUMBER 3 NULL YES int NULL NULL 10 0 NULL NULL int(3) unsigned select +NULL information_schema KEY_CACHES READS 10 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES READ_REQUESTS 9 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES UNUSED_BLOCKS 7 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES USED_BLOCKS 6 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES WRITES 12 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select +NULL information_schema KEY_CACHES WRITE_REQUESTS 11 0 NO bigint NULL NULL 19 0 NULL NULL bigint(21) unsigned select NULL information_schema KEY_COLUMN_USAGE COLUMN_NAME 7 NO varchar 64 192 NULL NULL utf8 utf8_general_ci varchar(64) select NULL information_schema KEY_COLUMN_USAGE CONSTRAINT_CATALOG 1 NULL YES varchar 512 1536 NULL NULL utf8 utf8_general_ci varchar(512) select NULL information_schema KEY_COLUMN_USAGE CONSTRAINT_NAME 3 NO varchar 64 192 NULL NULL utf8 utf8_general_ci varchar(64) select @@ -271,10 +283,12 @@ NULL information_schema PBXT_STATISTICS NULL information_schema PBXT_STATISTICS Name 2 NO varchar 40 120 NULL NULL utf8 utf8_general_ci varchar(40) select NULL information_schema PBXT_STATISTICS Value 3 0 NO bigint NULL NULL 19 0 NULL NULL bigint(8) select NULL information_schema PLUGINS PLUGIN_AUTHOR 8 NULL YES varchar 64 192 NULL NULL utf8 utf8_general_ci varchar(64) select +NULL information_schema PLUGINS PLUGIN_AUTH_VERSION 12 NULL YES varchar 80 240 NULL NULL utf8 utf8_general_ci varchar(80) select NULL information_schema PLUGINS PLUGIN_DESCRIPTION 9 NULL YES longtext 4294967295 4294967295 NULL NULL utf8 utf8_general_ci longtext select NULL information_schema PLUGINS PLUGIN_LIBRARY 6 NULL YES varchar 64 192 NULL NULL utf8 utf8_general_ci varchar(64) select NULL information_schema PLUGINS PLUGIN_LIBRARY_VERSION 7 NULL YES varchar 20 60 NULL NULL utf8 utf8_general_ci varchar(20) select NULL information_schema PLUGINS PLUGIN_LICENSE 10 NULL YES varchar 80 240 NULL NULL utf8 utf8_general_ci varchar(80) select +NULL information_schema PLUGINS PLUGIN_MATURITY 11 NULL YES varchar 12 36 NULL NULL utf8 utf8_general_ci varchar(12) select NULL information_schema PLUGINS PLUGIN_NAME 1 NO varchar 64 192 NULL NULL utf8 utf8_general_ci varchar(64) select NULL information_schema PLUGINS PLUGIN_STATUS 3 NO varchar 10 30 NULL NULL utf8 utf8_general_ci varchar(10) select NULL information_schema PLUGINS PLUGIN_TYPE 4 NO varchar 80 240 NULL NULL utf8 utf8_general_ci varchar(80) select @@ -741,6 +755,18 @@ NULL information_schema INNODB_TRX trx_w NULL information_schema INNODB_TRX trx_weight bigint NULL NULL NULL NULL bigint(21) unsigned NULL information_schema INNODB_TRX trx_mysql_thread_id bigint NULL NULL NULL NULL bigint(21) unsigned 3.0000 information_schema INNODB_TRX trx_query varchar 1024 3072 utf8 utf8_general_ci varchar(1024) +3.0000 information_schema KEY_CACHES KEY_CACHE_NAME varchar 192 576 utf8 utf8_general_ci varchar(192) +NULL information_schema KEY_CACHES PARTITIONS int NULL NULL NULL NULL int(3) unsigned +NULL information_schema KEY_CACHES PARTITION_NUMBER int NULL NULL NULL NULL int(3) unsigned +NULL information_schema KEY_CACHES FULL_SIZE bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES BLOCK_SIZE bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES USED_BLOCKS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES UNUSED_BLOCKS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES DIRTY_BLOCKS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES READ_REQUESTS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES READS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES WRITE_REQUESTS bigint NULL NULL NULL NULL bigint(21) unsigned +NULL information_schema KEY_CACHES WRITES bigint NULL NULL NULL NULL bigint(21) unsigned 3.0000 information_schema KEY_COLUMN_USAGE CONSTRAINT_CATALOG varchar 512 1536 utf8 utf8_general_ci varchar(512) 3.0000 information_schema KEY_COLUMN_USAGE CONSTRAINT_SCHEMA varchar 64 192 utf8 utf8_general_ci varchar(64) 3.0000 information_schema KEY_COLUMN_USAGE CONSTRAINT_NAME varchar 64 192 utf8 utf8_general_ci varchar(64) @@ -791,6 +817,8 @@ NULL information_schema PBXT_STATISTICS 3.0000 information_schema PLUGINS PLUGIN_AUTHOR varchar 64 192 utf8 utf8_general_ci varchar(64) 1.0000 information_schema PLUGINS PLUGIN_DESCRIPTION longtext 4294967295 4294967295 utf8 utf8_general_ci longtext 3.0000 information_schema PLUGINS PLUGIN_LICENSE varchar 80 240 utf8 utf8_general_ci varchar(80) +3.0000 information_schema PLUGINS PLUGIN_MATURITY varchar 12 36 utf8 utf8_general_ci varchar(12) +3.0000 information_schema PLUGINS PLUGIN_AUTH_VERSION varchar 80 240 utf8 utf8_general_ci varchar(80) NULL information_schema PROCESSLIST ID bigint NULL NULL NULL NULL bigint(4) 3.0000 information_schema PROCESSLIST USER varchar 16 48 utf8 utf8_general_ci varchar(16) 3.0000 information_schema PROCESSLIST HOST varchar 64 192 utf8 utf8_general_ci varchar(64) === modified file 'mysql-test/suite/funcs_1/r/is_columns_mysql.result' --- a/mysql-test/suite/funcs_1/r/is_columns_mysql.result 2010-03-29 15:13:53 +0000 +++ b/mysql-test/suite/funcs_1/r/is_columns_mysql.result 2010-04-08 17:19:01 +0000 @@ -49,7 +49,7 @@ NULL mysql event modified 9 0000-00-00 0 NULL mysql event name 2 NO char 64 192 NULL NULL utf8 utf8_general_ci char(64) PRI select,insert,update,references NULL mysql event on_completion 14 DROP NO enum 8 24 NULL NULL utf8 utf8_general_ci enum('DROP','PRESERVE') select,insert,update,references NULL mysql event originator 17 NULL NO int NULL NULL 10 0 NULL NULL int(10) unsigned select,insert,update,references -NULL mysql event sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') select,insert,update,references +NULL mysql event sql_mode 15 NO set 494 1482 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') select,insert,update,references NULL mysql event starts 11 NULL YES datetime NULL NULL NULL NULL NULL NULL datetime select,insert,update,references NULL mysql event status 13 ENABLED NO enum 18 54 NULL NULL utf8 utf8_general_ci enum('ENABLED','DISABLED','SLAVESIDE_DISABLED') select,insert,update,references NULL mysql event time_zone 18 SYSTEM NO char 64 64 NULL NULL latin1 latin1_swedish_ci char(64) select,insert,update,references @@ -124,7 +124,7 @@ NULL mysql proc returns 10 NULL NO longb NULL mysql proc security_type 8 DEFINER NO enum 7 21 NULL NULL utf8 utf8_general_ci enum('INVOKER','DEFINER') select,insert,update,references NULL mysql proc specific_name 4 NO char 64 192 NULL NULL utf8 utf8_general_ci char(64) select,insert,update,references NULL mysql proc sql_data_access 6 CONTAINS_SQL NO enum 17 51 NULL NULL utf8 utf8_general_ci enum('CONTAINS_SQL','NO_SQL','READS_SQL_DATA','MODIFIES_SQL_DATA') select,insert,update,references -NULL mysql proc sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') select,insert,update,references +NULL mysql proc sql_mode 15 NO set 494 1482 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') select,insert,update,references NULL mysql proc type 3 NULL NO enum 9 27 NULL NULL utf8 utf8_general_ci enum('FUNCTION','PROCEDURE') PRI select,insert,update,references NULL mysql procs_priv Db 2 NO char 64 192 NULL NULL utf8 utf8_bin char(64) PRI select,insert,update,references NULL mysql procs_priv Grantor 6 NO char 77 231 NULL NULL utf8 utf8_bin char(77) MUL select,insert,update,references @@ -330,7 +330,7 @@ NULL mysql event starts datetime NULL NU NULL mysql event ends datetime NULL NULL NULL NULL datetime 3.0000 mysql event status enum 18 54 utf8 utf8_general_ci enum('ENABLED','DISABLED','SLAVESIDE_DISABLED') 3.0000 mysql event on_completion enum 8 24 utf8 utf8_general_ci enum('DROP','PRESERVE') -3.0000 mysql event sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +3.0000 mysql event sql_mode set 494 1482 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') 3.0000 mysql event comment char 64 192 utf8 utf8_bin char(64) NULL mysql event originator int NULL NULL NULL NULL int(10) unsigned 1.0000 mysql event time_zone char 64 64 latin1 latin1_swedish_ci char(64) @@ -405,7 +405,7 @@ NULL mysql ndb_binlog_index schemaops bi 3.0000 mysql proc definer char 77 231 utf8 utf8_bin char(77) NULL mysql proc created timestamp NULL NULL NULL NULL timestamp NULL mysql proc modified timestamp NULL NULL NULL NULL timestamp -3.0000 mysql proc sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +3.0000 mysql proc sql_mode set 494 1482 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') 3.0000 mysql proc comment char 64 192 utf8 utf8_bin char(64) 3.0000 mysql proc character_set_client char 32 96 utf8 utf8_bin char(32) 3.0000 mysql proc collation_connection char 32 96 utf8 utf8_bin char(32) === modified file 'mysql-test/suite/funcs_1/r/is_columns_mysql_embedded.result' --- a/mysql-test/suite/funcs_1/r/is_columns_mysql_embedded.result 2009-05-19 16:43:50 +0000 +++ b/mysql-test/suite/funcs_1/r/is_columns_mysql_embedded.result 2010-04-08 12:10:05 +0000 @@ -49,7 +49,7 @@ NULL mysql event modified 9 0000-00-00 0 NULL mysql event name 2 NO char 64 192 NULL NULL utf8 utf8_general_ci char(64) PRI NULL mysql event on_completion 14 DROP NO enum 8 24 NULL NULL utf8 utf8_general_ci enum('DROP','PRESERVE') NULL mysql event originator 17 NULL NO int NULL NULL 10 0 NULL NULL int(10) unsigned -NULL mysql event sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +NULL mysql event sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NULL mysql event starts 11 NULL YES datetime NULL NULL NULL NULL NULL NULL datetime NULL mysql event status 13 ENABLED NO enum 18 54 NULL NULL utf8 utf8_general_ci enum('ENABLED','DISABLED','SLAVESIDE_DISABLED') NULL mysql event time_zone 18 SYSTEM NO char 64 64 NULL NULL latin1 latin1_swedish_ci char(64) @@ -124,7 +124,7 @@ NULL mysql proc returns 10 NULL NO longb NULL mysql proc security_type 8 DEFINER NO enum 7 21 NULL NULL utf8 utf8_general_ci enum('INVOKER','DEFINER') NULL mysql proc specific_name 4 NO char 64 192 NULL NULL utf8 utf8_general_ci char(64) NULL mysql proc sql_data_access 6 CONTAINS_SQL NO enum 17 51 NULL NULL utf8 utf8_general_ci enum('CONTAINS_SQL','NO_SQL','READS_SQL_DATA','MODIFIES_SQL_DATA') -NULL mysql proc sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +NULL mysql proc sql_mode 15 NO set 478 1434 NULL NULL utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') NULL mysql proc type 3 NULL NO enum 9 27 NULL NULL utf8 utf8_general_ci enum('FUNCTION','PROCEDURE') PRI NULL mysql procs_priv Db 2 NO char 64 192 NULL NULL utf8 utf8_bin char(64) PRI NULL mysql procs_priv Grantor 6 NO char 77 231 NULL NULL utf8 utf8_bin char(77) MUL @@ -327,7 +327,7 @@ NULL mysql event starts datetime NULL NU NULL mysql event ends datetime NULL NULL NULL NULL datetime 3.0000 mysql event status enum 18 54 utf8 utf8_general_ci enum('ENABLED','DISABLED','SLAVESIDE_DISABLED') 3.0000 mysql event on_completion enum 8 24 utf8 utf8_general_ci enum('DROP','PRESERVE') -3.0000 mysql event sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +3.0000 mysql event sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') 3.0000 mysql event comment char 64 192 utf8 utf8_bin char(64) NULL mysql event originator int NULL NULL NULL NULL int(10) unsigned 1.0000 mysql event time_zone char 64 64 latin1 latin1_swedish_ci char(64) @@ -402,7 +402,7 @@ NULL mysql ndb_binlog_index schemaops bi 3.0000 mysql proc definer char 77 231 utf8 utf8_bin char(77) NULL mysql proc created timestamp NULL NULL NULL NULL timestamp NULL mysql proc modified timestamp NULL NULL NULL NULL timestamp -3.0000 mysql proc sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') +3.0000 mysql proc sql_mode set 478 1434 utf8 utf8_general_ci set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') 3.0000 mysql proc comment char 64 192 utf8 utf8_bin char(64) 3.0000 mysql proc character_set_client char 32 96 utf8 utf8_bin char(32) 3.0000 mysql proc collation_connection char 32 96 utf8 utf8_bin char(32) === modified file 'mysql-test/suite/funcs_1/r/is_tables_is.result' --- a/mysql-test/suite/funcs_1/r/is_tables_is.result 2010-02-01 06:14:12 +0000 +++ b/mysql-test/suite/funcs_1/r/is_tables_is.result 2010-04-06 02:08:10 +0000 @@ -590,6 +590,29 @@ user_comment Separator ----------------------------------------------------- TABLE_CATALOG NULL TABLE_SCHEMA information_schema +TABLE_NAME KEY_CACHES +TABLE_TYPE SYSTEM VIEW +ENGINE MEMORY +VERSION 10 +ROW_FORMAT Fixed +TABLE_ROWS #TBLR# +AVG_ROW_LENGTH #ARL# +DATA_LENGTH #DL# +MAX_DATA_LENGTH #MDL# +INDEX_LENGTH #IL# +DATA_FREE #DF# +AUTO_INCREMENT NULL +CREATE_TIME #CRT# +UPDATE_TIME #UT# +CHECK_TIME #CT# +TABLE_COLLATION utf8_general_ci +CHECKSUM NULL +CREATE_OPTIONS #CO# +TABLE_COMMENT #TC# +user_comment +Separator ----------------------------------------------------- +TABLE_CATALOG NULL +TABLE_SCHEMA information_schema TABLE_NAME KEY_COLUMN_USAGE TABLE_TYPE SYSTEM VIEW ENGINE MEMORY @@ -1688,6 +1711,29 @@ user_comment Separator ----------------------------------------------------- TABLE_CATALOG NULL TABLE_SCHEMA information_schema +TABLE_NAME KEY_CACHES +TABLE_TYPE SYSTEM VIEW +ENGINE MEMORY +VERSION 10 +ROW_FORMAT Fixed +TABLE_ROWS #TBLR# +AVG_ROW_LENGTH #ARL# +DATA_LENGTH #DL# +MAX_DATA_LENGTH #MDL# +INDEX_LENGTH #IL# +DATA_FREE #DF# +AUTO_INCREMENT NULL +CREATE_TIME #CRT# +UPDATE_TIME #UT# +CHECK_TIME #CT# +TABLE_COLLATION utf8_general_ci +CHECKSUM NULL +CREATE_OPTIONS #CO# +TABLE_COMMENT #TC# +user_comment +Separator ----------------------------------------------------- +TABLE_CATALOG NULL +TABLE_SCHEMA information_schema TABLE_NAME KEY_COLUMN_USAGE TABLE_TYPE SYSTEM VIEW ENGINE MEMORY === modified file 'mysql-test/suite/pbxt/r/multi_statement.result' --- a/mysql-test/suite/pbxt/r/multi_statement.result 2009-04-02 10:03:14 +0000 +++ b/mysql-test/suite/pbxt/r/multi_statement.result 2010-04-06 22:47:08 +0000 @@ -47,5 +47,5 @@ i 3 show status like 'Slow_queries'|||| Variable_name Value -Slow_queries 0 +Slow_queries 3 drop table t1|||| === modified file 'mysql-test/suite/pbxt/r/mysqlshow.result' --- a/mysql-test/suite/pbxt/r/mysqlshow.result 2010-02-01 06:14:12 +0000 +++ b/mysql-test/suite/pbxt/r/mysqlshow.result 2010-03-31 18:45:41 +0000 @@ -91,6 +91,7 @@ Database: information_schema | GLOBAL_STATUS | | GLOBAL_VARIABLES | | INDEX_STATISTICS | +| KEY_CACHES | | KEY_COLUMN_USAGE | | PARTITIONS | | PLUGINS | @@ -144,6 +145,7 @@ Database: INFORMATION_SCHEMA | GLOBAL_STATUS | | GLOBAL_VARIABLES | | INDEX_STATISTICS | +| KEY_CACHES | | KEY_COLUMN_USAGE | | PARTITIONS | | PLUGINS | === added file 'mysql-test/suite/pbxt/t/multi_statement-master.opt' --- a/mysql-test/suite/pbxt/t/multi_statement-master.opt 1970-01-01 00:00:00 +0000 +++ b/mysql-test/suite/pbxt/t/multi_statement-master.opt 2010-04-06 22:47:08 +0000 @@ -0,0 +1,3 @@ +--slow-query-log +--slow-query-log-file=slow.log +--log-queries-not-using-indexes === modified file 'mysql-test/suite/pbxt/t/status.test' --- a/mysql-test/suite/pbxt/t/status.test 2009-04-02 10:03:14 +0000 +++ b/mysql-test/suite/pbxt/t/status.test 2010-04-07 15:26:30 +0000 @@ -30,7 +30,7 @@ connection con1; --send update t1 set n = 3; connection con2; -sleep 0.5; +sleep 5; unlock tables; connection con1; reap; === modified file 'mysql-test/suite/rpl/r/rpl_optimize.result' --- a/mysql-test/suite/rpl/r/rpl_optimize.result 2010-01-13 09:00:03 +0000 +++ b/mysql-test/suite/rpl/r/rpl_optimize.result 2010-03-31 19:12:21 +0000 @@ -4,6 +4,7 @@ reset master; reset slave; drop table if exists t1,t2,t3,t4,t5,t6,t7,t8,t9; start slave; +drop tables if exists t1; CALL mtr.add_suppression('Statement may not be safe to log in statement format.'); create table t1 (a int not null auto_increment primary key, b int, key(b)); INSERT INTO t1 (a) VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10); === modified file 'mysql-test/suite/rpl/t/rpl_optimize.test' --- a/mysql-test/suite/rpl/t/rpl_optimize.test 2010-03-04 08:03:07 +0000 +++ b/mysql-test/suite/rpl/t/rpl_optimize.test 2010-03-31 19:12:21 +0000 @@ -15,6 +15,10 @@ -- source include/not_staging.inc -- source include/master-slave.inc +--disable_warnings +drop tables if exists t1; +--enable_warnings + CALL mtr.add_suppression('Statement may not be safe to log in statement format.'); create table t1 (a int not null auto_increment primary key, b int, key(b)); @@ -51,10 +55,9 @@ sync_with_master; # won't work if slave connection master; # cleanup drop table t1; -connection slave; -sync_with_master; +sync_slave_with_master; # If the machine is so fast that slave syncs before OPTIMIZE -# starts, this test wil demonstrate nothing but will pass. +# starts, this test will demonstrate nothing but will pass. # End of 4.1 tests === modified file 'mysql-test/t/connect.test' --- a/mysql-test/t/connect.test 2009-09-07 20:50:10 +0000 +++ b/mysql-test/t/connect.test 2010-04-06 18:36:06 +0000 @@ -328,6 +328,34 @@ if ($error) --disconnect extracon2 --connection default +# +# A couple of plugin tests - for builtin plugins only +# +CREATE USER mysqltest_up1 IDENTIFIED VIA mysql_native_password using '*E8D46CE25265E545D225A8A6F1BAF642FEBEE5CB'; +CREATE USER mysqltest_up2 IDENTIFIED VIA mysql_old_password using '09301740536db389'; + +--replace_result $MASTER_MYSOCK MASTER_SOCKET $MASTER_MYPORT MASTER_PORT +--error ER_ACCESS_DENIED_ERROR +connect(pcon1,localhost,mysqltest_up1,foo,,$MASTER_MYPORT,); +--replace_result $MASTER_MYSOCK MASTER_SOCKET $MASTER_MYPORT MASTER_PORT +connect(pcon2,localhost,mysqltest_up1,bar,,$MASTER_MYPORT,); +connection pcon2; +select user(), current_user(); +disconnect pcon2; + +--replace_result $MASTER_MYSOCK MASTER_SOCKET $MASTER_MYPORT MASTER_PORT +--error ER_ACCESS_DENIED_ERROR +connect(pcon3,localhost,mysqltest_up2,newpw,,$MASTER_MYPORT,); +--replace_result $MASTER_MYSOCK MASTER_SOCKET $MASTER_MYPORT MASTER_PORT +connect(pcon4,localhost,mysqltest_up2,oldpw,,$MASTER_MYPORT,); +connection pcon4; +select user(), current_user(); +disconnect pcon4; + +connection default; +DROP USER mysqltest_up1@'%'; +DROP USER mysqltest_up2@'%'; + # Wait till all disconnects are completed --source include/wait_until_count_sessions.inc === added file 'mysql-test/t/create_options.test' --- a/mysql-test/t/create_options.test 1970-01-01 00:00:00 +0000 +++ b/mysql-test/t/create_options.test 2010-04-08 12:10:05 +0000 @@ -0,0 +1,68 @@ +--disable_warnings +drop table if exists t1; +--enable_warnings + +SET @OLD_SQL_MODE=@@SQL_MODE; +SET SQL_MODE='IGNORE_BAD_TABLE_OPTIONS'; + +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1='1v1'; +show create table t1; +drop table t1; + +--echo #reassiginig options in the same line +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1=1v1 TKEY1=DEFAULT tkey1=1v2 tkey2=2v1; +show create table t1; + +-- echo #add option +alter table t1 tkey4=4v1; +show create table t1; + +--echo #remove options +alter table t1 tkey3=DEFAULT tkey4=DEFAULT; +show create table t1; + +drop table t1; + +create table t1 (a int fkey1=v1, key akey (a) kkey1=v1) tkey1=1v1 tkey1=1v2 TKEY1=DEFAULT tkey2=2v1 tkey3=3v1; +show create table t1; + +--echo #change field with option with the same value +alter table t1 change a a int `FKEY1`='v1'; +show create table t1; +--echo #change field with option with a different value +alter table t1 change a a int fkey1=v2; +show create table t1; +--echo #new column no options +alter table t1 add column b int; +show create table t1; +--echo #new key with options +alter table t1 add key bkey (b) kkey2=v1; +show create table t1; +--echo #new column with options +alter table t1 add column c int fkey1=v1 fkey2=v2; +show create table t1; +--echo #new key no options +alter table t1 add key ckey (c); +show create table t1; +--echo #drop column +alter table t1 drop b; +show create table t1; +--echo #add column with options after delete +alter table t1 add column b int fkey2=v1; +show create table t1; +--echo #add key +alter table t1 add key bkey (b) kkey2=v2; +show create table t1; +drop table t1; + +#numeric (unquoted) value +create table t1 (a int) tkey1=100; +show create table t1; +drop table t1; + +--echo #error on unknown option +SET SQL_MODE=''; +--error ER_UNKNOWN_OPTION +create table t1 (a int fkey=vvv, key akey (a) dff=vvv) tkey1=1v1; + +SET @@SQL_MODE=@OLD_SQL_MODE; === modified file 'mysql-test/t/events_bugs.test' --- a/mysql-test/t/events_bugs.test 2009-03-11 20:30:56 +0000 +++ b/mysql-test/t/events_bugs.test 2010-04-08 12:10:05 +0000 @@ -1204,7 +1204,6 @@ create event e1 on schedule every 1 day select @@sql_mode; set @@sql_mode= @old_mode; # Rename SQL modes that differ in name between the server and the table definition. -select replace(@full_mode, '?', 'NOT_USED') into @full_mode; select replace(@full_mode, 'ALLOW_INVALID_DATES', 'INVALID_DATES') into @full_mode; select name from mysql.event where name = 'p' and sql_mode = @full_mode; drop event e1; === modified file 'mysql-test/t/func_time.test' --- a/mysql-test/t/func_time.test 2009-01-23 12:22:05 +0000 +++ b/mysql-test/t/func_time.test 2010-04-08 09:47:31 +0000 @@ -650,7 +650,7 @@ drop table t1; SET GLOBAL log_bin_trust_function_creators = 0; create table t1 (a datetime, i int, b datetime); -insert into t1 select sysdate(), sleep(1), sysdate() from dual; +insert into t1 select sysdate(), sleep(2), sysdate() from dual; select a != b from t1; drop table t1; === modified file 'mysql-test/t/key_cache.test' --- a/mysql-test/t/key_cache.test 2008-03-27 16:43:17 +0000 +++ b/mysql-test/t/key_cache.test 2010-04-07 15:05:09 +0000 @@ -1,11 +1,13 @@ # -# Test of multiple key caches +# Test of multiple key caches, simple an partitioned # --disable_warnings drop table if exists t1, t2, t3; --enable_warnings -SET @save_key_buffer=@@key_buffer_size; +SET @save_key_buffer_size=@@key_buffer_size; +SET @save_key_cache_block_size=@@key_cache_block_size; +SET @save_key_cache_partitions=@@key_cache_partitions; SELECT @@key_buffer_size, @@small.key_buffer_size; @@ -33,7 +35,7 @@ SELECT @@`default`.key_buffer_size; SELECT @@small.key_buffer_size; SELECT @@medium.key_buffer_size; -SET @@global.key_buffer_size=@save_key_buffer; +SET @@global.key_buffer_size=@save_key_buffer_size; # # Errors @@ -247,3 +249,270 @@ SET GLOBAL key_cache_block_size= @bug284 DROP TABLE t1; # End of 4.1 tests + +# +# Test cases for partitioned key caches +# + +# Test usage of the KEY_CACHE table from information schema +# for a simple key cache + +set global key_buffer_size=@save_key_buffer_size; +set global key_cache_block_size=@save_key_cache_block_size; +select @@key_buffer_size; +select @@key_cache_block_size; +select @@key_cache_partitions; + +create table t1 ( + p int not null auto_increment primary key, + a char(10)); +create table t2 ( + p int not null auto_increment primary key, + i int, a char(10), key k1(i), key k2(a)); + +select @@key_cache_partitions; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), + (3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +select * from t2; +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; + +--replace_result 1808 KEY_BLOCKS_UNUSED 1670 KEY_BLOCKS_UNUSED 1789 KEY_BLOCKS_UNUSED +show status like 'key_%'; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +delete from t2 where a='zzzz'; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +delete from t1; +delete from t2; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +# For the key cache with 2 partitions execute the same sequence of +# statements as for the simple cache above. +# The statistical information on the number of i/o requests and +# the number of is expected to be the same. + +set global key_cache_partitions=2; +select @@key_cache_partitions; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), + (3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +select * from t2; +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; + +--replace_result 1808 KEY_BLOCKS_UNUSED 1670 KEY_BLOCKS_UNUSED 1788 KEY_BLOCKS_UNUSED +show status like 'key_%'; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +delete from t1; +delete from t2; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +# Check that we can work with one partition with the same results + +set global key_cache_partitions=1; +select @@key_cache_partitions; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), + (3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +select * from t2; +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; + +--replace_result 1808 KEY_BLOCKS_UNUSED 1670 KEY_BLOCKS_UNUSED 1789 KEY_BLOCKS_UNUSED +show status like 'key_%'; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +delete from t1; +delete from t2; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +flush tables; flush status; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +# Switch back to 2 partitions + +set global key_buffer_size=32*1024; +select @@key_buffer_size; +set global key_cache_partitions=2; +select @@key_cache_partitions; +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +insert into t1 values (1, 'qqqq'), (2, 'yyyy'); +insert into t2 values (1, 1, 'qqqq'), (2, 1, 'pppp'), + (3, 1, 'yyyy'), (4, 3, 'zzzz'); +select * from t1; +select * from t2; +update t1 set p=3 where p=1; +update t2 set i=2 where i=1; + +--replace_column 7 # +select * from information_schema.key_caches where partition_number is null; + +# Add more rows to tables t1 and t2 + +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; +insert into t1(a) select a from t1; + +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; +insert into t2(i,a) select i,a from t2; + +--replace_column 6 # 7 # 10 # +select * from information_schema.key_caches where partition_number is null; + +select * from t1 where p between 1010 and 1020 ; +select * from t2 where p between 1010 and 1020 ; +--replace_column 6 # 7 # 10 # +select * from information_schema.key_caches where partition_number is null; + +flush tables; flush status; +update t1 set a='zzzz' where a='qqqq'; +update t2 set i=1 where i=2; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +# Now test how we can work with 7 partitions + +set global keycache1.key_buffer_size=256*1024; +select @@keycache1.key_buffer_size; +set global keycache1.key_cache_partitions=7; +select @@keycache1.key_cache_partitions; + +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; +--replace_column 7 # +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; + +cache index t1 key (`primary`) in keycache1; + +explain select p from t1 where p between 1010 and 1020; +select p from t1 where p between 1010 and 1020; +explain select i from t2 where p between 1010 and 1020; +select i from t2 where p between 1010 and 1020; +explain select count(*) from t1, t2 where t1.p = t2.i; +select count(*) from t1, t2 where t1.p = t2.i; + +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; +--replace_column 7 # +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; + +cache index t2 in keycache1; +update t2 set p=p+3000, i=2 where a='qqqq'; +--replace_column 7 # +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; + +set global keycache2.key_buffer_size=1024*1024; +cache index t2 in keycache2; +insert into t2 values (2000, 3, 'yyyy'); +--replace_column 7 # +select * from information_schema.key_caches where key_cache_name like "keycache2" + and partition_number is null; +--replace_column 7 # +select * from information_schema.key_caches where key_cache_name like "key%" + and partition_number is null; + +cache index t2 in keycache1; +update t2 set p=p+5000 where a='zzzz'; +select * from t2 where p between 1010 and 1020; +explain select p from t2 where p between 1010 and 1020; +select p from t2 where p between 1010 and 1020; +explain select i from t2 where a='yyyy' and i=3; +select i from t2 where a='yyyy' and i=3; +explain select a from t2 where a='yyyy' and i=3; +select a from t2 where a='yyyy' and i=3 ; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=2*1024; +insert into t2 values (7000, 3, 'yyyy'); +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=8*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; +insert into t2 values (8000, 3, 'yyyy'); +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_buffer_size=64*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=2*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=8*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_buffer_size=0; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=8*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_buffer_size=0; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_buffer_size=128*1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +set global keycache1.key_cache_block_size=1024; +--replace_column 6 # 7 # +select * from information_schema.key_caches where partition_number is null; + +drop table t1,t2; + +set global keycache1.key_buffer_size=0; +set global keycache2.key_buffer_size=0; + +set global key_buffer_size=@save_key_buffer_size; +set global key_cache_partitions=@save_key_cache_partitions; + +#End of 5.1 tests === modified file 'mysql-test/t/plugin.test' --- a/mysql-test/t/plugin.test 2010-03-03 14:44:14 +0000 +++ b/mysql-test/t/plugin.test 2010-04-08 17:19:01 +0000 @@ -84,4 +84,40 @@ set session sql_mode=@old_sql_mode; --error ER_INCORRECT_GLOBAL_LOCAL_VAR set session old=bla; +############################################################### +# engine-specific clauses in the CREATE TABLE: + +--echo #legal values +CREATE TABLE t1 ( a int complex='c,f,f,f' ) ENGINE=example ULL=10000 STR='dskj' one_or_two='one' YESNO=0; +show create table t1; +drop table t1; + +SET @OLD_SQL_MODE=@@SQL_MODE; +SET SQL_MODE='IGNORE_BAD_TABLE_OPTIONS'; + +--echo #illegal value fixed +CREATE TABLE t1 (a int) ENGINE=example ULL=10000000000000000000 one_or_two='ttt' YESNO=SSS; +show create table t1; + +--echo #alter table +alter table t1 ULL=10000000; +show create table t1; +alter table t1 change a a int complex='c,c,c'; +show create table t1; +drop table t1; + +--echo #illegal value error +SET SQL_MODE=''; +--error ER_BAD_OPTION_VALUE +CREATE TABLE t1 (a int) ENGINE=example ULL=10000000000000000000 one_or_two='ttt' YESNO=SSS; +SET @@SQL_MODE=@OLD_SQL_MODE; + +# +# The only preparable statement above was CREATE TABLE. +# We need to prepare another statement here to force the +# previous one to be deallocated (mysqltest reuses the same handle) +# and to unlock all thread-local plugin locks. Otherwise it won't +# uninstall. +# +select 1; UNINSTALL PLUGIN example; === added file 'mysql-test/t/select_pkeycache-master.opt' --- a/mysql-test/t/select_pkeycache-master.opt 1970-01-01 00:00:00 +0000 +++ b/mysql-test/t/select_pkeycache-master.opt 2010-02-16 16:41:11 +0000 @@ -0,0 +1 @@ +--key_cache_partitions=7 === added file 'mysql-test/t/select_pkeycache.test' --- a/mysql-test/t/select_pkeycache.test 1970-01-01 00:00:00 +0000 +++ b/mysql-test/t/select_pkeycache.test 2010-02-16 16:41:11 +0000 @@ -0,0 +1,8 @@ + +# +# Run select.test partitioned default key cache (with 7 partitions) +# (see setting the number of partitions in select_pkecache-master.opt) +# The result is expected the same as for select.test +# + +--source t/select.test === modified file 'mysql-test/t/sp.test' --- a/mysql-test/t/sp.test 2009-12-23 13:44:03 +0000 +++ b/mysql-test/t/sp.test 2010-04-08 12:10:05 +0000 @@ -8210,7 +8210,6 @@ call p(); select @@sql_mode; set @@sql_mode= @old_mode; # Rename SQL modes that differ in name between the server and the table definition. -select replace(@full_mode, '?', 'NOT_USED') into @full_mode; select replace(@full_mode, 'ALLOW_INVALID_DATES', 'INVALID_DATES') into @full_mode; select name from mysql.proc where name = 'p' and sql_mode = @full_mode; drop procedure p; === modified file 'mysys/mf_keycache.c' --- a/mysys/mf_keycache.c 2010-03-30 12:36:49 +0000 +++ b/mysys/mf_keycache.c 2010-04-06 11:02:24 +0000 @@ -13,12 +13,43 @@ along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + + /** - @file + @file + The file contains the following modules: + + Simple Key Cache Module + + Partitioned Key Cache Module + + Key Cache Interface Module + +*/ + +#include "mysys_priv.h" +#include "mysys_err.h" +#include <keycache.h> +#include "my_static.h" +#include <m_string.h> +#include <my_bit.h> +#include <errno.h> +#include <stdarg.h> + +/****************************************************************************** + Simple Key Cache Module + + The module contains implementations of all key cache interface functions + employed by partitioned key caches. + +******************************************************************************/ + +/* These functions handle keyblock cacheing for ISAM and MyISAM tables. One cache can handle many files. It must contain buffers of the same blocksize. + init_key_cache() should be used to init cache handler. The free list (free_block_list) is a stack like structure. @@ -37,9 +68,7 @@ blocks_unused is the sum of never used blocks in the pool and of currently free blocks. blocks_used is the number of blocks fetched from the pool and as such gives the maximum number of in-use blocks at any time. -*/ -/* Key Cache Locking ================= @@ -104,14 +133,77 @@ I/O finished. */ -#include "mysys_priv.h" -#include "mysys_err.h" -#include <keycache.h> -#include "my_static.h" -#include <m_string.h> -#include <my_bit.h> -#include <errno.h> -#include <stdarg.h> +/* declare structures that is used by st_key_cache */ + +struct st_block_link; +typedef struct st_block_link BLOCK_LINK; +struct st_keycache_page; +typedef struct st_keycache_page KEYCACHE_PAGE; +struct st_hash_link; +typedef struct st_hash_link HASH_LINK; + +/* info about requests in a waiting queue */ +typedef struct st_keycache_wqueue +{ + struct st_my_thread_var *last_thread; /* circular list of waiting threads */ +} KEYCACHE_WQUEUE; + +#define CHANGED_BLOCKS_HASH 128 /* must be power of 2 */ + +/* Control block for a simple (non-partitioned) key cache */ + +typedef struct st_simple_key_cache_cb +{ + my_bool key_cache_inited; /* <=> control block is allocated */ + my_bool in_resize; /* true during resize operation */ + my_bool resize_in_flush; /* true during flush of resize operation */ + my_bool can_be_used; /* usage of cache for read/write is allowed */ + size_t key_cache_mem_size; /* specified size of the cache memory */ + uint key_cache_block_size; /* size of the page buffer of a cache block */ + ulong min_warm_blocks; /* min number of warm blocks; */ + ulong age_threshold; /* age threshold for hot blocks */ + ulonglong keycache_time; /* total number of block link operations */ + uint hash_entries; /* max number of entries in the hash table */ + int hash_links; /* max number of hash links */ + int hash_links_used; /* number of hash links currently used */ + int disk_blocks; /* max number of blocks in the cache */ + ulong blocks_used; /* maximum number of concurrently used blocks */ + ulong blocks_unused; /* number of currently unused blocks */ + ulong blocks_changed; /* number of currently dirty blocks */ + ulong warm_blocks; /* number of blocks in warm sub-chain */ + ulong cnt_for_resize_op; /* counter to block resize operation */ + long blocks_available; /* number of blocks available in the LRU chain */ + HASH_LINK **hash_root; /* arr. of entries into hash table buckets */ + HASH_LINK *hash_link_root; /* memory for hash table links */ + HASH_LINK *free_hash_list; /* list of free hash links */ + BLOCK_LINK *free_block_list; /* list of free blocks */ + BLOCK_LINK *block_root; /* memory for block links */ + uchar HUGE_PTR *block_mem; /* memory for block buffers */ + BLOCK_LINK *used_last; /* ptr to the last block of the LRU chain */ + BLOCK_LINK *used_ins; /* ptr to the insertion block in LRU chain */ + pthread_mutex_t cache_lock; /* to lock access to the cache structure */ + KEYCACHE_WQUEUE resize_queue; /* threads waiting during resize operation */ + /* + Waiting for a zero resize count. Using a queue for symmetry though + only one thread can wait here. + */ + KEYCACHE_WQUEUE waiting_for_resize_cnt; + KEYCACHE_WQUEUE waiting_for_hash_link; /* waiting for a free hash link */ + KEYCACHE_WQUEUE waiting_for_block; /* requests waiting for a free block */ + BLOCK_LINK *changed_blocks[CHANGED_BLOCKS_HASH]; /* hash for dirty file bl.*/ + BLOCK_LINK *file_blocks[CHANGED_BLOCKS_HASH]; /* hash for other file bl.*/ + + /* Statistics variables. These are reset in reset_key_cache_counters(). */ + ulong global_blocks_changed; /* number of currently dirty blocks */ + ulonglong global_cache_w_requests;/* number of write requests (write hits) */ + ulonglong global_cache_write; /* number of writes from cache to files */ + ulonglong global_cache_r_requests;/* number of read requests (read hits) */ + ulonglong global_cache_read; /* number of reads from files to cache */ + + int blocks; /* max number of blocks in the cache */ + uint hash_factor; /* factor used to calculate hash function */ + my_bool in_init; /* Set to 1 in MySQL during init/resize */ +} SIMPLE_KEY_CACHE_CB; /* Some compilation flags have been added specifically for this module @@ -223,7 +315,8 @@ KEY_CACHE *dflt_key_cache= &dflt_key_cac #define FLUSH_CACHE 2000 /* sort this many blocks at once */ -static int flush_all_key_blocks(KEY_CACHE *keycache); +static int flush_all_key_blocks(SIMPLE_KEY_CACHE_CB *keycache); +static void end_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, my_bool cleanup); #ifdef THREAD static void wait_on_queue(KEYCACHE_WQUEUE *wqueue, pthread_mutex_t *mutex); @@ -232,15 +325,16 @@ static void release_whole_queue(KEYCACHE #define wait_on_queue(wqueue, mutex) do {} while (0) #define release_whole_queue(wqueue) do {} while (0) #endif -static void free_block(KEY_CACHE *keycache, BLOCK_LINK *block); +static void free_block(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block); #if !defined(DBUG_OFF) -static void test_key_cache(KEY_CACHE *keycache, +static void test_key_cache(SIMPLE_KEY_CACHE_CB *keycache, const char *where, my_bool lock); #endif - +#define KEYCACHE_BASE_EXPR(f, pos) \ + ((ulong) ((pos) / keycache->key_cache_block_size) + (ulong) (f)) #define KEYCACHE_HASH(f, pos) \ -(((ulong) ((pos) / keycache->key_cache_block_size) + \ - (ulong) (f)) & (keycache->hash_entries-1)) + ((KEYCACHE_BASE_EXPR(f, pos) / keycache->hash_factor) & \ + (keycache->hash_entries-1)) #define FILE_HASH(f) ((uint) (f) & (CHANGED_BLOCKS_HASH-1)) #define DEFAULT_KEYCACHE_DEBUG_LOG "keycache_debug.log" @@ -336,9 +430,10 @@ static int keycache_pthread_cond_signal( #define inline /* disabled inline for easier debugging */ static int fail_block(BLOCK_LINK *block); static int fail_hlink(HASH_LINK *hlink); -static int cache_empty(KEY_CACHE *keycache); +static int cache_empty(SIMPLE_KEY_CACHE_CB *keycache); #endif + static inline uint next_power(uint value) { return (uint) my_round_up_to_next_power((uint32) value) << 1; @@ -346,19 +441,32 @@ static inline uint next_power(uint value /* - Initialize a key cache + Initialize a simple key cache SYNOPSIS - init_key_cache() - keycache pointer to a key cache data structure - key_cache_block_size size of blocks to keep cached data - use_mem total memory to use for the key cache - division_limit division limit (may be zero) - age_threshold age threshold (may be zero) + init_simple_key_cache() + keycache pointer to the control block of a simple key cache + key_cache_block_size size of blocks to keep cached data + use_mem memory to use for the key cache buferrs/structures + division_limit division limit (may be zero) + age_threshold age threshold (may be zero) + + DESCRIPTION + This function is the implementation of the init_key_cache interface + function that is employed by simple (non-partitioned) key caches. + The function builds a simple key cache and initializes the control block + structure of the type SIMPLE_KEY_CACHE_CB that is used for this key cache. + The parameter keycache is supposed to point to this structure. + The parameter key_cache_block_size specifies the size of the blocks in + the key cache to be built. The parameters division_limit and age_threshhold + determine the initial values of those characteristics of the key cache + that are used for midpoint insertion strategy. The parameter use_mem + specifies the total amount of memory to be allocated for key cache blocks + and auxiliary structures. RETURN VALUE number of blocks in the key cache, if successful, - 0 - otherwise. + <= 0 - otherwise. NOTES. if keycache->key_cache_inited != 0 we assume that the key cache @@ -367,17 +475,17 @@ static inline uint next_power(uint value It's assumed that no two threads call this function simultaneously referring to the same key cache handle. - */ -int init_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, - size_t use_mem, uint division_limit, - uint age_threshold) +static +int init_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold) { ulong blocks, hash_links; size_t length; int error; - DBUG_ENTER("init_key_cache"); + DBUG_ENTER("init_simple_key_cache"); DBUG_ASSERT(key_cache_block_size >= 512); KEYCACHE_DEBUG_OPEN; @@ -387,12 +495,15 @@ int init_key_cache(KEY_CACHE *keycache, DBUG_RETURN(0); } + keycache->blocks_used= keycache->blocks_unused= 0; + keycache->global_blocks_changed= 0; keycache->global_cache_w_requests= keycache->global_cache_r_requests= 0; keycache->global_cache_read= keycache->global_cache_write= 0; keycache->disk_blocks= -1; if (! keycache->key_cache_inited) { keycache->key_cache_inited= 1; + keycache->hash_factor= 1; /* Initialize these variables once only. Their value must survive re-initialization during resizing. @@ -534,51 +645,42 @@ err: /* - Resize a key cache + Prepare for resizing a simple key cache SYNOPSIS - resize_key_cache() - keycache pointer to a key cache data structure - key_cache_block_size size of blocks to keep cached data - use_mem total memory to use for the new key cache - division_limit new division limit (if not zero) - age_threshold new age threshold (if not zero) - - RETURN VALUE - number of blocks in the key cache, if successful, - 0 - otherwise. + prepare_resize_simple_key_cache() + keycache pointer to the control block of a simple key cache + with_resize_queue <=> resize queue is used + release_lock <=> release the key cache lock before return - NOTES. - The function first compares the memory size and the block size parameters - with the key cache values. + DESCRIPTION + This function flushes all dirty pages from a simple key cache and after + this it destroys the key cache calling end_simple_key_cache. The function + takes the parameter keycache as a pointer to the control block + structure of the type SIMPLE_KEY_CACHE_CB for this key cache. + The parameter with_resize_queue determines weather the resize queue is + involved (MySQL server never uses this queue). The parameter release_lock + says weather the key cache lock must be released before return from + the function. - If they differ the function free the the memory allocated for the - old key cache blocks by calling the end_key_cache function and - then rebuilds the key cache with new blocks by calling - init_key_cache. + RETURN VALUE + 0 - on success, + 1 - otherwise. - The function starts the operation only when all other threads - performing operations with the key cache let her to proceed - (when cnt_for_resize=0). + NOTES + This function is the called by resize_simple_key_cache and + resize_partitioned_key_cache that resize simple and partitioned key caches + respectively. */ -int resize_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, - size_t use_mem, uint division_limit, - uint age_threshold) +static +int prepare_resize_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, + my_bool with_resize_queue, + my_bool release_lock) { - int blocks; - DBUG_ENTER("resize_key_cache"); - - if (!keycache->key_cache_inited) - DBUG_RETURN(keycache->disk_blocks); - - if(key_cache_block_size == keycache->key_cache_block_size && - use_mem == keycache->key_cache_mem_size) - { - change_key_cache_param(keycache, division_limit, age_threshold); - DBUG_RETURN(keycache->disk_blocks); - } - + int res= 0; + DBUG_ENTER("prepare_resize_simple_key_cache"); + keycache_pthread_mutex_lock(&keycache->cache_lock); #ifdef THREAD @@ -588,7 +690,7 @@ int resize_key_cache(KEY_CACHE *keycache one resizer only. In set_var.cc keycache->in_init is used to block multiple attempts. */ - while (keycache->in_resize) + while (with_resize_queue && keycache->in_resize) { /* purecov: begin inspected */ wait_on_queue(&keycache->resize_queue, &keycache->cache_lock); @@ -613,8 +715,8 @@ int resize_key_cache(KEY_CACHE *keycache { /* TODO: if this happens, we should write a warning in the log file ! */ keycache->resize_in_flush= 0; - blocks= 0; keycache->can_be_used= 0; + res= 1; goto finish; } DBUG_ASSERT(cache_empty(keycache)); @@ -640,29 +742,144 @@ int resize_key_cache(KEY_CACHE *keycache #else KEYCACHE_DBUG_ASSERT(keycache->cnt_for_resize_op == 0); #endif - - /* - Free old cache structures, allocate new structures, and initialize - them. Note that the cache_lock mutex and the resize_queue are left - untouched. We do not lose the cache_lock and will release it only at - the end of this function. - */ - end_key_cache(keycache, 0); /* Don't free mutex */ - /* The following will work even if use_mem is 0 */ - blocks= init_key_cache(keycache, key_cache_block_size, use_mem, - division_limit, age_threshold); + + end_simple_key_cache(keycache, 0); finish: + if (release_lock) + keycache_pthread_mutex_unlock(&keycache->cache_lock); + DBUG_RETURN(res); +} + + +/* + Finalize resizing a simple key cache + + SYNOPSIS + finish_resize_simple_key_cache() + keycache pointer to the control block of a simple key cache + with_resize_queue <=> resize queue is used + acquire_lock <=> acquire the key cache lock at start + + DESCRIPTION + This function performs finalizing actions for the operation of + resizing a simple key cache. The function takes the parameter + keycache as a pointer to the control block structure of the type + SIMPLE_KEY_CACHE_CB for this key cache. The function sets the flag + in_resize in this structure to FALSE. + The parameter with_resize_queue determines weather the resize queue + is involved (MySQL server never uses this queue). + The parameter acquire_lock says weather the key cache lock must be + acquired at the start of the function. + + RETURN VALUE + none + + NOTES + This function is the called by resize_simple_key_cache and + resize_partitioned_key_cache that resize simple and partitioned key caches + respectively. +*/ + +static +void finish_resize_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, + my_bool with_resize_queue, + my_bool acquire_lock) +{ + DBUG_ENTER("finish_resize_simple_key_cache"); + + if (acquire_lock) + keycache_pthread_mutex_lock(&keycache->cache_lock); + + safe_mutex_assert_owner(&keycache->cache_lock); + /* Mark the resize finished. This allows other threads to start a resize or to request new cache blocks. */ keycache->in_resize= 0; - - /* Signal waiting threads. */ - release_whole_queue(&keycache->resize_queue); + + if (with_resize_queue) + { + /* Signal waiting threads. */ + release_whole_queue(&keycache->resize_queue); + } keycache_pthread_mutex_unlock(&keycache->cache_lock); + + DBUG_VOID_RETURN; +} + + +/* + Resize a simple key cache + + SYNOPSIS + resize_simple_key_cache() + keycache pointer to the control block of a simple key cache + key_cache_block_size size of blocks to keep cached data + use_mem memory to use for the key cache buffers/structures + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + This function is the implementation of the resize_key_cache interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for the simple key + cache to be resized. + The parameter key_cache_block_size specifies the new size of the blocks in + the key cache. The parameters division_limit and age_threshold + determine the new initial values of those characteristics of the key cache + that are used for midpoint insertion strategy. The parameter use_mem + specifies the total amount of memory to be allocated for key cache blocks + and auxiliary structures in the new key cache. + + RETURN VALUE + number of blocks in the key cache, if successful, + 0 - otherwise. + + NOTES. + The function first calls the function prepare_resize_simple_key_cache + to flush all dirty blocks from key cache, to free memory used + for key cache blocks and auxiliary structures. After this the + function builds a new key cache with new parameters. + + This implementation doesn't block the calls and executions of other + functions from the key cache interface. However it assumes that the + calls of resize_simple_key_cache itself are serialized. + + The function starts the operation only when all other threads + performing operations with the key cache let her to proceed + (when cnt_for_resize=0). +*/ + +static +int resize_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold) +{ + int blocks= 0; + DBUG_ENTER("resize_simple_key_cache"); + + if (!keycache->key_cache_inited) + DBUG_RETURN(blocks); + + /* + Note that the cache_lock mutex and the resize_queue are left untouched. + We do not lose the cache_lock and will release it only at the end of + this function. + */ + if (prepare_resize_simple_key_cache(keycache, 1, 0)) + goto finish; + + /* The following will work even if use_mem is 0 */ + blocks= init_simple_key_cache(keycache, key_cache_block_size, use_mem, + division_limit, age_threshold); + +finish: + finish_resize_simple_key_cache(keycache, 1, 0); + DBUG_RETURN(blocks); } @@ -670,7 +887,7 @@ finish: /* Increment counter blocking resize key cache operation */ -static inline void inc_counter_for_resize_op(KEY_CACHE *keycache) +static inline void inc_counter_for_resize_op(SIMPLE_KEY_CACHE_CB *keycache) { keycache->cnt_for_resize_op++; } @@ -680,35 +897,47 @@ static inline void inc_counter_for_resiz Decrement counter blocking resize key cache operation; Signal the operation to proceed when counter becomes equal zero */ -static inline void dec_counter_for_resize_op(KEY_CACHE *keycache) +static inline void dec_counter_for_resize_op(SIMPLE_KEY_CACHE_CB *keycache) { if (!--keycache->cnt_for_resize_op) release_whole_queue(&keycache->waiting_for_resize_cnt); } + /* - Change the key cache parameters + Change key cache parameters of a simple key cache SYNOPSIS - change_key_cache_param() - keycache pointer to a key cache data structure - division_limit new division limit (if not zero) - age_threshold new age threshold (if not zero) + change_simple_key_cache_param() + keycache pointer to the control block of a simple key cache + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + This function is the implementation of the change_key_cache_param interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for the simple key + cache where new values of the division limit and the age threshold used + for midpoint insertion strategy are to be set. The parameters + division_limit and age_threshold provide these new values. RETURN VALUE none NOTES. - Presently the function resets the key cache parameters - concerning midpoint insertion strategy - division_limit and - age_threshold. + Presently the function resets the key cache parameters concerning + midpoint insertion strategy - division_limit and age_threshold. + This function changes some parameters of a given key cache without + reformatting it. The function does not touch the contents the key + cache blocks. */ -void change_key_cache_param(KEY_CACHE *keycache, uint division_limit, - uint age_threshold) +static +void change_simple_key_cache_param(SIMPLE_KEY_CACHE_CB *keycache, uint division_limit, + uint age_threshold) { - DBUG_ENTER("change_key_cache_param"); - + DBUG_ENTER("change_simple_key_cache_param"); keycache_pthread_mutex_lock(&keycache->cache_lock); if (division_limit) keycache->min_warm_blocks= (keycache->disk_blocks * @@ -722,20 +951,31 @@ void change_key_cache_param(KEY_CACHE *k /* - Remove key_cache from memory + Destroy a simple key cache SYNOPSIS - end_key_cache() - keycache key cache handle - cleanup Complete free (Free also mutex for key cache) + end_simple_key_cache() + keycache pointer to the control block of a simple key cache + cleanup <=> complete free (free also mutex for key cache) + + DESCRIPTION + This function is the implementation of the end_key_cache interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for the simple key + cache to be destroyed. + The function frees the memory allocated for the key cache blocks and + auxiliary structures. If the value of the parameter cleanup is TRUE + then even the key cache mutex is freed. RETURN VALUE none */ -void end_key_cache(KEY_CACHE *keycache, my_bool cleanup) +static +void end_simple_key_cache(SIMPLE_KEY_CACHE_CB *keycache, my_bool cleanup) { - DBUG_ENTER("end_key_cache"); + DBUG_ENTER("end_simple_key_cache"); DBUG_PRINT("enter", ("key_cache: 0x%lx", (long) keycache)); if (!keycache->key_cache_inited) @@ -1026,7 +1266,7 @@ static inline void link_changed(BLOCK_LI void */ -static void link_to_file_list(KEY_CACHE *keycache, +static void link_to_file_list(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block, int file, my_bool unlink_block) { @@ -1067,7 +1307,7 @@ static void link_to_file_list(KEY_CACHE void */ -static void link_to_changed_list(KEY_CACHE *keycache, +static void link_to_changed_list(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block) { DBUG_ASSERT(block->status & BLOCK_IN_USE); @@ -1122,8 +1362,8 @@ static void link_to_changed_list(KEY_CAC not linked in the LRU ring. */ -static void link_block(KEY_CACHE *keycache, BLOCK_LINK *block, my_bool hot, - my_bool at_end) +static void link_block(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block, + my_bool hot, my_bool at_end) { BLOCK_LINK *ins; BLOCK_LINK **pins; @@ -1243,7 +1483,7 @@ static void link_block(KEY_CACHE *keycac See NOTES for link_block */ -static void unlink_block(KEY_CACHE *keycache, BLOCK_LINK *block) +static void unlink_block(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block) { DBUG_ASSERT((block->status & ~BLOCK_CHANGED) == (BLOCK_READ | BLOCK_IN_USE)); DBUG_ASSERT(block->hash_link); /*backptr to block NULL from free_block()*/ @@ -1301,7 +1541,8 @@ static void unlink_block(KEY_CACHE *keyc RETURN void */ -static void reg_requests(KEY_CACHE *keycache, BLOCK_LINK *block, int count) +static void reg_requests(SIMPLE_KEY_CACHE_CB *keycache, + BLOCK_LINK *block, int count) { DBUG_ASSERT(block->status & BLOCK_IN_USE); DBUG_ASSERT(block->hash_link); @@ -1344,7 +1585,7 @@ static void reg_requests(KEY_CACHE *keyc not linked in the LRU ring. */ -static void unreg_request(KEY_CACHE *keycache, +static void unreg_request(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block, int at_end) { DBUG_ASSERT(block->status & (BLOCK_READ | BLOCK_IN_USE)); @@ -1433,7 +1674,7 @@ static void remove_reader(BLOCK_LINK *bl signals on its termination */ -static void wait_for_readers(KEY_CACHE *keycache, +static void wait_for_readers(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block) { #ifdef THREAD @@ -1482,7 +1723,7 @@ static inline void link_hash(HASH_LINK * Remove a hash link from the hash table */ -static void unlink_hash(KEY_CACHE *keycache, HASH_LINK *hash_link) +static void unlink_hash(SIMPLE_KEY_CACHE_CB *keycache, HASH_LINK *hash_link) { KEYCACHE_DBUG_PRINT("unlink_hash", ("fd: %u pos_ %lu #requests=%u", (uint) hash_link->file,(ulong) hash_link->diskpos, hash_link->requests)); @@ -1538,7 +1779,7 @@ static void unlink_hash(KEY_CACHE *keyca Get the hash link for a page */ -static HASH_LINK *get_hash_link(KEY_CACHE *keycache, +static HASH_LINK *get_hash_link(SIMPLE_KEY_CACHE_CB *keycache, int file, my_off_t filepos) { reg1 HASH_LINK *hash_link, **start; @@ -1659,7 +1900,7 @@ restart: waits until first of this operations links any block back. */ -static BLOCK_LINK *find_key_block(KEY_CACHE *keycache, +static BLOCK_LINK *find_key_block(SIMPLE_KEY_CACHE_CB *keycache, File file, my_off_t filepos, int init_hits_left, int wrmode, int *page_st) @@ -2419,7 +2660,7 @@ restart: portion is less than read_length, but not less than min_length. */ -static void read_block(KEY_CACHE *keycache, +static void read_block(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block, uint read_length, uint min_length, my_bool primary) { @@ -2507,43 +2748,60 @@ static void read_block(KEY_CACHE *keycac /* - Read a block of data from a cached file into a buffer; + Read a block of data from a simple key cache into a buffer SYNOPSIS - key_cache_read() - keycache pointer to a key cache data structure - file handler for the file for the block of data to be read - filepos position of the block of data in the file - level determines the weight of the data - buff buffer to where the data must be placed - length length of the buffer - block_length length of the block in the key cache buffer - return_buffer return pointer to the key cache buffer with the data + simple_key_cache_read() + keycache pointer to the control block of a simple key cache + file handler for the file for the block of data to be read + filepos position of the block of data in the file + level determines the weight of the data + buff buffer to where the data must be placed + length length of the buffer + block_length length of the read data from a key cache block + return_buffer return pointer to the key cache buffer with the data + DESCRIPTION + This function is the implementation of the key_cache_read interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for a simple key + cache. + In a general case the function reads a block of data from the key cache + into the buffer buff of the size specified by the parameter length. The + beginning of the block of data to be read is specified by the parameters + file and filepos. The length of the read data is the same as the length + of the buffer. The data is read into the buffer in key_cache_block_size + increments. If the next portion of the data is not found in any key cache + block, first it is read from file into the key cache. + If the parameter return_buffer is not ignored and its value is TRUE, and + the data to be read of the specified size block_length can be read from one + key cache buffer, then the function returns a pointer to the data in the + key cache buffer. + The function takse into account parameters block_length and return buffer + only in a single-threaded environment. + The parameter 'level' is used only by the midpoint insertion strategy + when the data or its portion cannot be found in the key cache. + RETURN VALUE - Returns address from where the data is placed if sucessful, 0 - otherwise. + Returns address from where the data is placed if successful, 0 - otherwise. - NOTES. - The function ensures that a block of data of size length from file - positioned at filepos is in the buffers for some key cache blocks. - Then the function either copies the data into the buffer buff, or, - if return_buffer is TRUE, it just returns the pointer to the key cache - buffer with the data. + NOTES Filepos must be a multiple of 'block_length', but it doesn't have to be a multiple of key_cache_block_size; */ -uchar *key_cache_read(KEY_CACHE *keycache, - File file, my_off_t filepos, int level, - uchar *buff, uint length, - uint block_length __attribute__((unused)), - int return_buffer __attribute__((unused))) +uchar *simple_key_cache_read(SIMPLE_KEY_CACHE_CB *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length __attribute__((unused)), + int return_buffer __attribute__((unused))) { my_bool locked_and_incremented= FALSE; int error=0; uchar *start= buff; - DBUG_ENTER("key_cache_read"); + DBUG_ENTER("simple_key_cache_read"); DBUG_PRINT("enter", ("fd: %u pos: %lu length: %u", (uint) file, (ulong) filepos, length)); @@ -2738,28 +2996,47 @@ end: /* - Insert a block of file data from a buffer into key cache + Insert a block of file data from a buffer into a simple key cache SYNOPSIS - key_cache_insert() - keycache pointer to a key cache data structure + simple_key_cache_insert() + keycache pointer to the control block of a simple key cache file handler for the file to insert data from filepos position of the block of data in the file to insert level determines the weight of the data buff buffer to read data from length length of the data in the buffer - NOTES - This is used by MyISAM to move all blocks from a index file to the key - cache - + DESCRIPTION + This function is the implementation of the key_cache_insert interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for a simple key + cache. + The function writes a block of file data from a buffer into the key cache. + The buffer is specified with the parameters buff and length - the pointer + to the beginning of the buffer and its size respectively. It's assumed + the buffer contains the data from 'file' allocated from the position + filepos. The data is copied from the buffer in key_cache_block_size + increments. + The parameter level is used to set one characteristic for the key buffers + loaded with the data from buff. The characteristic is used only by the + midpoint insertion strategy. + RETURN VALUE 0 if a success, 1 - otherwise. + + NOTES + The function is used by MyISAM to move all blocks from a index file to + the key cache. It can be performed in parallel with reading the file data + from the key buffers by other threads. + */ -int key_cache_insert(KEY_CACHE *keycache, - File file, my_off_t filepos, int level, - uchar *buff, uint length) +static +int simple_key_cache_insert(SIMPLE_KEY_CACHE_CB *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length) { int error= 0; DBUG_ENTER("key_cache_insert"); @@ -2979,43 +3256,64 @@ int key_cache_insert(KEY_CACHE *keycache /* - Write a buffer into a cached file. + Write a buffer into a simple key cache SYNOPSIS - key_cache_write() - keycache pointer to a key cache data structure - file handler for the file to write data to - filepos position in the file to write data to - level determines the weight of the data - buff buffer with the data - length length of the buffer - dont_write if is 0 then all dirty pages involved in writing - should have been flushed from key cache + simple_key_cache_write() + keycache pointer to the control block of a simple key cache + file handler for the file to write data to + file_extra maps of key cache partitions containing + dirty pages from file + filepos position in the file to write data to + level determines the weight of the data + buff buffer with the data + length length of the buffer + dont_write if is 0 then all dirty pages involved in writing + should have been flushed from key cache + DESCRIPTION + This function is the implementation of the key_cache_write interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for a simple key + cache. + In a general case the function copies data from a buffer into the key + cache. The buffer is specified with the parameters buff and length - + the pointer to the beginning of the buffer and its size respectively. + It's assumed the buffer contains the data to be written into 'file' + starting from the position filepos. The data is copied from the buffer + in key_cache_block_size increments. + If the value of the parameter dont_write is FALSE then the function + also writes the data into file. + The parameter level is used to set one characteristic for the key buffers + filled with the data from buff. The characteristic is employed only by + the midpoint insertion strategy. + The parameter file_extra currently makes sense only for simple key caches + that are elements of a partitioned key cache. It provides a pointer to the + shared bitmap of the partitions that may contains dirty pages for the file. + This bitmap is used to optimize the function + flush_partitioned_key_cache_blocks. + RETURN VALUE 0 if a success, 1 - otherwise. - NOTES. - The function copies the data of size length from buff into buffers - for key cache blocks that are assigned to contain the portion of - the file starting with position filepos. - It ensures that this data is flushed to the file if dont_write is FALSE. - Filepos must be a multiple of 'block_length', but it doesn't - have to be a multiple of key_cache_block_size; - - dont_write is always TRUE in the server (info->lock_type is never F_UNLCK). + NOTES + This implementation exploits the fact that the function is called only + when a thread has got an exclusive lock for the key file. */ -int key_cache_write(KEY_CACHE *keycache, - File file, my_off_t filepos, int level, - uchar *buff, uint length, - uint block_length __attribute__((unused)), - int dont_write) +static +int simple_key_cache_write(SIMPLE_KEY_CACHE_CB *keycache, + File file, void *file_extra __attribute__((unused)), + my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length __attribute__((unused)), + int dont_write) { my_bool locked_and_incremented= FALSE; int error=0; - DBUG_ENTER("key_cache_write"); + DBUG_ENTER("simple_key_cache_write"); DBUG_PRINT("enter", ("fd: %u pos: %lu length: %u block_length: %u" " key_block_length: %u", @@ -3330,7 +3628,7 @@ end: Block must have a request registered on it. */ -static void free_block(KEY_CACHE *keycache, BLOCK_LINK *block) +static void free_block(SIMPLE_KEY_CACHE_CB *keycache, BLOCK_LINK *block) { KEYCACHE_THREAD_TRACE("free block"); KEYCACHE_DBUG_PRINT("free_block", @@ -3470,7 +3768,7 @@ static int cmp_sec_link(BLOCK_LINK **a, free used blocks if requested */ -static int flush_cached_blocks(KEY_CACHE *keycache, +static int flush_cached_blocks(SIMPLE_KEY_CACHE_CB *keycache, File file, BLOCK_LINK **cache, BLOCK_LINK **end, enum flush_type type) @@ -3514,9 +3812,9 @@ static int flush_cached_blocks(KEY_CACHE (BLOCK_READ | BLOCK_IN_FLUSH | BLOCK_CHANGED | BLOCK_IN_USE)); block->status|= BLOCK_IN_FLUSHWRITE; keycache_pthread_mutex_unlock(&keycache->cache_lock); - error= my_pwrite(file, block->buffer+block->offset, + error= my_pwrite(file, block->buffer + block->offset, block->length - block->offset, - block->hash_link->diskpos+ block->offset, + block->hash_link->diskpos + block->offset, MYF(MY_NABP | MY_WAIT_IF_FULL)); keycache_pthread_mutex_lock(&keycache->cache_lock); keycache->global_cache_write++; @@ -3576,7 +3874,7 @@ static int flush_cached_blocks(KEY_CACHE /* - Flush all key blocks for a file to disk, but don't do any mutex locks. + Flush all key blocks for a file to disk, but don't do any mutex locks SYNOPSIS flush_key_blocks_int() @@ -3598,7 +3896,7 @@ static int flush_cached_blocks(KEY_CACHE 1 error */ -static int flush_key_blocks_int(KEY_CACHE *keycache, +static int flush_key_blocks_int(SIMPLE_KEY_CACHE_CB *keycache, File file, enum flush_type type) { BLOCK_LINK *cache_buff[FLUSH_CACHE],**cache; @@ -4040,22 +4338,46 @@ err: /* - Flush all blocks for a file to disk + Flush all blocks for a file from key buffers of a simple key cache SYNOPSIS - flush_key_blocks() - keycache pointer to a key cache data structure - file handler for the file to flush to - flush_type type of the flush + flush_simple_key_blocks() + keycache pointer to the control block of a simple key cache + file handler for the file to flush to + file_extra maps of key cache partitions containing + dirty pages from file (not used) + flush_type type of the flush operation + DESCRIPTION + This function is the implementation of the flush_key_blocks interface + function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type S_KEY_CACHE_CB for a simple key + cache. + In a general case the function flushes the data from all dirty key + buffers related to the file 'file' into this file. The function does + exactly this if the value of the parameter type is FLUSH_KEEP. If the + value of this parameter is FLUSH_RELEASE, the function additionally + releases the key buffers containing data from 'file' for new usage. + If the value of the parameter type is FLUSH_IGNORE_CHANGED the function + just releases the key buffers containing data from 'file'. + The parameter file_extra currently is not used by this function. + RETURN 0 ok 1 error + + NOTES + This implementation exploits the fact that the function is called only + when a thread has got an exclusive lock for the key file. */ -int flush_key_blocks(KEY_CACHE *keycache, - File file, enum flush_type type) +static +int flush_simple_key_cache_blocks(SIMPLE_KEY_CACHE_CB *keycache, + File file, + void *file_extra __attribute__((unused)), + enum flush_type type) { int res= 0; DBUG_ENTER("flush_key_blocks"); @@ -4109,7 +4431,7 @@ int flush_key_blocks(KEY_CACHE *keycache != 0 Error */ -static int flush_all_key_blocks(KEY_CACHE *keycache) +static int flush_all_key_blocks(SIMPLE_KEY_CACHE_CB *keycache) { BLOCK_LINK *block; uint total_found; @@ -4212,37 +4534,43 @@ static int flush_all_key_blocks(KEY_CACH /* - Reset the counters of a key cache. + Reset the counters of a simple key cache SYNOPSIS - reset_key_cache_counters() - name the name of a key cache - key_cache pointer to the key kache to be reset + reset_simple_key_cache_counters() + name the name of a key cache + keycache pointer to the control block of a simple key cache DESCRIPTION - This procedure is used by process_key_caches() to reset the counters of all - currently used key caches, both the default one and the named ones. + This function is the implementation of the reset_key_cache_counters + interface function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type S_KEY_CACHE_CB for a simple key cache. + This function resets the values of all statistical counters for the key + cache to 0. + The parameter name is currently not used. RETURN 0 on success (always because it can't fail) */ -int reset_key_cache_counters(const char *name __attribute__((unused)), - KEY_CACHE *key_cache) +static +int reset_simple_key_cache_counters(const char *name __attribute__((unused)), + SIMPLE_KEY_CACHE_CB *keycache) { - DBUG_ENTER("reset_key_cache_counters"); - if (!key_cache->key_cache_inited) + DBUG_ENTER("reset_simple_key_cache_counters"); + if (!keycache->key_cache_inited) { DBUG_PRINT("info", ("Key cache %s not initialized.", name)); DBUG_RETURN(0); } DBUG_PRINT("info", ("Resetting counters for key cache %s.", name)); - key_cache->global_blocks_changed= 0; /* Key_blocks_not_flushed */ - key_cache->global_cache_r_requests= 0; /* Key_read_requests */ - key_cache->global_cache_read= 0; /* Key_reads */ - key_cache->global_cache_w_requests= 0; /* Key_write_requests */ - key_cache->global_cache_write= 0; /* Key_writes */ + keycache->global_blocks_changed= 0; /* Key_blocks_not_flushed */ + keycache->global_cache_r_requests= 0; /* Key_read_requests */ + keycache->global_cache_read= 0; /* Key_reads */ + keycache->global_cache_w_requests= 0; /* Key_write_requests */ + keycache->global_cache_write= 0; /* Key_writes */ DBUG_RETURN(0); } @@ -4251,9 +4579,10 @@ int reset_key_cache_counters(const char /* Test if disk-cache is ok */ -static void test_key_cache(KEY_CACHE *keycache __attribute__((unused)), - const char *where __attribute__((unused)), - my_bool lock __attribute__((unused))) +static +void test_key_cache(SIMPLE_KEY_CACHE_CB *keycache __attribute__((unused)), + const char *where __attribute__((unused)), + my_bool lock __attribute__((unused))) { /* TODO */ } @@ -4265,7 +4594,7 @@ static void test_key_cache(KEY_CACHE *ke #define MAX_QUEUE_LEN 100 -static void keycache_dump(KEY_CACHE *keycache) +static void keycache_dump(SIMPLE_KEY_CACHE_CB *keycache) { FILE *keycache_dump_file=fopen(KEYCACHE_DUMP_FILE, "w"); struct st_my_thread_var *last; @@ -4505,7 +4834,7 @@ static int fail_hlink(HASH_LINK *hlink) return 0; /* Let the assert fail. */ } -static int cache_empty(KEY_CACHE *keycache) +static int cache_empty(SIMPLE_KEY_CACHE_CB *keycache) { int errcnt= 0; int idx; @@ -4543,3 +4872,1702 @@ static int cache_empty(KEY_CACHE *keycac } #endif + +/* + Get statistics for a simple key cache + + SYNOPSIS + get_simple_key_cache_statistics() + keycache pointer to the control block of a simple key cache + partition_no partition number (not used) + key_cache_stats OUT pointer to the structure for the returned statistics + + DESCRIPTION + This function is the implementation of the get_key_cache_statistics + interface function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for a simple key + cache. This function returns the statistical data for the key cache. + The parameter partition_no is not used by this function. + + RETURN + none +*/ + +static +void get_simple_key_cache_statistics(SIMPLE_KEY_CACHE_CB *keycache, + uint partition_no __attribute__((unused)), + KEY_CACHE_STATISTICS *keycache_stats) +{ + DBUG_ENTER("simple_get_key_cache_statistics"); + + keycache_stats->mem_size= (longlong) keycache->key_cache_mem_size; + keycache_stats->block_size= (longlong) keycache->key_cache_block_size; + keycache_stats->blocks_used= keycache->blocks_used; + keycache_stats->blocks_unused= keycache->blocks_unused; + keycache_stats->blocks_changed= keycache->global_blocks_changed; + keycache_stats->read_requests= keycache->global_cache_r_requests; + keycache_stats->reads= keycache->global_cache_read; + keycache_stats->write_requests= keycache->global_cache_w_requests; + keycache_stats->writes= keycache->global_cache_write; + DBUG_VOID_RETURN; +} + + +/* + Offsets of the statistical values in the control block for a simple key cache + The first NO_LONG_KEY_CACHE_STAT_VARIABLES=3 are of the ulong type while the + remaining are of the ulonglong type. + */ +static size_t simple_key_cache_stat_var_offsets[]= +{ + offsetof(SIMPLE_KEY_CACHE_CB, blocks_used), + offsetof(SIMPLE_KEY_CACHE_CB, blocks_unused), + offsetof(SIMPLE_KEY_CACHE_CB, global_blocks_changed), + offsetof(SIMPLE_KEY_CACHE_CB, global_cache_w_requests), + offsetof(SIMPLE_KEY_CACHE_CB, global_cache_write), + offsetof(SIMPLE_KEY_CACHE_CB, global_cache_r_requests), + offsetof(SIMPLE_KEY_CACHE_CB, global_cache_read) +}; + + +/* + Get the value of a statistical variable for a simple key cache + + SYNOPSIS + get_simple_key_cache_stat_value() + keycache pointer to the control block of a simple key cache + var_no the ordered number of a statistical variable + + DESCRIPTION + This function is the implementation of the get_simple_key_cache_stat_value + interface function that is employed by simple (non-partitioned) key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type SIMPLE_KEY_CACHE_CB for a simple key + cache. This function returns the value of the statistical variable var_no + for this key cache. The variables are numbered starting from 0 to 6. + + RETURN + The value of the specified statistical variable + +*/ + +static +ulonglong get_simple_key_cache_stat_value(SIMPLE_KEY_CACHE_CB *keycache, + uint var_no) +{ + size_t var_ofs= simple_key_cache_stat_var_offsets[var_no]; + ulonglong res= 0; + DBUG_ENTER("get_simple_key_cache_stat_value"); + + if (var_no < 3) + res= (ulonglong) (*(long *) ((char *) keycache + var_ofs)); + else + res= *(ulonglong *) ((char *) keycache + var_ofs); + + DBUG_RETURN(res); +} + + +/* + The array of pointer to the key cache interface functions used for simple + key caches. Any simple key cache objects including those incorporated into + partitioned keys caches exploit this array. + + The current implementation of these functions allows to call them from + the MySQL server code directly. We don't do it though. +*/ + +static KEY_CACHE_FUNCS simple_key_cache_funcs = +{ + (INIT_KEY_CACHE) init_simple_key_cache, + (RESIZE_KEY_CACHE) resize_simple_key_cache, + (CHANGE_KEY_CACHE_PARAM) change_simple_key_cache_param, + (KEY_CACHE_READ) simple_key_cache_read, + (KEY_CACHE_INSERT) simple_key_cache_insert, + (KEY_CACHE_WRITE) simple_key_cache_write, + (FLUSH_KEY_BLOCKS) flush_simple_key_cache_blocks, + (RESET_KEY_CACHE_COUNTERS) reset_simple_key_cache_counters, + (END_KEY_CACHE) end_simple_key_cache, + (GET_KEY_CACHE_STATISTICS) get_simple_key_cache_statistics, + (GET_KEY_CACHE_STAT_VALUE) get_simple_key_cache_stat_value +}; + + +/****************************************************************************** + Partitioned Key Cache Module + + The module contains implementations of all key cache interface functions + employed by partitioned key caches. + + A partitioned key cache is a collection of structures for simple key caches + called key cache partitions. Any page from a file can be placed into a buffer + of only one partition. The number of the partition is calculated from + the file number and the position of the page in the file, and it's always the + same for the page. The function that maps pages into partitions takes care + of even distribution of pages among partitions. + + Partition key cache mitigate one of the major problem of simple key cache: + thread contention for key cache lock (mutex). Every call of a key cache + interface function must acquire this lock. So threads compete for this lock + even in the case when they have acquired shared locks for the file and + pages they want read from are in the key cache buffers. + When working with a partitioned key cache any key cache interface function + that needs only one page has to acquire the key cache lock only for the + partition the page is ascribed to. This makes the chances for threads not + compete for the same key cache lock better. Unfortunately if we use a + partitioned key cache with N partitions for B-tree indexes we can't say + that the chances becomes N times less. The fact is that any index lookup + operation requires reading from the root page that, for any index, is always + ascribed to the same partition. To resolve this problem we should have + employed more sophisticated mechanisms of working with root pages. + + Currently the number of partitions in a partitioned key cache is limited + by 64. We could increase this limit. Simultaneously we would have to increase + accordingly the size of the bitmap dirty_part_map from the MYISAM_SHARE + structure. + +******************************************************************************/ + +/* Control block for a partitioned key cache */ + +typedef struct st_partitioned_key_cache_cb +{ + my_bool key_cache_inited; /*<=> control block is allocated */ + SIMPLE_KEY_CACHE_CB **partition_array; /* the key cache partitions */ + size_t key_cache_mem_size; /* specified size of the cache memory */ + uint key_cache_block_size; /* size of the page buffer of a cache block */ + uint partitions; /* number of partitions in the key cache */ +} PARTITIONED_KEY_CACHE_CB; + +static +void end_partitioned_key_cache(PARTITIONED_KEY_CACHE_CB *keycache, + my_bool cleanup); + +static int +reset_partitioned_key_cache_counters(const char *name, + PARTITIONED_KEY_CACHE_CB *keycache); + +/* + Determine the partition to which the index block to read is ascribed + + SYNOPSIS + get_key_cache_partition() + keycache pointer to the control block of a partitioned key cache + file handler for the file for the block of data to be read + filepos position of the block of data in the file + + DESCRIPTION + The function determines the number of the partition in whose buffer the + block from 'file' at the position filepos has to be placed for reading. + The function returns the control block of the simple key cache for this + partition to the caller. + + RETURN VALUE + The pointer to the control block of the partition to which the specified + file block is ascribed. +*/ + +static +SIMPLE_KEY_CACHE_CB * +get_key_cache_partition(PARTITIONED_KEY_CACHE_CB *keycache, + File file, my_off_t filepos) +{ + uint i= KEYCACHE_BASE_EXPR(file, filepos) % keycache->partitions; + return keycache->partition_array[i]; +} + + +/* + Determine the partition to which the index block to write is ascribed + + SYNOPSIS + get_key_cache_partition() + keycache pointer to the control block of a partitioned key cache + file handler for the file for the block of data to be read + filepos position of the block of data in the file + dirty_part_map pointer to the bitmap of dirty partitions for the file + + DESCRIPTION + The function determines the number of the partition in whose buffer the + block from 'file' at the position filepos has to be placed for writing and + marks the partition as dirty in the dirty_part_map bitmap. + The function returns the control block of the simple key cache for this + partition to the caller. + + RETURN VALUE + The pointer to the control block of the partition to which the specified + file block is ascribed. +*/ + +static SIMPLE_KEY_CACHE_CB +*get_key_cache_partition_for_write(PARTITIONED_KEY_CACHE_CB *keycache, + File file, my_off_t filepos, + ulonglong* dirty_part_map) +{ + uint i= KEYCACHE_BASE_EXPR( file, filepos) % keycache->partitions; + *dirty_part_map|= 1<<i; + return keycache->partition_array[i]; +} + + +/* + Initialize a partitioned key cache + + SYNOPSIS + init_partitioned_key_cache() + keycache pointer to the control block of a partitioned key cache + key_cache_block_size size of blocks to keep cached data + use_mem total memory to use for all key cache partitions + division_limit division limit (may be zero) + age_threshold age threshold (may be zero) + + DESCRIPTION + This function is the implementation of the init_key_cache interface function + that is employed by partitioned key caches. + The function builds and initializes an array of simple key caches, and then + initializes the control block structure of the type PARTITIONED_KEY_CACHE_CB + that is used for a partitioned key cache. The parameter keycache is + supposed to point to this structure. The number of partitions in the + partitioned key cache to be built must be passed through the field + 'partitions' of this structure. The parameter key_cache_block_size specifies + the size of the blocks in the the simple key caches to be built. + The parameters division_limit and age_threshold determine the initial + values of those characteristics of the simple key caches that are used for + midpoint insertion strategy. The parameter use_mem specifies the total + amount of memory to be allocated for the key cache blocks in all simple key + caches and for all auxiliary structures. + + RETURN VALUE + total number of blocks in key cache partitions, if successful, + <= 0 - otherwise. + + NOTES + If keycache->key_cache_inited != 0 then we assume that the memory for + the array of partitions has been already allocated. + + It's assumed that no two threads call this function simultaneously + referring to the same key cache handle. +*/ + +static +int init_partitioned_key_cache(PARTITIONED_KEY_CACHE_CB *keycache, + uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold) +{ + int i; + size_t mem_per_cache; + size_t mem_decr; + int cnt; + SIMPLE_KEY_CACHE_CB *partition; + SIMPLE_KEY_CACHE_CB **partition_ptr; + uint partitions= keycache->partitions; + int blocks= 0; + DBUG_ENTER("partitioned_init_key_cache"); + + keycache->key_cache_block_size = key_cache_block_size; + + if (keycache->key_cache_inited) + partition_ptr= keycache->partition_array; + else + { + if(!(partition_ptr= + (SIMPLE_KEY_CACHE_CB **) my_malloc(sizeof(SIMPLE_KEY_CACHE_CB *) * + partitions, MYF(MY_WME)))) + DBUG_RETURN(-1); + bzero(partition_ptr, sizeof(SIMPLE_KEY_CACHE_CB *) * partitions); + keycache->partition_array= partition_ptr; + } + + mem_decr= mem_per_cache / 5; + mem_per_cache = use_mem / partitions; + + for (i= 0; i < (int) partitions; i++) + { + my_bool key_cache_inited= keycache->key_cache_inited; + if (key_cache_inited) + partition= *partition_ptr; + else + { + if (!(partition= + (SIMPLE_KEY_CACHE_CB *) my_malloc(sizeof(SIMPLE_KEY_CACHE_CB), + MYF(MY_WME)))) + continue; + partition->key_cache_inited= 0; + } + + cnt= init_simple_key_cache(partition, key_cache_block_size, mem_per_cache, + division_limit, age_threshold); + if (cnt <= 0) + { + end_simple_key_cache(partition, 1); + if (!key_cache_inited) + { + my_free(partition, MYF(0)); + partition= 0; + } + if ((i == 0 && cnt < 0) || i > 0) + { + /* + Here we have two cases: + 1. i == 0 and cnt < 0 + cnt < 0 => mem_per_cache is not big enough to allocate minimal + number of key blocks in the key cache of the partition. + Decrease the the number of the partitions by 1 and start again. + 2. i > 0 + There is not enough memory for one of the succeeding partitions. + Just skip this partition decreasing the number of partitions in + the key cache by one. + Do not change the value of mem_per_cache in both cases. + */ + if (key_cache_inited) + { + my_free(partition, MYF(0)); + partition= 0; + if(key_cache_inited) + memmove(partition_ptr, partition_ptr+1, + sizeof(partition_ptr)*(partitions-i-1)); + } + if (!--partitions) + break; + } + else + { + /* + We come here when i == 0 && cnt == 0. + cnt == 0 => the memory allocator fails to allocate a block of + memory of the size mem_per_cache. Decrease the value of + mem_per_cache without changing the current number of partitions + and start again. Make sure that such a decrease may happen not + more than 5 times in total. + */ + if (use_mem <= mem_decr) + break; + use_mem-= mem_decr; + } + i--; + mem_per_cache= use_mem/partitions; + continue; + } + else + { + blocks+= cnt; + *partition_ptr++= partition; + } + } + + keycache->partitions= partitions= partition_ptr-keycache->partition_array; + keycache->key_cache_mem_size= mem_per_cache * partitions; + for (i= 0; i < (int) partitions; i++) + keycache->partition_array[i]->hash_factor= partitions; + + keycache->key_cache_inited= 1; + + if (!partitions) + blocks= -1; + + DBUG_RETURN(blocks); +} + + +/* + Resize a partitioned key cache + + SYNOPSIS + resize_partitioned_key_cache() + keycache pointer to the control block of a partitioned key cache + key_cache_block_size size of blocks to keep cached data + use_mem total memory to use for the new key cache + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + This function is the implementation of the resize_key_cache interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for the + partitioned key cache to be resized. + The parameter key_cache_block_size specifies the new size of the blocks in + the simple key caches that comprise the partitioned key cache. + The parameters division_limit and age_threshold determine the new initial + values of those characteristics of the simple key cache that are used for + midpoint insertion strategy. The parameter use-mem specifies the total + amount of memory to be allocated for the key cache blocks in all new + simple key caches and for all auxiliary structures. + + RETURN VALUE + number of blocks in the key cache, if successful, + 0 - otherwise. + + NOTES. + The function first calls prepare_resize_simple_key_cache for each simple + key cache effectively flushing all dirty pages from it and destroying + the key cache. Then init_partitioned_key_cache is called. This call builds + a new array of simple key caches containing the same number of elements + as the old one. After this the function calls the function + finish_resize_simple_key_cache for each simple key cache from this array. + + This implementation doesn't block the calls and executions of other + functions from the key cache interface. However it assumes that the + calls of resize_partitioned_key_cache itself are serialized. +*/ + +static +int resize_partitioned_key_cache(PARTITIONED_KEY_CACHE_CB *keycache, + uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold) +{ + uint i; + uint partitions= keycache->partitions; + my_bool cleanup= use_mem == 0; + int blocks= -1; + int err= 0; + DBUG_ENTER("partitioned_resize_key_cache"); + if (cleanup) + { + end_partitioned_key_cache(keycache, 0); + DBUG_RETURN(-1); + } + for (i= 0; i < partitions; i++) + { + err|= prepare_resize_simple_key_cache(keycache->partition_array[i], 0, 1); + } + if (!err) + blocks= init_partitioned_key_cache(keycache, key_cache_block_size, + use_mem, division_limit, age_threshold); + if (blocks > 0) + { + for (i= 0; i < partitions; i++) + { + finish_resize_simple_key_cache(keycache->partition_array[i], 0, 1); + } + } + DBUG_RETURN(blocks); +} + + +/* + Change key cache parameters of a partitioned key cache + + SYNOPSIS + partitioned_change_key_cache_param() + keycache pointer to the control block of a partitioned key cache + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + This function is the implementation of the change_key_cache_param interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for the simple + key cache where new values of the division limit and the age threshold used + for midpoint insertion strategy are to be set. The parameters + division_limit and age_threshold provide these new values. + + RETURN VALUE + none + + NOTES + The function just calls change_simple_key_cache_param for each element from + the array of simple caches that comprise the partitioned key cache. +*/ + +static +void change_partitioned_key_cache_param(PARTITIONED_KEY_CACHE_CB *keycache, + uint division_limit, + uint age_threshold) +{ + uint i; + uint partitions= keycache->partitions; + DBUG_ENTER("partitioned_change_key_cache_param"); + for (i= 0; i < partitions; i++) + { + change_simple_key_cache_param(keycache->partition_array[i], division_limit, + age_threshold); + } + DBUG_VOID_RETURN; +} + + +/* + Destroy a partitioned key cache + + SYNOPSIS + end_partitioned_key_cache() + keycache pointer to the control block of a partitioned key cache + cleanup <=> complete free (free also control block structures + for all simple key caches) + + DESCRIPTION + This function is the implementation of the end_key_cache interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for the + partitioned key cache to be destroyed. + The function frees the memory allocated for the cache blocks and + auxiliary structures used by simple key caches that comprise the + partitioned key cache. If the value of the parameter cleanup is TRUE + then even the memory used for control blocks of the simple key caches + and the array of pointers to them are freed. + + RETURN VALUE + none +*/ + +static +void end_partitioned_key_cache(PARTITIONED_KEY_CACHE_CB *keycache, + my_bool cleanup) +{ + uint i; + uint partitions= keycache->partitions; + DBUG_ENTER("partitioned_end_key_cache"); + DBUG_PRINT("enter", ("key_cache: 0x%lx", (long) keycache)); + + for (i= 0; i < partitions; i++) + { + end_simple_key_cache(keycache->partition_array[i], cleanup); + } + if (cleanup) + { + for (i= 0; i < partitions; i++) + my_free((uchar*) keycache->partition_array[i], MYF(0)); + my_free((uchar*) keycache->partition_array, MYF(0)); + keycache->key_cache_inited= 0; + } + DBUG_VOID_RETURN; +} + + +/* + Read a block of data from a partitioned key cache into a buffer + + SYNOPSIS + + partitioned_key_cache_read() + keycache pointer to the control block of a partitioned key cache + file handler for the file for the block of data to be read + filepos position of the block of data in the file + level determines the weight of the data + buff buffer to where the data must be placed + length length of the buffer + block_length length of the read data from a key cache block + return_buffer return pointer to the key cache buffer with the data + + DESCRIPTION + This function is the implementation of the key_cache_read interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a + partitioned key cache. + In a general case the function reads a block of data from the key cache + into the buffer buff of the size specified by the parameter length. The + beginning of the block of data to be read is specified by the parameters + file and filepos. The length of the read data is the same as the length + of the buffer. The data is read into the buffer in key_cache_block_size + increments. To read each portion the function first finds out in what + partition of the key cache this portion(page) is to be saved, and calls + simple_key_cache_read with the pointer to the corresponding simple key as + its first parameter. + If the parameter return_buffer is not ignored and its value is TRUE, and + the data to be read of the specified size block_length can be read from one + key cache buffer, then the function returns a pointer to the data in the + key cache buffer. + The function takes into account parameters block_length and return buffer + only in a single-threaded environment. + The parameter 'level' is used only by the midpoint insertion strategy + when the data or its portion cannot be found in the key cache. + + RETURN VALUE + Returns address from where the data is placed if successful, 0 - otherwise. +*/ + +static +uchar *partitioned_key_cache_read(PARTITIONED_KEY_CACHE_CB *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length __attribute__((unused)), + int return_buffer __attribute__((unused))) +{ + uint r_length; + uint offset= (uint) (filepos % keycache->key_cache_block_size); + uchar *start= buff; + DBUG_ENTER("partitioned_key_cache_read"); + DBUG_PRINT("enter", ("fd: %u pos: %lu length: %u", + (uint) file, (ulong) filepos, length)); + +#ifndef THREAD + if (block_length > keycache->key_cache_block_size || offset) + return_buffer=0; +#endif + + /* Read data in key_cache_block_size increments */ + do + { + SIMPLE_KEY_CACHE_CB *partition= get_key_cache_partition(keycache, + file, filepos); + uchar *ret_buff= 0; + r_length= length; + set_if_smaller(r_length, keycache->key_cache_block_size - offset); + ret_buff= simple_key_cache_read((void *) partition, + file, filepos, level, + buff, r_length, + block_length, return_buffer); + if (ret_buff == 0) + DBUG_RETURN(0); +#ifndef THREAD + /* This is only true if we were able to read everything in one block */ + if (return_buffer) + DBUG_RETURN(ret_buff); +#endif + filepos+= r_length; + buff+= r_length; + offset= 0; + } while ((length-= r_length)); + + DBUG_RETURN(start); +} + + +/* + Insert a block of file data from a buffer into a partitioned key cache + + SYNOPSIS + partitioned_key_cache_insert() + keycache pointer to the control block of a partitioned key cache + file handler for the file to insert data from + filepos position of the block of data in the file to insert + level determines the weight of the data + buff buffer to read data from + length length of the data in the buffer + + DESCRIPTION + This function is the implementation of the key_cache_insert interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a + partitioned key cache. + The function writes a block of file data from a buffer into the key cache. + The buffer is specified with the parameters buff and length - the pointer + to the beginning of the buffer and its size respectively. It's assumed + that the buffer contains the data from 'file' allocated from the position + filepos. The data is copied from the buffer in key_cache_block_size + increments. For every portion of data the function finds out in what simple + key cache from the array of partitions the data must be stored, and after + this calls simple_key_cache_insert to copy the data into a key buffer of + this simple key cache. + The parameter level is used to set one characteristic for the key buffers + loaded with the data from buff. The characteristic is used only by the + midpoint insertion strategy. + + RETURN VALUE + 0 if a success, 1 - otherwise. + + NOTES + The function is used by MyISAM to move all blocks from a index file to + the key cache. It can be performed in parallel with reading the file data + from the key buffers by other threads. +*/ + +static +int partitioned_key_cache_insert(PARTITIONED_KEY_CACHE_CB *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length) +{ + uint w_length; + uint offset= (uint) (filepos % keycache->key_cache_block_size); + DBUG_ENTER("partitioned_key_cache_insert"); + DBUG_PRINT("enter", ("fd: %u pos: %lu length: %u", + (uint) file,(ulong) filepos, length)); + + + /* Write data in key_cache_block_size increments */ + do + { + SIMPLE_KEY_CACHE_CB *partition= get_key_cache_partition(keycache, + file, filepos); + w_length= length; + set_if_smaller(w_length, keycache->key_cache_block_size - offset); + if (simple_key_cache_insert((void *) partition, + file, filepos, level, + buff, w_length)) + DBUG_RETURN(1); + + filepos+= w_length; + buff+= w_length; + offset = 0; + } while ((length-= w_length)); + + DBUG_RETURN(0); +} + + +/* + Write data from a buffer into a partitioned key cache + + SYNOPSIS + + partitioned_key_cache_write() + keycache pointer to the control block of a partitioned key cache + file handler for the file to write data to + filepos position in the file to write data to + level determines the weight of the data + buff buffer with the data + length length of the buffer + dont_write if is 0 then all dirty pages involved in writing + should have been flushed from key cache + file_extra maps of key cache partitions containing + dirty pages from file + + DESCRIPTION + This function is the implementation of the key_cache_write interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a + partitioned key cache. + In a general case the function copies data from a buffer into the key + cache. The buffer is specified with the parameters buff and length - + the pointer to the beginning of the buffer and its size respectively. + It's assumed the buffer contains the data to be written into 'file' + starting from the position filepos. The data is copied from the buffer + in key_cache_block_size increments. For every portion of data the + function finds out in what simple key cache from the array of partitions + the data must be stored, and after this calls simple_key_cache_write to + copy the data into a key buffer of this simple key cache. + If the value of the parameter dont_write is FALSE then the function + also writes the data into file. + The parameter level is used to set one characteristic for the key buffers + filled with the data from buff. The characteristic is employed only by + the midpoint insertion strategy. + The parameter file_expra provides a pointer to the shared bitmap of + the partitions that may contains dirty pages for the file. This bitmap + is used to optimize the function flush_partitioned_key_cache_blocks. + + RETURN VALUE + 0 if a success, 1 - otherwise. + + NOTES + This implementation exploits the fact that the function is called only + when a thread has got an exclusive lock for the key file. +*/ + +static +int partitioned_key_cache_write(PARTITIONED_KEY_CACHE_CB *keycache, + File file, void *file_extra, + my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length __attribute__((unused)), + int dont_write) +{ + uint w_length; + ulonglong *part_map= (ulonglong *) file_extra; + uint offset= (uint) (filepos % keycache->key_cache_block_size); + DBUG_ENTER("partitioned_key_cache_write"); + DBUG_PRINT("enter", + ("fd: %u pos: %lu length: %u block_length: %u" + " key_block_length: %u", + (uint) file, (ulong) filepos, length, block_length, + keycache ? keycache->key_cache_block_size : 0)); + + + /* Write data in key_cache_block_size increments */ + do + { + SIMPLE_KEY_CACHE_CB *partition= get_key_cache_partition_for_write(keycache, + file, + filepos, + part_map); + w_length = length; + set_if_smaller(w_length, keycache->key_cache_block_size - offset ); + if (simple_key_cache_write(partition, + file, 0, filepos, level, + buff, w_length, block_length, + dont_write)) + DBUG_RETURN(1); + + filepos+= w_length; + buff+= w_length; + offset= 0; + } while ((length-= w_length)); + + DBUG_RETURN(0); +} + + +/* + Flush all blocks for a file from key buffers of a partitioned key cache + + SYNOPSIS + + flush_partitioned_key_cache_blocks() + keycache pointer to the control block of a partitioned key cache + file handler for the file to flush to + file_extra maps of key cache partitions containing + dirty pages from file (not used) + flush_type type of the flush operation + + DESCRIPTION + This function is the implementation of the flush_key_blocks interface + function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a + partitioned key cache. + In a general case the function flushes the data from all dirty key + buffers related to the file 'file' into this file. The function does + exactly this if the value of the parameter type is FLUSH_KEEP. If the + value of this parameter is FLUSH_RELEASE, the function additionally + releases the key buffers containing data from 'file' for new usage. + If the value of the parameter type is FLUSH_IGNORE_CHANGED the function + just releases the key buffers containing data from 'file'. + The function performs the operation by calling the function + flush_simple_key_cache_blocks for the elements of the array of the + simple key caches that comprise the partitioned key_cache. If the value + of the parameter type is FLUSH_KEEP s_flush_key_blocks is called only + for the partitions with possibly dirty pages marked in the bitmap + pointed to by the parameter file_extra. + + RETURN + 0 ok + 1 error + + NOTES + This implementation exploits the fact that the function is called only + when a thread has got an exclusive lock for the key file. +*/ + +static +int flush_partitioned_key_cache_blocks(PARTITIONED_KEY_CACHE_CB *keycache, + File file, void *file_extra, + enum flush_type type) +{ + uint i; + uint partitions= keycache->partitions; + int err= 0; + ulonglong *dirty_part_map= (ulonglong *) file_extra; + DBUG_ENTER("partitioned_flush_key_blocks"); + DBUG_PRINT("enter", ("keycache: 0x%lx", (long) keycache)); + + for (i= 0; i < partitions; i++) + { + SIMPLE_KEY_CACHE_CB *partition= keycache->partition_array[i]; + if ((type == FLUSH_KEEP || type == FLUSH_FORCE_WRITE) && + !((*dirty_part_map) & ((ulonglong) 1 << i))) + continue; + err|= test(flush_simple_key_cache_blocks(partition, file, 0, type)); + } + *dirty_part_map= 0; + + DBUG_RETURN(err); +} + + +/* + Reset the counters of a partitioned key cache + + SYNOPSIS + reset_partitioned_key_cache_counters() + name the name of a key cache + keycache pointer to the control block of a partitioned key cache + + DESCRIPTION + This function is the implementation of the reset_key_cache_counters + interface function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a partitioned + key cache. + This function resets the values of the statistical counters of the simple + key caches comprising partitioned key cache to 0. It does it by calling + reset_simple_key_cache_counters for each key cache partition. + The parameter name is currently not used. + + RETURN + 0 on success (always because it can't fail) +*/ + +static int +reset_partitioned_key_cache_counters(const char *name __attribute__((unused)), + PARTITIONED_KEY_CACHE_CB *keycache) +{ + uint i; + uint partitions= keycache->partitions; + DBUG_ENTER("partitioned_reset_key_cache_counters"); + + for (i = 0; i < partitions; i++) + { + reset_simple_key_cache_counters(name, keycache->partition_array[i]); + } + DBUG_RETURN(0); +} + + +/* + Get statistics for a partition key cache + + SYNOPSIS + get_partitioned_key_cache_statistics() + keycache pointer to the control block of a partitioned key cache + partition_no partition number to get statistics for + key_cache_stats OUT pointer to the structure for the returned statistics + + DESCRIPTION + This function is the implementation of the get_key_cache_statistics + interface function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for + a partitioned key cache. + If the value of the parameter partition_no is equal to 0 then aggregated + statistics for all partitions is returned in the fields of the + structure key_cache_stat of the type KEY_CACHE_STATISTICS . Otherwise + the function returns data for the partition number partition_no of the + key cache in the structure key_cache_stat. (Here partitions are numbered + starting from 1.) + + RETURN + none +*/ + +static +void +get_partitioned_key_cache_statistics(PARTITIONED_KEY_CACHE_CB *keycache, + uint partition_no, + KEY_CACHE_STATISTICS *keycache_stats) +{ + uint i; + SIMPLE_KEY_CACHE_CB *partition; + uint partitions= keycache->partitions; + DBUG_ENTER("get_partitioned_key_cache_statistics"); + + if (partition_no != 0) + { + partition= keycache->partition_array[partition_no-1]; + get_simple_key_cache_statistics((void *) partition, 0, keycache_stats); + DBUG_VOID_RETURN; + } + bzero(keycache_stats, sizeof(KEY_CACHE_STATISTICS)); + keycache_stats->mem_size= (longlong) keycache->key_cache_mem_size; + keycache_stats->block_size= (longlong) keycache->key_cache_block_size; + for (i = 0; i < partitions; i++) + { + partition= keycache->partition_array[i]; + keycache_stats->blocks_used+= partition->blocks_used; + keycache_stats->blocks_unused+= partition->blocks_unused; + keycache_stats->blocks_changed+= partition->global_blocks_changed; + keycache_stats->read_requests+= partition->global_cache_r_requests; + keycache_stats->reads+= partition->global_cache_read; + keycache_stats->write_requests+= partition->global_cache_w_requests; + keycache_stats->writes+= partition->global_cache_write; + } + DBUG_VOID_RETURN; +} + +/* + Get the value of a statistical variable for a partitioned key cache + + SYNOPSIS + get_partitioned_key_cache_stat_value() + keycache pointer to the control block of a partitioned key cache + var_no the ordered number of a statistical variable + + DESCRIPTION + This function is the implementation of the get_key_cache_stat_value + interface function that is employed by partitioned key caches. + The function takes the parameter keycache as a pointer to the + control block structure of the type PARTITIONED_KEY_CACHE_CB for a + partitioned key cache. + This function returns the value of the statistical variable var_no + for this key cache. The variables are numbered starting from 0 to 6. + The returned value is calculated as the sum of the values of the + statistical variable with number var_no for all simple key caches that + comprise the partitioned key cache. + + RETURN + The value of the specified statistical variable +*/ + +static +ulonglong +get_partitioned_key_cache_stat_value(PARTITIONED_KEY_CACHE_CB *keycache, + uint var_no) +{ + uint i; + uint partitions= keycache->partitions; + size_t var_ofs= simple_key_cache_stat_var_offsets[var_no]; + ulonglong res= 0; + DBUG_ENTER("get_partitioned_key_cache_stat_value"); + + if (var_no < NUM_LONG_KEY_CACHE_STAT_VARIABLES) + { + for (i = 0; i < partitions; i++) + { + SIMPLE_KEY_CACHE_CB *partition= keycache->partition_array[i]; + res+= (ulonglong) (*(long *) ((char *) partition + var_ofs)); + } + } + else + { + for (i = 0; i < partitions; i++) + { + SIMPLE_KEY_CACHE_CB *partition= keycache->partition_array[i]; + res+= *(ulonglong *) ((char *) partition + var_ofs); + } + } + DBUG_RETURN(res); +} + + +/* + The array of pointers to the key cache interface functions used by + partitioned key caches. Any partitioned key cache object caches exploits + this array. + + The current implementation of these functions does not allow to call + them from the MySQL server code directly. The key cache interface + wrappers must be used for this purpose. +*/ + +static KEY_CACHE_FUNCS partitioned_key_cache_funcs = +{ + (INIT_KEY_CACHE) init_partitioned_key_cache, + (RESIZE_KEY_CACHE) resize_partitioned_key_cache, + (CHANGE_KEY_CACHE_PARAM) change_partitioned_key_cache_param, + (KEY_CACHE_READ) partitioned_key_cache_read, + (KEY_CACHE_INSERT) partitioned_key_cache_insert, + (KEY_CACHE_WRITE) partitioned_key_cache_write, + (FLUSH_KEY_BLOCKS) flush_partitioned_key_cache_blocks, + (RESET_KEY_CACHE_COUNTERS) reset_partitioned_key_cache_counters, + (END_KEY_CACHE) end_partitioned_key_cache, + (GET_KEY_CACHE_STATISTICS) get_partitioned_key_cache_statistics, + (GET_KEY_CACHE_STAT_VALUE) get_partitioned_key_cache_stat_value +}; + + +/****************************************************************************** + Key Cache Interface Module + + The module contains wrappers for all key cache interface functions. + + Currently there are key caches of two types: simple key caches and + partitioned key caches. Each type (class) has its own implementation of the + basic key cache operations used the MyISAM storage engine. The pointers + to the implementation functions are stored in two static structures of the + type KEY_CACHE_FUNC: simple_key_cache_funcs - for simple key caches, and + partitioned_key_cache_funcs - for partitioned key caches. When a key cache + object is created the constructor procedure init_key_cache places a pointer + to the corresponding table into one of its fields. The procedure also + initializes a control block for the key cache oject and saves the pointer + to this block in another field of the key cache object. + When a key cache wrapper function is invoked for a key cache object to + perform a basic key cache operation it looks into the interface table + associated with the key cache oject and calls the corresponding + implementation of the operation. It passes the saved key cache control + block to this implementation. If, for some reasons, the control block + has not been fully initialized yet, the wrapper function either does not + do anything or, in the case when it perform a read/write operation, the + function do it directly through the system i/o functions. + + As we can see the model with which the key cache interface is supported + as quite conventional for interfaces in general. + +******************************************************************************/ + + +/* + Initialize a key cache + + SYNOPSIS + init_key_cache() + keycache pointer to the key cache to be initialized + key_cache_block_size size of blocks to keep cached data + use_mem total memory to use for cache buffers/structures + division_limit division limit (may be zero) + age_threshold age threshold (may be zero) + partitions number of partitions in the key cache + + DESCRIPTION + The function creates a control block structure for a key cache and + places the pointer to this block in the structure keycache. + If the value of the parameter 'partitions' is 0 then a simple key cache + is created. Otherwise a partitioned key cache with the specified number + of partitions is created. + The parameter key_cache_block_size specifies the size of the blocks in + the key cache to be created. The parameters division_limit and + age_threshold determine the initial values of those characteristics of + the key cache that are used for midpoint insertion strategy. The parameter + use_mem specifies the total amount of memory to be allocated for the + key cache buffers and for all auxiliary structures. + + RETURN VALUE + total number of blocks in key cache partitions, if successful, + <= 0 - otherwise. + + NOTES + if keycache->key_cache_inited != 0 we assume that the memory + for the control block of the key cache has been already allocated. + + It's assumed that no two threads call this function simultaneously + referring to the same key cache handle. +*/ + +int init_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold, uint partitions) +{ + void *keycache_cb; + int blocks; + if (keycache->key_cache_inited) + keycache_cb= keycache->keycache_cb; + else + { + if (partitions == 0) + { + if (!(keycache_cb= (void *) my_malloc(sizeof(SIMPLE_KEY_CACHE_CB), + MYF(0)))) + return 0; + ((SIMPLE_KEY_CACHE_CB *) keycache_cb)->key_cache_inited= 0; + keycache->key_cache_type= SIMPLE_KEY_CACHE; + keycache->interface_funcs= &simple_key_cache_funcs; + } + else + { + if (!(keycache_cb= (void *) my_malloc(sizeof(PARTITIONED_KEY_CACHE_CB), + MYF(0)))) + return 0; + ((PARTITIONED_KEY_CACHE_CB *) keycache_cb)->key_cache_inited= 0; + keycache->key_cache_type= PARTITIONED_KEY_CACHE; + keycache->interface_funcs= &partitioned_key_cache_funcs; + } + keycache->keycache_cb= keycache_cb; + keycache->key_cache_inited= 1; + } + + if (partitions != 0) + { + ((PARTITIONED_KEY_CACHE_CB *) keycache_cb)->partitions= partitions; + } + keycache->can_be_used= 0; + blocks= keycache->interface_funcs->init(keycache_cb, key_cache_block_size, + use_mem, division_limit, + age_threshold); + keycache->partitions= partitions ? + ((PARTITIONED_KEY_CACHE_CB *) keycache_cb)->partitions : + 0; + DBUG_ASSERT(partitions <= MAX_KEY_CACHE_PARTITIONS); + if (blocks > 0) + keycache->can_be_used= 1; + return blocks; +} + + +/* + Resize a key cache + + SYNOPSIS + resize_key_cache() + keycache pointer to the key cache to be resized + key_cache_block_size size of blocks to keep cached data + use_mem total memory to use for the new key cache + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + The function operates over the key cache key cache. + The parameter key_cache_block_size specifies the new size of the block + buffers in the key cache. The parameters division_limit and age_threshold + determine the new initial values of those characteristics of the key cache + that are used for midpoint insertion strategy. The parameter use_mem + specifies the total amount of memory to be allocated for the key cache + buffers and for all auxiliary structures. + + RETURN VALUE + number of blocks in the key cache, if successful, + 0 - otherwise. + + NOTES + The function does not block the calls and executions of other functions + from the key cache interface. However it assumes that the calls of + resize_key_cache itself are serialized. + + Currently the function is called when the values of the variables + key_buffer_size and/or key_cache_block_size are being reset for + the key cache keycache. +*/ + +int resize_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, + size_t use_mem, uint division_limit, uint age_threshold) +{ + int blocks= -1; + if (keycache->key_cache_inited) + { + if ((uint) keycache->param_partitions != keycache->partitions && use_mem) + blocks= repartition_key_cache(keycache, + key_cache_block_size, use_mem, + division_limit, age_threshold, + (uint) keycache->param_partitions); + else + { + blocks= keycache->interface_funcs->resize(keycache->keycache_cb, + key_cache_block_size, + use_mem, division_limit, + age_threshold); + + if (keycache->partitions) + keycache->partitions= + ((PARTITIONED_KEY_CACHE_CB *)(keycache->keycache_cb))->partitions; + } + + keycache->can_be_used= (blocks >= 0); + } + return blocks; +} + + +/* + Change key cache parameters of a key cache + + SYNOPSIS + change_key_cache_param() + keycache pointer to the key cache to change parameters for + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + + DESCRIPTION + The function sets new values of the division limit and the age threshold + used when the key cache keycach employs midpoint insertion strategy. + The parameters division_limit and age_threshold provide these new values. + + RETURN VALUE + none + + NOTES + Currently the function is called when the values of the variables + key_cache_division_limit and/or key_cache_age_threshold are being reset + for the key cache keycache. +*/ + +void change_key_cache_param(KEY_CACHE *keycache, uint division_limit, + uint age_threshold) +{ + if (keycache->key_cache_inited) + { + + keycache->interface_funcs->change_param(keycache->keycache_cb, + division_limit, + age_threshold); + } +} + + +/* + Destroy a key cache + + SYNOPSIS + end_key_cache() + keycache pointer to the key cache to be destroyed + cleanup <=> complete free + + DESCRIPTION + The function frees the memory allocated for the cache blocks and + auxiliary structures used by the key cache keycache. If the value + of the parameter cleanup is TRUE then all resources used by the key + cache are to be freed. + + RETURN VALUE + none +*/ + +void end_key_cache(KEY_CACHE *keycache, my_bool cleanup) +{ + if (keycache->key_cache_inited) + { + keycache->interface_funcs->end(keycache->keycache_cb, cleanup); + if (cleanup) + { + if (keycache->keycache_cb) + { + my_free((uchar *) keycache->keycache_cb, MYF(0)); + keycache->keycache_cb= 0; + } + keycache->key_cache_inited= 0; + } + keycache->can_be_used= 0; + } +} + + +/* + Read a block of data from a key cache into a buffer + + SYNOPSIS + + key_cache_read() + keycache pointer to the key cache to read data from + file handler for the file for the block of data to be read + filepos position of the block of data in the file + level determines the weight of the data + buff buffer to where the data must be placed + length length of the buffer + block_length length of the data read from a key cache block + return_buffer return pointer to the key cache buffer with the data + + DESCRIPTION + The function operates over buffers of the key cache keycache. + In a general case the function reads a block of data from the key cache + into the buffer buff of the size specified by the parameter length. The + beginning of the block of data to be read is specified by the parameters + file and filepos. The length of the read data is the same as the length + of the buffer. + If the parameter return_buffer is not ignored and its value is TRUE, and + the data to be read of the specified size block_length can be read from one + key cache buffer, then the function returns a pointer to the data in the + key cache buffer. + The parameter 'level' is used only by the midpoint insertion strategy + when the data or its portion cannot be found in the key cache. + The function reads data into the buffer directly from file if the control + block of the key cache has not been initialized yet. + + RETURN VALUE + Returns address from where the data is placed if successful, 0 - otherwise. + + NOTES. + Filepos must be a multiple of 'block_length', but it doesn't + have to be a multiple of key_cache_block_size; +*/ + +uchar *key_cache_read(KEY_CACHE *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length, int return_buffer) +{ + if (keycache->key_cache_inited && keycache->can_be_used) + return keycache->interface_funcs->read(keycache->keycache_cb, + file, filepos, level, + buff, length, + block_length, return_buffer); + + /* We can't use mutex here as the key cache may not be initialized */ + keycache->global_cache_r_requests++; + keycache->global_cache_read++; + + if (my_pread(file, (uchar*) buff, length, filepos, MYF(MY_NABP))) + return (uchar *) 0; + + return buff; +} + + +/* + Insert a block of file data from a buffer into a key cache + + SYNOPSIS + key_cache_insert() + keycache pointer to the key cache to insert data into + file handler for the file to insert data from + filepos position of the block of data in the file to insert + level determines the weight of the data + buff buffer to read data from + length length of the data in the buffer + + DESCRIPTION + The function operates over buffers of the key cache keycache. + The function writes a block of file data from a buffer into the key cache. + The buffer is specified with the parameters buff and length - the pointer + to the beginning of the buffer and its size respectively. It's assumed + that the buffer contains the data from 'file' allocated from the position + filepos. + The parameter level is used to set one characteristic for the key buffers + loaded with the data from buff. The characteristic is used only by the + midpoint insertion strategy. + + RETURN VALUE + 0 if a success, 1 - otherwise. + + NOTES + The function is used by MyISAM to move all blocks from a index file to + the key cache. + It is assumed that it may be performed in parallel with reading the file + data from the key buffers by other threads. +*/ + +int key_cache_insert(KEY_CACHE *keycache, + File file, my_off_t filepos, int level, + uchar *buff, uint length) +{ + if (keycache->key_cache_inited && keycache->can_be_used) + return keycache->interface_funcs->insert(keycache->keycache_cb, + file, filepos, level, + buff, length); + return 0; +} + + +/* + Write data from a buffer into a key cache + + SYNOPSIS + + key_cache_write() + keycache pointer to the key cache to write data to + file handler for the file to write data to + filepos position in the file to write data to + level determines the weight of the data + buff buffer with the data + length length of the buffer + dont_write if is 0 then all dirty pages involved in writing + should have been flushed from key cache + file_extra pointer to optional file attributes + + DESCRIPTION + The function operates over buffers of the key cache keycache. + In a general case the function writes data from a buffer into the key + cache. The buffer is specified with the parameters buff and length - + the pointer to the beginning of the buffer and its size respectively. + It's assumed the buffer contains the data to be written into 'file' + starting from the position filepos. + If the value of the parameter dont_write is FALSE then the function + also writes the data into file. + The parameter level is used to set one characteristic for the key buffers + filled with the data from buff. The characteristic is employed only by + the midpoint insertion strategy. + The parameter file_expra may point to additional file attributes used + for optimization or other purposes. + The function writes data from the buffer directly into file if the control + block of the key cache has not been initialized yet. + + RETURN VALUE + 0 if a success, 1 - otherwise. + + NOTES + This implementation may exploit the fact that the function is called only + when a thread has got an exclusive lock for the key file. +*/ + +int key_cache_write(KEY_CACHE *keycache, + File file, void *file_extra, + my_off_t filepos, int level, + uchar *buff, uint length, + uint block_length, int force_write) +{ + if (keycache->key_cache_inited && keycache->can_be_used) + return keycache->interface_funcs->write(keycache->keycache_cb, + file, file_extra, + filepos, level, + buff, length, + block_length, force_write); + + /* We can't use mutex here as the key cache may not be initialized */ + keycache->global_cache_w_requests++; + keycache->global_cache_write++; + if (my_pwrite(file, buff, length, filepos, MYF(MY_NABP | MY_WAIT_IF_FULL))) + return 1; + + return 0; +} + + +/* + Flush all blocks for a file from key buffers of a key cache + + SYNOPSIS + + flush_key_blocks() + keycache pointer to the key cache whose blocks are to be flushed + file handler for the file to flush to + file_extra maps of key cache (used for partitioned key caches) + flush_type type of the flush operation + + DESCRIPTION + The function operates over buffers of the key cache keycache. + In a general case the function flushes the data from all dirty key + buffers related to the file 'file' into this file. The function does + exactly this if the value of the parameter type is FLUSH_KEEP. If the + value of this parameter is FLUSH_RELEASE, the function additionally + releases the key buffers containing data from 'file' for new usage. + If the value of the parameter type is FLUSH_IGNORE_CHANGED the function + just releases the key buffers containing data from 'file'. + If the value of the parameter type is FLUSH_KEEP the function may use + the value of the parameter file_extra pointing to possibly dirty + partitions to optimize the operation for partitioned key caches. + + RETURN + 0 ok + 1 error + + NOTES + Any implementation of the function may exploit the fact that the function + is called only when a thread has got an exclusive lock for the key file. +*/ + +int flush_key_blocks(KEY_CACHE *keycache, + int file, void *file_extra, + enum flush_type type) +{ + if (keycache->key_cache_inited) + return keycache->interface_funcs->flush(keycache->keycache_cb, + file, file_extra, type); + return 0; +} + + +/* + Reset the counters of a key cache + + SYNOPSIS + reset_key_cache_counters() + name the name of a key cache (unused) + keycache pointer to the key cache for which to reset counters + + DESCRIPTION + This function resets the values of the statistical counters for the key + cache keycache. + The parameter name is currently not used. + + RETURN + 0 on success (always because it can't fail) + + NOTES + This procedure is used by process_key_caches() to reset the counters of all + currently used key caches, both the default one and the named ones. +*/ + +int reset_key_cache_counters(const char *name __attribute__((unused)), + KEY_CACHE *keycache) +{ + if (keycache->key_cache_inited) + { + + return keycache->interface_funcs->reset_counters(name, + keycache->keycache_cb); + } + return 0; +} + + +/* + Get statistics for a key cache + + SYNOPSIS + get_key_cache_statistics() + keycache pointer to the key cache to get statistics for + partition_no partition number to get statistics for + key_cache_stats OUT pointer to the structure for the returned statistics + + DESCRIPTION + If the value of the parameter partition_no is equal to 0 then statistics + for the whole key cache keycache (aggregated statistics) is returned in the + fields of the structure key_cache_stat of the type KEY_CACHE_STATISTICS. + Otherwise the value of the parameter partition_no makes sense only for + a partitioned key cache. In this case the function returns statistics + for the partition with the specified number partition_no. + + RETURN + none +*/ + +void get_key_cache_statistics(KEY_CACHE *keycache, uint partition_no, + KEY_CACHE_STATISTICS *key_cache_stats) +{ + if (keycache->key_cache_inited) + { + keycache->interface_funcs->get_stats(keycache->keycache_cb, + partition_no, key_cache_stats); + } +} + + +/* + Get the value of a statistical variable for a key cache + + SYNOPSIS + get_key_cache_stat_value() + keycache pointer to the key cache to get statistics for + var_no the ordered number of a statistical variable + + DESCRIPTION + This function returns the value of the statistical variable var_no for + the key cache keycache. The variables are numbered starting from 0 to 6. + + RETURN + The value of the specified statistical variable. + + NOTES + Currently for any key cache the function can return values for the + following 7 statistical variables: + + Name Number + + blocks_used 0 + blocks_unused 1 + blocks_changed 2 + read_requests 3 + reads 4 + write_requests 5 + writes 6 +*/ + +ulonglong get_key_cache_stat_value(KEY_CACHE *keycache, uint var_no) +{ + if (keycache->key_cache_inited) + { + return keycache->interface_funcs->get_stat_val(keycache->keycache_cb, + var_no); + } + else + return 0; +} + + +/* + Repartition a key cache + + SYNOPSIS + repartition_key_cache() + keycache pointer to the key cache to be repartitioned + key_cache_block_size size of blocks to keep cached data + use_mem total memory to use for the new key cache + division_limit new division limit (if not zero) + age_threshold new age threshold (if not zero) + partitions new number of partitions in the key cache + + DESCRIPTION + The function operates over the key cache keycache. + The parameter partitions specifies the number of partitions in the key + cache after repartitioning. If the value of this parameter is 0 then + a simple key cache must be created instead of the old one. + The parameter key_cache_block_size specifies the new size of the block + buffers in the key cache. The parameters division_limit and age_threshold + determine the new initial values of those characteristics of the key cache + that are used for midpoint insertion strategy. The parameter use_mem + specifies the total amount of memory to be allocated for the new key + cache buffers and for all auxiliary structures. + + RETURN VALUE + number of blocks in the key cache, if successful, + 0 - otherwise. + + NOTES + The function does not block the calls and executions of other functions + from the key cache interface. However it assumes that the calls of + resize_key_cache itself are serialized. + + Currently the function is called when the value of the variable + key_cache_partitions is being reset for the key cache keycache. +*/ + +int repartition_key_cache(KEY_CACHE *keycache, uint key_cache_block_size, + size_t use_mem, uint division_limit, + uint age_threshold, uint partitions) +{ + uint blocks= -1; + if (keycache->key_cache_inited) + { + keycache->interface_funcs->resize(keycache->keycache_cb, + key_cache_block_size, 0, + division_limit, age_threshold); + end_key_cache(keycache, 1); + blocks= init_key_cache(keycache, key_cache_block_size, use_mem, + division_limit, age_threshold, partitions); + } + return blocks; +} + === modified file 'mysys/my_gethostbyname.c' --- a/mysys/my_gethostbyname.c 2010-03-30 12:36:49 +0000 +++ b/mysys/my_gethostbyname.c 2010-03-31 20:50:54 +0000 @@ -95,7 +95,7 @@ struct hostent * my_gethostbyname_r(const char *name, struct hostent *result __attribute__((unused)), char *buffer __attribute__((unused)), - int buflen__attribute__((unused)), + int buflen __attribute__((unused)), int *h_errnop) { struct hostent *hp; === modified file 'plugin/auth/auth_socket.c' --- a/plugin/auth/auth_socket.c 2010-03-29 15:13:53 +0000 +++ b/plugin/auth/auth_socket.c 2010-04-01 14:34:51 +0000 @@ -99,4 +99,21 @@ mysql_declare_plugin(socket_auth) NULL } mysql_declare_plugin_end; +maria_declare_plugin(socket_auth) +{ + MYSQL_AUTHENTICATION_PLUGIN, + &socket_auth_handler, + "socket_peercred", + "Sergei Golubchik", + "Unix Socket based authentication", + PLUGIN_LICENSE_GPL, + NULL, + NULL, + 0x0100, + NULL, + NULL, + "1.0", + MariaDB_PLUGIN_MATURITY_BETA +} +mysql_declare_plugin_end; === modified file 'plugin/auth/dialog.c' --- a/plugin/auth/dialog.c 2010-03-29 15:13:53 +0000 +++ b/plugin/auth/dialog.c 2010-04-08 13:58:04 +0000 @@ -51,6 +51,8 @@ #define PASSWORD_QUESTION "\4" #define LAST_PASSWORD "\5" +typedef unsigned char uchar; + /********************* SERVER SIDE ****************************************/ /** @@ -63,7 +65,9 @@ static int two_questions(MYSQL_PLUGIN_VI int pkt_len; /* send a password question */ - if (vio->write_packet(vio, PASSWORD_QUESTION "Password, please:", 18)) + if (vio->write_packet(vio, + (const uchar*) (PASSWORD_QUESTION "Password, please:"), + 18)) return CR_ERROR; /* read the answer */ @@ -73,11 +77,12 @@ static int two_questions(MYSQL_PLUGIN_VI info->password_used = 1; /* fail if the password is wrong */ - if (strcmp(pkt, info->auth_string)) + if (strcmp((char*) pkt, info->auth_string)) return CR_ERROR; /* send the last, ordinary, question */ - if (vio->write_packet(vio, LAST_QUESTION "Are you sure ?", 15)) + if (vio->write_packet(vio, + (const uchar*) (LAST_QUESTION "Are you sure ?"), 15)) return CR_ERROR; /* read the answer */ @@ -85,7 +90,7 @@ static int two_questions(MYSQL_PLUGIN_VI return CR_ERROR; /* check the reply */ - return strcmp(pkt, "yes, of course") ? CR_ERROR : CR_OK; + return strcmp((char*) pkt, "yes, of course") ? CR_ERROR : CR_OK; } static struct st_mysql_auth two_handler= @@ -108,7 +113,8 @@ static int three_attempts(MYSQL_PLUGIN_V for (i= 0; i < 3; i++) { /* send the prompt */ - if (vio->write_packet(vio, PASSWORD_QUESTION "Password, please:", 18)) + if (vio->write_packet(vio, + (const uchar*) (PASSWORD_QUESTION "Password, please:"), 18)) return CR_ERROR; /* read the password */ @@ -121,7 +127,7 @@ static int three_attempts(MYSQL_PLUGIN_V finish, if the password is correct. note, that we did not mark the prompt packet as "last" */ - if (strcmp(pkt, info->auth_string) == 0) + if (strcmp((char*) pkt, info->auth_string) == 0) return CR_OK; } @@ -165,6 +171,38 @@ mysql_declare_plugin(dialog) NULL } mysql_declare_plugin_end; +maria_declare_plugin(dialog) +{ + MYSQL_AUTHENTICATION_PLUGIN, + &two_handler, + "two_questions", + "Sergei Golubchik", + "Dialog plugin demo 1", + PLUGIN_LICENSE_GPL, + NULL, + NULL, + 0x0100, + NULL, + NULL, + "1.0", + MariaDB_PLUGIN_MATURITY_BETA +}, +{ + MYSQL_AUTHENTICATION_PLUGIN, + &three_handler, + "three_attempts", + "Sergei Golubchik", + "Dialog plugin demo 2", + PLUGIN_LICENSE_GPL, + NULL, + NULL, + 0x0100, + NULL, + NULL, + "1.0", + MariaDB_PLUGIN_MATURITY_BETA +} +maria_declare_plugin_end; /********************* CLIENT SIDE ***************************************/ /* @@ -186,13 +224,19 @@ static mysql_authentication_dialog_ask_t static char *builtin_ask(MYSQL *mysql __attribute__((unused)), int type __attribute__((unused)), const char *prompt, - char *buf, int buf_len __attribute__((unused))) + char *buf, int buf_len) { + int len; + fputs(prompt, stdout); fputc(' ', stdout); - if (gets(buf) == 0) + if (fgets(buf, buf_len, stdin) == 0) return 0; + len= strlen(buf); + if (len && buf[len-1]=='\n') + buf[len-1]=0; + return buf; } @@ -255,12 +299,12 @@ static int perform_dialog(MYSQL_PLUGIN_V if ((cmd >> 1) == 2 && *pkt == 0) reply= mysql->passwd; else - reply= ask(mysql, cmd >> 1, pkt, reply_buf, sizeof(reply_buf)); + reply= ask(mysql, cmd >> 1, (char*) pkt, reply_buf, sizeof(reply_buf)); if (!reply) return CR_ERROR; } /* send the reply to the server */ - res= vio->write_packet(vio, reply, strlen(reply)+1); + res= vio->write_packet(vio, (uchar*) reply, strlen(reply)+1); if (reply != mysql->passwd && reply != reply_buf) free(reply); === modified file 'plugin/daemon_example/daemon_example.cc' --- a/plugin/daemon_example/daemon_example.cc 2007-06-27 14:49:12 +0000 +++ b/plugin/daemon_example/daemon_example.cc 2010-04-01 14:34:51 +0000 @@ -200,3 +200,20 @@ mysql_declare_plugin(daemon_example) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(daemon_example) +{ + MYSQL_DAEMON_PLUGIN, + &daemon_example_plugin, + "daemon_example", + "Brian Aker", + "Daemon example, creates a heartbeat beat file in mysql-heartbeat.log", + PLUGIN_LICENSE_GPL, + daemon_example_plugin_init, /* Plugin Init */ + daemon_example_plugin_deinit, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_EXPERIMENTAL /* maturity */ +} +maria_declare_plugin_end; === modified file 'plugin/fulltext/plugin_example.c' --- a/plugin/fulltext/plugin_example.c 2010-01-04 17:54:42 +0000 +++ b/plugin/fulltext/plugin_example.c 2010-04-01 14:34:51 +0000 @@ -270,4 +270,21 @@ mysql_declare_plugin(ftexample) NULL } mysql_declare_plugin_end; +maria_declare_plugin(ftexample) +{ + MYSQL_FTPARSER_PLUGIN, /* type */ + &simple_parser_descriptor, /* descriptor */ + "simple_parser", /* name */ + "MySQL AB", /* author */ + "Simple Full-Text Parser", /* description */ + PLUGIN_LICENSE_GPL, + simple_parser_plugin_init, /* init function (when loaded) */ + simple_parser_plugin_deinit,/* deinit function (when unloaded) */ + 0x0001, /* version */ + simple_status, /* status variables */ + simple_system_variables, /* system variables */ + "0.01", /* string version */ + MariaDB_PLUGIN_MATURITY_EXPERIMENTAL /* maturity */ +} +maria_declare_plugin_end; === modified file 'scripts/mysql_system_tables.sql' --- a/scripts/mysql_system_tables.sql 2010-03-29 15:13:53 +0000 +++ b/scripts/mysql_system_tables.sql 2010-04-08 12:10:05 +0000 @@ -60,7 +60,7 @@ CREATE TABLE IF NOT EXISTS time_zone_tra CREATE TABLE IF NOT EXISTS time_zone_leap_second ( Transition_time bigint signed NOT NULL, Correction int signed NOT NULL, PRIMARY KEY TranTime (Transition_time) ) engine=MyISAM CHARACTER SET utf8 comment='Leap seconds information for time zones'; -CREATE TABLE IF NOT EXISTS proc (db char(64) collate utf8_bin DEFAULT '' NOT NULL, name char(64) DEFAULT '' NOT NULL, type enum('FUNCTION','PROCEDURE') NOT NULL, specific_name char(64) DEFAULT '' NOT NULL, language enum('SQL') DEFAULT 'SQL' NOT NULL, sql_data_access enum( 'CONTAINS_SQL', 'NO_SQL', 'READS_SQL_DATA', 'MODIFIES_SQL_DATA') DEFAULT 'CONTAINS_SQL' NOT NULL, is_deterministic enum('YES','NO') DEFAULT 'NO' NOT NULL, security_type enum('INVOKER','DEFINER') DEFAULT 'DEFINER' NOT NULL, param_list blob NOT NULL, returns longblob DEFAULT '' NOT NULL, body longblob NOT NULL, definer char(77) collate utf8_bin DEFAULT '' NOT NULL, created timestamp, modified timestamp, sql_mode set( 'REAL_AS_FLOAT', 'PIPES_AS_CONCAT', 'ANSI_QUOTES', 'IGNORE_SPACE', 'NOT_USED', 'ONLY_FULL_GROUP_BY', 'NO_UNSIGNED_SUBTRACTION', 'NO_DIR_IN_CREATE', 'POSTGRESQL', 'ORACLE', 'MSSQL', 'DB2', 'MAXDB', 'NO_KEY_OPTIONS', 'NO_TABLE_OPTIONS', 'NO_FIELD_OPTIONS', 'MYSQL323', 'MYSQL40', 'ANSI', 'NO_AUTO_VALUE_ON_ZERO', 'NO_BACKSLASH_ESCAPES', 'STRICT_TRANS_TABLES', 'STRICT_ALL_TABLES', 'NO_ZERO_IN_DATE', 'NO_ZERO_DATE', 'INVALID_DATES', 'ERROR_FOR_DIVISION_BY_ZERO', 'TRADITIONAL', 'NO_AUTO_CREATE_USER', 'HIGH_NOT_PRECEDENCE', 'NO_ENGINE_SUBSTITUTION', 'PAD_CHAR_TO_FULL_LENGTH') DEFAULT '' NOT NULL, comment char(64) collate utf8_bin DEFAULT '' NOT NULL, character_set_client char(32) collate utf8_bin, collation_connection char(32) collate utf8_bin, db_collation char(32) collate utf8_bin, body_utf8 longblob, PRIMARY KEY (db,name,type)) engine=MyISAM character set utf8 comment='Stored Procedures'; +CREATE TABLE IF NOT EXISTS proc (db char(64) collate utf8_bin DEFAULT '' NOT NULL, name char(64) DEFAULT '' NOT NULL, type enum('FUNCTION','PROCEDURE') NOT NULL, specific_name char(64) DEFAULT '' NOT NULL, language enum('SQL') DEFAULT 'SQL' NOT NULL, sql_data_access enum( 'CONTAINS_SQL', 'NO_SQL', 'READS_SQL_DATA', 'MODIFIES_SQL_DATA') DEFAULT 'CONTAINS_SQL' NOT NULL, is_deterministic enum('YES','NO') DEFAULT 'NO' NOT NULL, security_type enum('INVOKER','DEFINER') DEFAULT 'DEFINER' NOT NULL, param_list blob NOT NULL, returns longblob DEFAULT '' NOT NULL, body longblob NOT NULL, definer char(77) collate utf8_bin DEFAULT '' NOT NULL, created timestamp, modified timestamp, sql_mode set( 'REAL_AS_FLOAT', 'PIPES_AS_CONCAT', 'ANSI_QUOTES', 'IGNORE_SPACE', 'IGNORE_BAD_TABLE_OPTIONS', 'ONLY_FULL_GROUP_BY', 'NO_UNSIGNED_SUBTRACTION', 'NO_DIR_IN_CREATE', 'POSTGRESQL', 'ORACLE', 'MSSQL', 'DB2', 'MAXDB', 'NO_KEY_OPTIONS', 'NO_TABLE_OPTIONS', 'NO_FIELD_OPTIONS', 'MYSQL323', 'MYSQL40', 'ANSI', 'NO_AUTO_VALUE_ON_ZERO', 'NO_BACKSLASH_ESCAPES', 'STRICT_TRANS_TABLES', 'STRICT_ALL_TABLES', 'NO_ZERO_IN_DATE', 'NO_ZERO_DATE', 'INVALID_DATES', 'ERROR_FOR_DIVISION_BY_ZERO', 'TRADITIONAL', 'NO_AUTO_CREATE_USER', 'HIGH_NOT_PRECEDENCE', 'NO_ENGINE_SUBSTITUTION', 'PAD_CHAR_TO_FULL_LENGTH') DEFAULT '' NOT NULL, comment char(64) collate utf8_bin DEFAULT '' NOT NULL, character_set_client char(32) collate utf8_bin, collation_connection char(32) collate utf8_bin, db_collation char(32) collate utf8_bin, body_utf8 longblob, PRIMARY KEY (db,name,type)) engine=MyISAM character set utf8 comment='Stored Procedures'; CREATE TABLE IF NOT EXISTS procs_priv ( Host char(60) binary DEFAULT '' NOT NULL, Db char(64) binary DEFAULT '' NOT NULL, User char(16) binary DEFAULT '' NOT NULL, Routine_name char(64) COLLATE utf8_general_ci DEFAULT '' NOT NULL, Routine_type enum('FUNCTION','PROCEDURE') NOT NULL, Grantor char(77) DEFAULT '' NOT NULL, Proc_priv set('Execute','Alter Routine','Grant') COLLATE utf8_general_ci DEFAULT '' NOT NULL, Timestamp timestamp(14), PRIMARY KEY (Host,Db,User,Routine_name,Routine_type), KEY Grantor (Grantor) ) engine=MyISAM CHARACTER SET utf8 COLLATE utf8_bin comment='Procedure privileges'; @@ -80,7 +80,7 @@ PREPARE stmt FROM @str; EXECUTE stmt; DROP PREPARE stmt; -CREATE TABLE IF NOT EXISTS event ( db char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', name char(64) CHARACTER SET utf8 NOT NULL default '', body longblob NOT NULL, definer char(77) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', execute_at DATETIME default NULL, interval_value int(11) default NULL, interval_field ENUM('YEAR','QUARTER','MONTH','DAY','HOUR','MINUTE','WEEK','SECOND','MICROSECOND','YEAR_MONTH','DAY_HOUR','DAY_MINUTE','DAY_SECOND','HOUR_MINUTE','HOUR_SECOND','MINUTE_SECOND','DAY_MICROSECOND','HOUR_MICROSECOND','MINUTE_MICROSECOND','SECOND_MICROSECOND') default NULL, created TIMESTAMP NOT NULL, modified TIMESTAMP NOT NULL, last_executed DATETIME default NULL, starts DATETIME default NULL, ends DATETIME default NULL, status ENUM('ENABLED','DISABLED','SLAVESIDE_DISABLED') NOT NULL default 'ENABLED', on_completion ENUM('DROP','PRESERVE') NOT NULL default 'DROP', sql_mode set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') DEFAULT '' NOT NULL, comment char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', originator INTEGER UNSIGNED NOT NULL, time_zone char(64) CHARACTER SET latin1 NOT NULL DEFAULT 'SYSTEM', character_set_client char(32) collate utf8_bin, collation_connection char(32) collate utf8_bin, db_collation char(32) collate utf8_bin, body_utf8 longblob, PRIMARY KEY (db, name) ) ENGINE=MyISAM DEFAULT CHARSET=utf8 COMMENT 'Events'; +CREATE TABLE IF NOT EXISTS event ( db char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', name char(64) CHARACTER SET utf8 NOT NULL default '', body longblob NOT NULL, definer char(77) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', execute_at DATETIME default NULL, interval_value int(11) default NULL, interval_field ENUM('YEAR','QUARTER','MONTH','DAY','HOUR','MINUTE','WEEK','SECOND','MICROSECOND','YEAR_MONTH','DAY_HOUR','DAY_MINUTE','DAY_SECOND','HOUR_MINUTE','HOUR_SECOND','MINUTE_SECOND','DAY_MICROSECOND','HOUR_MICROSECOND','MINUTE_MICROSECOND','SECOND_MICROSECOND') default NULL, created TIMESTAMP NOT NULL, modified TIMESTAMP NOT NULL, last_executed DATETIME default NULL, starts DATETIME default NULL, ends DATETIME default NULL, status ENUM('ENABLED','DISABLED','SLAVESIDE_DISABLED') NOT NULL default 'ENABLED', on_completion ENUM('DROP','PRESERVE') NOT NULL default 'DROP', sql_mode set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES','IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION','NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB','NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40','ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES','STRICT_ALL_TABLES','NO_ZERO_IN_DATE','NO_ZERO_DATE','INVALID_DATES','ERROR_FOR_DIVISION_BY_ZERO','TRADITIONAL','NO_AUTO_CREATE_USER','HIGH_NOT_PRECEDENCE','NO_ENGINE_SUBSTITUTION','PAD_CHAR_TO_FULL_LENGTH') DEFAULT '' NOT NULL, comment char(64) CHARACTER SET utf8 COLLATE utf8_bin NOT NULL default '', originator INTEGER UNSIGNED NOT NULL, time_zone char(64) CHARACTER SET latin1 NOT NULL DEFAULT 'SYSTEM', character_set_client char(32) collate utf8_bin, collation_connection char(32) collate utf8_bin, db_collation char(32) collate utf8_bin, body_utf8 longblob, PRIMARY KEY (db, name) ) ENGINE=MyISAM DEFAULT CHARSET=utf8 COMMENT 'Events'; CREATE TABLE IF NOT EXISTS ndb_binlog_index (Position BIGINT UNSIGNED NOT NULL, File VARCHAR(255) NOT NULL, epoch BIGINT UNSIGNED NOT NULL, inserts BIGINT UNSIGNED NOT NULL, updates BIGINT UNSIGNED NOT NULL, deletes BIGINT UNSIGNED NOT NULL, schemaops BIGINT UNSIGNED NOT NULL, PRIMARY KEY(epoch)) ENGINE=MYISAM; === modified file 'scripts/mysql_system_tables_fix.sql' --- a/scripts/mysql_system_tables_fix.sql 2010-03-29 15:13:53 +0000 +++ b/scripts/mysql_system_tables_fix.sql 2010-04-08 12:10:05 +0000 @@ -368,7 +368,7 @@ ALTER TABLE proc MODIFY name char(64) DE 'PIPES_AS_CONCAT', 'ANSI_QUOTES', 'IGNORE_SPACE', - 'NOT_USED', + 'IGNORE_BAD_TABLE_OPTIONS', 'ONLY_FULL_GROUP_BY', 'NO_UNSIGNED_SUBTRACTION', 'NO_DIR_IN_CREATE', @@ -482,14 +482,14 @@ ALTER TABLE db MODIFY Event_priv enum('N ALTER TABLE event DROP PRIMARY KEY; ALTER TABLE event ADD PRIMARY KEY(db, name); # Add sql_mode column just in case. -ALTER TABLE event ADD sql_mode set ('NOT_USED') AFTER on_completion; +ALTER TABLE event ADD sql_mode set ('IGNORE_BAD_TABLE_OPTIONS') AFTER on_completion; # Update list of sql_mode values. ALTER TABLE event MODIFY sql_mode set('REAL_AS_FLOAT', 'PIPES_AS_CONCAT', 'ANSI_QUOTES', 'IGNORE_SPACE', - 'NOT_USED', + 'IGNORE_BAD_TABLE_OPTIONS', 'ONLY_FULL_GROUP_BY', 'NO_UNSIGNED_SUBTRACTION', 'NO_DIR_IN_CREATE', === modified file 'server-tools/CMakeLists.txt' --- a/server-tools/CMakeLists.txt 2007-02-14 11:42:11 +0000 +++ b/server-tools/CMakeLists.txt 2010-04-02 09:20:09 +0000 @@ -27,7 +27,7 @@ ADD_EXECUTABLE(mysqlmanager buffer.cc co user_management_commands.cc ../../sql/net_serv.cc ../../sql-common/pack.c ../../sql/password.c ../../sql/sql_state.c ../../sql-common/client.c ../../libmysql/get_password.c - ../../libmysql/errmsg.c) + ../../libmysql/errmsg.c ../../sql-common/client_plugin.c) ADD_DEPENDENCIES(mysqlmanager GenError) TARGET_LINK_LIBRARIES(mysqlmanager dbug mysys strings taocrypt vio yassl zlib wsock32) === modified file 'server-tools/instance-manager/CMakeLists.txt' --- a/server-tools/instance-manager/CMakeLists.txt 2007-12-27 17:16:02 +0000 +++ b/server-tools/instance-manager/CMakeLists.txt 2010-04-02 09:20:09 +0000 @@ -28,7 +28,7 @@ ADD_EXECUTABLE(mysqlmanager buffer.cc co user_management_commands.cc ../../mysys/my_rnd.c ../../sql/net_serv.cc ../../sql-common/pack.c ../../sql/password.c ../../sql/sql_state.c ../../sql-common/client.c ../../libmysql/get_password.c - ../../libmysql/errmsg.c) + ../../libmysql/errmsg.c ../../sql-common/client_plugin.c) ADD_DEPENDENCIES(mysqlmanager GenError) TARGET_LINK_LIBRARIES(mysqlmanager debug dbug mysys strings taocrypt vio yassl zlib wsock32) === modified file 'sql-common/client.c' --- a/sql-common/client.c 2010-03-31 08:17:03 +0000 +++ b/sql-common/client.c 2010-04-09 09:26:17 +0000 @@ -2345,7 +2345,7 @@ void mpvio_info(Vio *vio, MYSQL_PLUGIN_V case VIO_TYPE_SSL: { struct sockaddr addr; - socklen_t addrlen= sizeof(addr); + SOCKET_SIZE_TYPE addrlen= sizeof(addr); if (getsockname(vio->sd, &addr, &addrlen)) return; info->protocol= addr.sa_family == AF_UNIX ? @@ -2360,7 +2360,7 @@ void mpvio_info(Vio *vio, MYSQL_PLUGIN_V return; case VIO_TYPE_SHARED_MEMORY: info->protocol= MYSQL_VIO_MEMORY; - info->handle= vio->handle_client_file_map; /* or what ? */ + info->handle= vio->handle_file_map; /* or what ? */ return; #endif default: DBUG_ASSERT(0); @@ -2394,7 +2394,7 @@ static void client_mpvio_info(MYSQL_PLUG */ int run_plugin_auth(MYSQL *mysql, char *data, uint data_len, - char *data_plugin, const char *db) + const char *data_plugin, const char *db) { const char *auth_plugin_name; auth_plugin_t *auth_plugin; @@ -2540,7 +2540,8 @@ CLI_MYSQL_REAL_CONNECT(MYSQL *mysql,cons char buff[NAME_LEN+USERNAME_LENGTH+100]; int scramble_data_len, pkt_scramble_len; char *end, *host_info=0, *server_version_end, *pkt_end; - char *scramble_data, *scramble_plugin; + char *scramble_data; + const char *scramble_plugin; my_socket sock; in_addr_t ip_addr; struct sockaddr_in sock_addr; === modified file 'sql/CMakeLists.txt' --- a/sql/CMakeLists.txt 2010-03-29 15:13:53 +0000 +++ b/sql/CMakeLists.txt 2010-04-08 12:10:05 +0000 @@ -77,6 +77,7 @@ SET (SQL_SOURCE rpl_rli.cc rpl_mi.cc sql_servers.cc sql_connect.cc scheduler.cc sql_profile.cc event_parse_data.cc opt_table_elimination.cc + create_options.cc ${PROJECT_SOURCE_DIR}/sql/sql_yacc.cc ${PROJECT_SOURCE_DIR}/sql/sql_yacc.h ${PROJECT_SOURCE_DIR}/include/mysqld_error.h === modified file 'sql/Makefile.am' --- a/sql/Makefile.am 2010-03-29 15:13:53 +0000 +++ b/sql/Makefile.am 2010-04-08 12:10:05 +0000 @@ -78,7 +78,8 @@ noinst_HEADERS = item.h item_func.h item sql_plugin.h authors.h event_parse_data.h \ event_data_objects.h event_scheduler.h \ sql_partition.h partition_info.h partition_element.h \ - contributors.h sql_servers.h + contributors.h sql_servers.h \ + create_options.h mysqld_SOURCES = sql_lex.cc sql_handler.cc sql_partition.cc \ item.cc item_sum.cc item_buff.cc item_func.cc \ @@ -124,7 +125,7 @@ mysqld_SOURCES = sql_lex.cc sql_handler. sql_plugin.cc sql_binlog.cc \ sql_builtin.cc sql_tablespace.cc partition_info.cc \ sql_servers.cc event_parse_data.cc \ - opt_table_elimination.cc + opt_table_elimination.cc create_options.cc nodist_mysqld_SOURCES = mini_client_errors.c pack.c client.c my_time.c my_user.c client_plugin.c === modified file 'sql/client_settings.h' --- a/sql/client_settings.h 2010-03-29 15:13:53 +0000 +++ b/sql/client_settings.h 2010-04-02 09:20:09 +0000 @@ -29,7 +29,6 @@ #define mysql_master_send_query(A, B, C) 1 #define mysql_slave_send_query(A, B, C) 1 #define mysql_rpl_probe(mysql) 0 -#undef HAVE_SMEM #undef _CUSTOMCONFIG_ #define mysql_server_init(a,b,c) mysql_client_plugin_init() === added file 'sql/create_options.cc' --- a/sql/create_options.cc 1970-01-01 00:00:00 +0000 +++ b/sql/create_options.cc 2010-04-08 12:10:05 +0000 @@ -0,0 +1,612 @@ +/* Copyright (C) 2010 Monty Program Ab + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; version 2 of the License. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + +/** + @file + + Engine defined options of tables/fields/keys in CREATE/ALTER TABLE. +*/ + +#include "mysql_priv.h" +#include "create_options.h" +#include <my_getopt.h> + +#define FRM_QUOTED_VALUE 0x8000 + +/** + Links this item to the given list end + + @param start The list beginning or NULL + @param end The list last element or does not matter +*/ + +void engine_option_value::link(engine_option_value **start, + engine_option_value **end) +{ + DBUG_ENTER("engine_option_value::link"); + DBUG_PRINT("enter", ("name: '%s' (%u) value: '%s' (%u)", + name.str, (uint) name.length, + value.str, (uint) value.length)); + engine_option_value *opt; + /* check duplicates to avoid writing them to frm*/ + for(opt= *start; + opt && ((opt->parsed && !opt->value.str) || + my_strnncoll(system_charset_info, + (uchar *)name.str, name.length, + (uchar*)opt->name.str, opt->name.length)); + opt= opt->next) /* no-op */; + if (opt) + { + opt->value.str= NULL; /* remove previous value */ + opt->parsed= TRUE; /* and don't issue warnings for it anymore */ + } + /* + Add this option to the end of the list + + @note: We add even if it is opt->value.str == NULL because it can be + ALTER TABLE to remove the option. + */ + if (*start) + { + (*end)->next= this; + *end= this; + } + else + { + /* + note that is *start == 0, the value of *end does not matter, + it can be uninitialized. + */ + *start= *end= this; + } + DBUG_VOID_RETURN; +} + +static bool report_wrong_value(THD *thd, const char *name, const char *val, + my_bool suppress_warning) +{ + if (!(thd->variables.sql_mode & MODE_IGNORE_BAD_TABLE_OPTIONS)) + { + my_error(ER_BAD_OPTION_VALUE, MYF(0), val, name); + return 1; + } + + /* + We may need to suppress warnings to avoid duplicate messages + about the same option (option list is parsed more than once during + CREATE/ALTER table). + Errors are not suppressed, as they abort the execution on the first parsing. + */ + if (!suppress_warning) + push_warning_printf(thd, MYSQL_ERROR::WARN_LEVEL_WARN, + ER_BAD_OPTION_VALUE, + ER(ER_BAD_OPTION_VALUE), val, name); + return 0; +} + +static bool report_unknown_option(THD *thd, engine_option_value *val, + my_bool suppress_warning) +{ + DBUG_ENTER("report_unknown_option"); + if (val->parsed) + { + DBUG_PRINT("info", ("parsed => exiting")); + DBUG_RETURN(FALSE); + } + + if (!(thd->variables.sql_mode & MODE_IGNORE_BAD_TABLE_OPTIONS)) + { + my_error(ER_UNKNOWN_OPTION, MYF(0), val->name.str); + DBUG_RETURN(TRUE); + } + + if (!suppress_warning) + push_warning_printf(thd, MYSQL_ERROR::WARN_LEVEL_WARN, + ER_UNKNOWN_OPTION, + ER(ER_UNKNOWN_OPTION), + val->name.str); + DBUG_RETURN(FALSE); +} + +static bool set_one_value(ha_create_table_option *opt, + THD *thd, LEX_STRING *value, void *base, + my_bool suppress_warning, + MEM_ROOT *root) +{ + DBUG_ENTER("set_one_value"); + DBUG_PRINT("enter", ("opt: 0x%lx type: %u name '%s' value: '%s'", + (ulong) opt, + opt->type, opt->name, + (value->str ? value->str : "<DEFAULT>"))); + switch (opt->type) + { + case HA_OPTION_TYPE_ULL: + { + ulonglong *val= (ulonglong*)((char*)base + opt->offset); + if (!value->str) + { + *val= opt->def_value; + DBUG_RETURN(0); + } + + my_option optp= { opt->name, 1, 0, (uchar **)val, 0, 0, GET_ULL, + REQUIRED_ARG, opt->def_value, opt->min_value, opt->max_value, + 0, opt->block_size, 0}; + + ulonglong orig_val= strtoull(value->str, NULL, 10); + my_bool unused; + *val= orig_val; + *val= getopt_ull_limit_value(*val, &optp, &unused); + if (*val == orig_val) + DBUG_RETURN(0); + + DBUG_RETURN(report_wrong_value(thd, opt->name, value->str, + suppress_warning)); + } + case HA_OPTION_TYPE_STRING: + { + char **val= (char **)((char *)base + opt->offset); + if (!value->str) + { + *val= 0; + DBUG_RETURN(0); + } + + if (!(*val= strmake_root(root, value->str, value->length))) + DBUG_RETURN(1); + DBUG_RETURN(0); + } + case HA_OPTION_TYPE_ENUM: + { + uint *val= (uint *)((char *)base + opt->offset), num; + + *val= opt->def_value; + if (!value->str) + DBUG_RETURN(0); + + const char *start= opt->values, *end; + + num= 0; + while (*start) + { + for (end=start; + *end && *end != ','; + end+= my_mbcharlen(system_charset_info, *end)) /* no-op */; + if (!my_strnncoll(system_charset_info, + (uchar*)start, end-start, + (uchar*)value->str, value->length)) + { + *val= num; + DBUG_RETURN(0); + } + if (*end) *end++; + start= end; + num++; + } + + DBUG_RETURN(report_wrong_value(thd, opt->name, value->str, + suppress_warning)); + } + case HA_OPTION_TYPE_BOOL: + { + bool *val= (bool *)((char *)base + opt->offset); + *val= opt->def_value; + + if (!value->str) + DBUG_RETURN(0); + + if (!my_strnncoll(system_charset_info, + (const uchar*)"NO", 2, + (uchar *)value->str, value->length) || + !my_strnncoll(system_charset_info, + (const uchar*)"OFF", 3, + (uchar *)value->str, value->length) || + !my_strnncoll(system_charset_info, + (const uchar*)"0", 1, + (uchar *)value->str, value->length)) + { + *val= FALSE; + DBUG_RETURN(FALSE); + } + + if (!my_strnncoll(system_charset_info, + (const uchar*)"YES", 3, + (uchar *)value->str, value->length) || + !my_strnncoll(system_charset_info, + (const uchar*)"ON", 2, + (uchar *)value->str, value->length) || + !my_strnncoll(system_charset_info, + (const uchar*)"1", 1, + (uchar *)value->str, value->length)) + { + *val= TRUE; + DBUG_RETURN(FALSE); + } + + DBUG_RETURN(report_wrong_value(thd, opt->name, value->str, + suppress_warning)); + } + } + DBUG_ASSERT(0); + my_error(ER_UNKNOWN_ERROR, MYF(0)); + DBUG_RETURN(1); +} + +static const size_t ha_option_type_sizeof[]= +{ sizeof(ulonglong), sizeof(char *), sizeof(uint), sizeof(bool)}; + +/** + Creates option structure and parses list of options in it + + @param thd thread handler + @param option_struct where to store pointer on the option struct + @param option_list list of options given by user + @param rules list of option description by engine + @param suppress_warning second parse so we do not need warnings + @param root MEM_ROOT where allocate memory + + @retval TRUE Error + @retval FALSE OK +*/ + +my_bool parse_option_list(THD* thd, void **option_struct, + engine_option_value *option_list, + ha_create_table_option *rules, + my_bool suppress_warning, + MEM_ROOT *root) +{ + ha_create_table_option *opt; + size_t option_struct_size= 0; + engine_option_value *val= option_list; + DBUG_ENTER("parse_option_list"); + DBUG_PRINT("enter", + ("struct: 0x%lx list: 0x%lx rules: 0x%lx suppres %u root 0x%lx", + (ulong) *option_struct, (ulong)option_list, (ulong)rules, + (uint) suppress_warning, (ulong) root)); + + if (rules) + { + LEX_STRING default_val= {NULL, 0}; + for (opt= rules; opt->name; opt++) + set_if_bigger(option_struct_size, opt->offset + + ha_option_type_sizeof[opt->type]); + + *option_struct= alloc_root(root, option_struct_size); + + /* set all values to default */ + for (opt= rules; opt->name; opt++) + set_one_value(opt, thd, &default_val, *option_struct, + suppress_warning, root); + } + + for (; val; val= val->next) + { + for (opt= rules; opt && opt->name; opt++) + { + if (my_strnncoll(system_charset_info, + (uchar*)opt->name, opt->name_length, + (uchar*)val->name.str, val->name.length)) + continue; + + if (set_one_value(opt, thd, &val->value, + *option_struct, suppress_warning || val->parsed, root)) + DBUG_RETURN(TRUE); + val->parsed= true; + break; + } + if (report_unknown_option(thd, val, suppress_warning)) + DBUG_RETURN(TRUE); + val->parsed= true; + } + + DBUG_RETURN(FALSE); +} + + +/** + Parses all table/fields/keys options + + @param thd thread handler + @param file handler of the table + @parem share descriptor of the table + + @retval TRUE Error + @retval FALSE OK +*/ + +my_bool parse_engine_table_options(THD *thd, handlerton *ht, + TABLE_SHARE *share) +{ + MEM_ROOT *root= &share->mem_root; + DBUG_ENTER("parse_engine_table_options"); + + if (parse_option_list(thd, &share->option_struct, share->option_list, + ht->table_options, TRUE, root)) + DBUG_RETURN(TRUE); + + for (Field **field= share->field; *field; field++) + { + if (parse_option_list(thd, &(*field)->option_struct, (*field)->option_list, + ht->field_options, TRUE, root)) + DBUG_RETURN(TRUE); + } + + for (uint index= 0; index < share->keys; index ++) + { + if (parse_option_list(thd, &share->key_info[index].option_struct, + share->key_info[index].option_list, + ht->index_options, TRUE, root)) + DBUG_RETURN(TRUE); + } + + DBUG_RETURN(FALSE); +} + + +/** + Returns representation length of key and value in the frm file +*/ + +uint engine_option_value::frm_length() +{ + /* + 1 byte - name length + 2 bytes - value length + + if value.str is NULL, this option is not written to frm (=DEFAULT) + */ + return value.str ? 1 + name.length + 2 + value.length : 0; +} + + +/** + Returns length of representation of option list in the frm file +*/ + +static uint option_list_frm_length(engine_option_value *opt) +{ + uint res= 0; + + for (; opt; opt= opt->next) + res+= opt->frm_length(); + + return res; +} + + +/** + Calculates length of options image in the .frm + + @param table_option_list list of table options + @param create_fields field descriptors list + @param keys number of keys + @param key_info array of key descriptors + + @returns length of image in frm +*/ + +uint engine_table_options_frm_length(engine_option_value *table_option_list, + List<Create_field> &create_fields, + uint keys, KEY *key_info) +{ + List_iterator<Create_field> it(create_fields); + Create_field *field; + uint res, index; + DBUG_ENTER("engine_table_options_frm_length"); + + res= option_list_frm_length(table_option_list); + + while ((field= it++)) + res+= option_list_frm_length(field->option_list); + + for (index= 0; index < keys; index++, key_info++) + res+= option_list_frm_length(key_info->option_list); + + /* + if there's at least one option somewhere (res > 0) + we write option lists for all fields and keys, zero-terminated. + If there're no options we write nothing at all (backward compatibility) + */ + DBUG_RETURN(res ? res + 1 + create_fields.elements + keys : 0); +} + + +/** + Writes image of the key and value to the frm image buffer + + @param buff pointer to the buffer free space beginning + + @returns pointer to byte after last recorded in the buffer +*/ + +uchar *engine_option_value::frm_image(uchar *buff) +{ + if (value.str) + { + *buff++= name.length; + memcpy(buff, name.str, name.length); + buff+= name.length; + int2store(buff, value.length | (quoted_value ? FRM_QUOTED_VALUE : 0)); + buff+= 2; + memcpy(buff, (const uchar *) value.str, value.length); + buff+= value.length; + } + return buff; +} + +/** + Writes image of the key and value to the frm image buffer + + @param buff pointer to the buffer to store the options in + @param opt list of options; + + @returns pointer to the end of the stored data in the buffer +*/ +static uchar *option_list_frm_image(uchar *buff, engine_option_value *opt) +{ + for (; opt; opt= opt->next) + buff= opt->frm_image(buff); + + *buff++= 0; + return buff; +} + + +/** + Writes options image in the .frm buffer + + @param buff pointer to the buffer + @param table_option_list list of table options + @param create_fields field descriptors list + @param keys number of keys + @param key_info array of key descriptors + + @returns pointer to byte after last recorded in the buffer +*/ + +uchar *engine_table_options_frm_image(uchar *buff, + engine_option_value *table_option_list, + List<Create_field> &create_fields, + uint keys, KEY *key_info) +{ + List_iterator<Create_field> it(create_fields); + Create_field *field; + KEY *key_info_end= key_info + keys; + DBUG_ENTER("engine_table_options_frm_image"); + + buff= option_list_frm_image(buff, table_option_list); + + while ((field= it++)) + buff= option_list_frm_image(buff, field->option_list); + + while (key_info < key_info_end) + buff= option_list_frm_image(buff, (key_info++)->option_list); + + DBUG_RETURN(buff); +} + +/** + Reads name and value from buffer, then link it in the list + + @param buff the buffer to read from + @param start The list beginning or NULL + @param end The list last element or does not matter + @param root MEM_ROOT for allocating + + @returns pointer to byte after last recorded in the buffer +*/ +uchar *engine_option_value::frm_read(const uchar *buff, engine_option_value **start, + engine_option_value **end, MEM_ROOT *root) +{ + LEX_STRING name, value; + uint len; + + name.length= buff[0]; + buff++; + if (!(name.str= strmake_root(root, (const char*)buff, name.length))) + return NULL; + buff+= name.length; + len= uint2korr(buff); + value.length= len & ~FRM_QUOTED_VALUE; + buff+= 2; + if (!(value.str= strmake_root(root, (const char*)buff, value.length))) + return NULL; + buff+= value.length; + + engine_option_value *ptr=new (root) + engine_option_value(name, value, len & FRM_QUOTED_VALUE, start, end); + if (!ptr) + return NULL; + + return (uchar *)buff; +} + + +/** + Reads options from this buffer + + @param buff the buffer to read from + @param length buffer length + @param share table descriptor + @param root MEM_ROOT for allocating + + @retval TRUE Error + @retval FALSE OK +*/ + +my_bool engine_table_options_frm_read(const uchar *buff, uint length, + TABLE_SHARE *share) +{ + const uchar *buff_end= buff + length; + engine_option_value *end; + MEM_ROOT *root= &share->mem_root; + uint count; + DBUG_ENTER("engine_table_options_frm_read"); + + while (buff < buff_end && *buff) + { + if (!(buff= engine_option_value::frm_read(buff, &share->option_list, &end, + root))) + DBUG_RETURN(TRUE); + } + buff++; + + for (count=0; count < share->fields; count++) + { + while (buff < buff_end && *buff) + { + if (!(buff= engine_option_value::frm_read(buff, + &share->field[count]->option_list, + &end, root))) + DBUG_RETURN(TRUE); + } + buff++; + } + + for (count=0; count < share->keys; count++) + { + while (buff < buff_end && *buff) + { + if (!(buff= engine_option_value::frm_read(buff, + &share->key_info[count].option_list, + &end, root))) + DBUG_RETURN(TRUE); + } + buff++; + } + + DBUG_RETURN(buff != buff_end); +} + +/** + Merges two lists of engine_option_value's with duplicate removal. +*/ + +engine_option_value *merge_engine_table_options(engine_option_value *first, + engine_option_value *second, + MEM_ROOT *root) +{ + engine_option_value *end, *opt; + DBUG_ENTER("merge_engine_table_options"); + + /* find last element */ + if (first && second) + for (end= first; end->next; end= end->next) /* no-op */; + + for (opt= second; opt; opt= opt->next) + new (root) engine_option_value(opt->name, opt->value, opt->quoted_value, + &first, &end); + DBUG_RETURN(first); +} === added file 'sql/create_options.h' --- a/sql/create_options.h 1970-01-01 00:00:00 +0000 +++ b/sql/create_options.h 2010-04-08 12:10:05 +0000 @@ -0,0 +1,92 @@ +/* Copyright (C) 2010 Monty Program Ab + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; version 2 of the License. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + +/** + @file + + Engine defined options of tables/fields/keys in CREATE/ALTER TABLE. +*/ + +#ifndef SQL_CREATE_OPTIONS_INCLUDED +#define SQL_CREATE_OPTIONS_INCLUDED + +#include "handler.h" + +class engine_option_value: public Sql_alloc +{ + public: + LEX_STRING name; + LEX_STRING value; + engine_option_value *next; ///< parser puts them in a FIFO linked list + bool parsed; ///< to detect unrecognized options + bool quoted_value; ///< option=VAL vs. option='VAL' + + engine_option_value(LEX_STRING &name_arg, LEX_STRING &value_arg, bool quoted, + engine_option_value **start, engine_option_value **end) : + name(name_arg), value(value_arg), + next(NULL), parsed(false), quoted_value(quoted) + { + link(start, end); + } + engine_option_value(LEX_STRING &name_arg, + engine_option_value **start, engine_option_value **end) : + name(name_arg), value(null_lex_str), + next(NULL), parsed(false), quoted_value(false) + { + link(start, end); + } + engine_option_value(LEX_STRING &name_arg, ulonglong value_arg, + engine_option_value **start, engine_option_value **end, + MEM_ROOT *root) : + name(name_arg), next(NULL), parsed(false), quoted_value(false) + { + if ((value.str= (char *)alloc_root(root, 22))) + { + value.length= longlong10_to_str(value_arg, value.str, 10) - value.str; + link(start, end); + } + } + static uchar *frm_read(const uchar *buff, engine_option_value **start, + engine_option_value **end, MEM_ROOT *root); + void link(engine_option_value **start, engine_option_value **end); + uint frm_length(); + uchar *frm_image(uchar *buff); +}; + +typedef struct st_key KEY; +class Create_field; + +my_bool parse_engine_table_options(THD *thd, handlerton *ht, + TABLE_SHARE *share); +my_bool parse_option_list(THD* thd, void **option_struct, + engine_option_value *option_list, + ha_create_table_option *rules, + my_bool suppress_warning, + MEM_ROOT *root); +my_bool engine_table_options_frm_read(const uchar *buff, + uint length, + TABLE_SHARE *share); +engine_option_value *merge_engine_table_options(engine_option_value *source, + engine_option_value *changes, + MEM_ROOT *root); + +uint engine_table_options_frm_length(engine_option_value *table_option_list, + List<Create_field> &create_fields, + uint keys, KEY *key_info); +uchar *engine_table_options_frm_image(uchar *buff, + engine_option_value *table_option_list, + List<Create_field> &create_fields, + uint keys, KEY *key_info); +#endif === modified file 'sql/event_db_repository.cc' --- a/sql/event_db_repository.cc 2010-03-15 11:51:23 +0000 +++ b/sql/event_db_repository.cc 2010-04-08 12:10:05 +0000 @@ -105,7 +105,8 @@ const TABLE_FIELD_TYPE event_table_field { { C_STRING_WITH_LEN("sql_mode") }, { C_STRING_WITH_LEN("set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES'," - "'IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION'," + "'IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY'," + "'NO_UNSIGNED_SUBTRACTION'," "'NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB'," "'NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40'," "'ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES'," === modified file 'sql/events.cc' --- a/sql/events.cc 2010-03-04 08:03:07 +0000 +++ b/sql/events.cc 2010-04-06 22:47:08 +0000 @@ -1263,8 +1263,9 @@ Events::load_events_from_db(THD *thd) } } } - sql_print_information("Event Scheduler: Loaded %d event%s", - count, (count == 1) ? "" : "s"); + if (global_system_variables.log_warnings) + sql_print_information("Event Scheduler: Loaded %d event%s", + count, (count == 1) ? "" : "s"); ret= FALSE; end: === modified file 'sql/field.cc' --- a/sql/field.cc 2010-03-17 02:32:31 +0000 +++ b/sql/field.cc 2010-04-08 12:10:05 +0000 @@ -1306,14 +1306,13 @@ String *Field::val_int_as_str(String *va Field::Field(uchar *ptr_arg,uint32 length_arg,uchar *null_ptr_arg, uchar null_bit_arg, utype unireg_check_arg, const char *field_name_arg) - :ptr(ptr_arg), null_ptr(null_ptr_arg), - table(0), orig_table(0), table_name(0), - field_name(field_name_arg), - key_start(0), part_of_key(0), part_of_key_not_clustered(0), - part_of_sortkey(0), unireg_check(unireg_check_arg), - field_length(length_arg), null_bit(null_bit_arg), - is_created_from_null_item(FALSE), - vcol_info(0), stored_in_db(TRUE) + :ptr(ptr_arg), null_ptr(null_ptr_arg), table(0), orig_table(0), + table_name(0), field_name(field_name_arg), option_list(0), + option_struct(0), key_start(0), part_of_key(0), + part_of_key_not_clustered(0), part_of_sortkey(0), + unireg_check(unireg_check_arg), field_length(length_arg), + null_bit(null_bit_arg), is_created_from_null_item(FALSE), vcol_info(0), + stored_in_db(TRUE) { flags=null_ptr ? 0: NOT_NULL_FLAG; comment.str= (char*) ""; @@ -9567,7 +9566,8 @@ bool Create_field::init(THD *thd, char * Item *fld_on_update_value, LEX_STRING *fld_comment, char *fld_change, List<String> *fld_interval_list, CHARSET_INFO *fld_charset, uint fld_geom_type, - Virtual_column_info *fld_vcol_info) + Virtual_column_info *fld_vcol_info, + engine_option_value *create_opt) { uint sign_len, allowed_type_modifier= 0; ulong max_field_charlength= MAX_FIELD_CHARLENGTH; @@ -9578,6 +9578,7 @@ bool Create_field::init(THD *thd, char * field_name= fld_name; def= fld_default_value; flags= fld_type_modifier; + option_list= create_opt; unireg_check= (fld_type_modifier & AUTO_INCREMENT_FLAG ? Field::NEXT_NUMBER : Field::NONE); decimals= fld_decimals ? (uint)atoi(fld_decimals) : 0; @@ -10217,6 +10218,8 @@ Create_field::Create_field(Field *old_fi decimals= old_field->decimals(); vcol_info= old_field->vcol_info; stored_in_db= old_field->stored_in_db; + option_list= old_field->option_list; + option_struct= old_field->option_struct; /* Fix if the original table had 4 byte pointer blobs */ if (flags & BLOB_FLAG) @@ -10291,6 +10294,19 @@ Create_field::Create_field(Field *old_fi /** + Makes a clone of this object for ALTER/CREATE TABLE + + @param mem_root MEM_ROOT where to clone the field +*/ + +Create_field *Create_field::clone(MEM_ROOT *mem_root) const +{ + Create_field *res= new (mem_root) Create_field(*this); + return res; +} + + +/** maximum possible display length for blob. @return === modified file 'sql/field.h' --- a/sql/field.h 2010-03-15 11:51:23 +0000 +++ b/sql/field.h 2010-04-08 12:10:05 +0000 @@ -137,6 +137,9 @@ public: struct st_table *table; // Pointer for table struct st_table *orig_table; // Pointer to original table const char **table_name, *field_name; + /** reference to the list of options or NULL */ + engine_option_value *option_list; + void *option_struct; /* structure with parsed options */ LEX_STRING comment; /* Field is part of the following keys */ key_map key_start, part_of_key, part_of_key_not_clustered; @@ -2145,6 +2148,9 @@ public: CHARSET_INFO *charset; Field::geometry_type geom_type; Field *field; // For alter table + engine_option_value *option_list; + /** structure with parsed options (for comparing fields in ALTER TABLE) */ + void *option_struct; uint8 row,col,sc_length,interval_id; // For rea_create_table uint offset,pack_flag; @@ -2162,11 +2168,11 @@ public: */ bool stored_in_db; - Create_field() :after(0) {} + Create_field() :after(0), option_list(NULL), option_struct(NULL) + {} Create_field(Field *field, Field *orig_field); /* Used to make a clone of this object for ALTER/CREATE TABLE */ - Create_field *clone(MEM_ROOT *mem_root) const - { return new (mem_root) Create_field(*this); } + Create_field *clone(MEM_ROOT *mem_root) const; void create_length_to_internal_length(void); /* Init for a tmp table field. To be extended if need be. */ @@ -2178,8 +2184,8 @@ public: char *decimals, uint type_modifier, Item *default_value, Item *on_update_value, LEX_STRING *comment, char *change, List<String> *interval_list, CHARSET_INFO *cs, - uint uint_geom_type, - Virtual_column_info *vcol_info); + uint uint_geom_type, Virtual_column_info *vcol_info, + engine_option_value *option_list); bool field_flags_are_binary() { === modified file 'sql/ha_ndbcluster.cc' --- a/sql/ha_ndbcluster.cc 2010-03-29 15:13:53 +0000 +++ b/sql/ha_ndbcluster.cc 2010-04-01 14:34:51 +0000 @@ -10564,6 +10564,23 @@ mysql_declare_plugin(ndbcluster) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(ndbcluster) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &ndbcluster_storage_engine, + ndbcluster_hton_name, + "MySQL AB", + "Clustered, fault-tolerant tables", + PLUGIN_LICENSE_GPL, + ndbcluster_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + ndb_status_variables_export,/* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_GAMMA /* maturity */ +} +maria_declare_plugin_end; #else int Sun_ar_require_a_symbol_here= 0; === modified file 'sql/ha_partition.cc' --- a/sql/ha_partition.cc 2010-03-15 11:51:23 +0000 +++ b/sql/ha_partition.cc 2010-04-08 12:10:05 +0000 @@ -52,6 +52,7 @@ #endif #include "mysql_priv.h" +#include "create_options.h" #ifdef WITH_PARTITION_STORAGE_ENGINE #include "ha_partition.h" @@ -1218,7 +1219,9 @@ int ha_partition::prepare_new_partition( DBUG_ENTER("prepare_new_partition"); if ((error= set_up_table_before_create(tbl, part_name, create_info, - 0, p_elem))) + 0, p_elem)) || + parse_engine_table_options(ha_thd(), file->ht, + file->table_share)) goto error_create; if ((error= file->ha_create(part_name, tbl, create_info))) { @@ -1869,6 +1872,8 @@ uint ha_partition::del_ren_cre_table(con { if ((error= set_up_table_before_create(table_arg, from_buff, create_info, i, NULL)) || + parse_engine_table_options(ha_thd(), (*file)->ht, + (*file)->table_share) || ((error= (*file)->ha_create(from_buff, table_arg, create_info)))) goto create_error; } @@ -6715,5 +6720,22 @@ mysql_declare_plugin(partition) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(partition) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &partition_storage_engine, + "partition", + "Mikael Ronstrom, MySQL AB", + "Partition Storage Engine Helper", + PLUGIN_LICENSE_GPL, + partition_initialize, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100, /* 1.0 */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; #endif === modified file 'sql/handler.cc' --- a/sql/handler.cc 2010-03-31 08:17:03 +0000 +++ b/sql/handler.cc 2010-04-09 09:26:17 +0000 @@ -24,6 +24,7 @@ #endif #include "mysql_priv.h" +#include "create_options.h" #include "rpl_filter.h" #include <myisampack.h> #include "myisam.h" @@ -3719,7 +3720,11 @@ int ha_create_table(THD *thd, const char name= get_canonical_filename(table.file, share.path.str, name_buff); + if (parse_engine_table_options(thd, table.file->ht, &share)) + goto err; + error= table.file->ha_create(name, &table, create_info); + VOID(closefrm(&table, 0)); if (error) { @@ -3826,11 +3831,13 @@ int ha_init_key_cache(const char *name, uint tmp_block_size= (uint) key_cache->param_block_size; uint division_limit= key_cache->param_division_limit; uint age_threshold= key_cache->param_age_threshold; + uint partitions= key_cache->param_partitions; pthread_mutex_unlock(&LOCK_global_system_variables); DBUG_RETURN(!init_key_cache(key_cache, tmp_block_size, tmp_buff_size, - division_limit, age_threshold)); + division_limit, age_threshold, + partitions)); } DBUG_RETURN(0); } @@ -3860,10 +3867,12 @@ int ha_resize_key_cache(KEY_CACHE *key_c /** - Change parameters for key cache (like size) + Change parameters for key cache (like division_limit) */ int ha_change_key_cache_param(KEY_CACHE *key_cache) { + DBUG_ENTER("ha_change_key_cache_param"); + if (key_cache->key_cache_inited) { pthread_mutex_lock(&LOCK_global_system_variables); @@ -3872,9 +3881,35 @@ int ha_change_key_cache_param(KEY_CACHE pthread_mutex_unlock(&LOCK_global_system_variables); change_key_cache_param(key_cache, division_limit, age_threshold); } - return 0; + DBUG_RETURN(0); } + +/** + Repartition key cache +*/ +int ha_repartition_key_cache(KEY_CACHE *key_cache) +{ + DBUG_ENTER("ha_repartition_key_cache"); + + if (key_cache->key_cache_inited) + { + pthread_mutex_lock(&LOCK_global_system_variables); + size_t tmp_buff_size= (size_t) key_cache->param_buff_size; + long tmp_block_size= (long) key_cache->param_block_size; + uint division_limit= key_cache->param_division_limit; + uint age_threshold= key_cache->param_age_threshold; + uint partitions= key_cache->param_partitions; + pthread_mutex_unlock(&LOCK_global_system_variables); + DBUG_RETURN(!repartition_key_cache(key_cache, tmp_block_size, + tmp_buff_size, + division_limit, age_threshold, + partitions)); + } + DBUG_RETURN(0); +} + + /** Free memory allocated by a key cache. */ @@ -4420,6 +4455,8 @@ int handler::index_read_idx_map(uchar * enum ha_rkey_function find_flag) { int error, error1; + LINT_INIT(error1); + error= index_init(index, 0); if (!error) { === modified file 'sql/handler.h' --- a/sql/handler.h 2010-03-25 13:33:39 +0000 +++ b/sql/handler.h 2010-04-08 12:10:05 +0000 @@ -16,6 +16,9 @@ /* Definitions for parameters to do with handler-routines */ +#ifndef SQL_HANDLER_INCLUDED +#define SQL_HANDLER_INCLUDED + #ifdef USE_PRAGMA_INTERFACE #pragma interface /* gcc class implementation */ #endif @@ -549,6 +552,103 @@ struct handler_log_file_data { enum log_status status; }; +/* + Definitions for engine-specific table/field/index options in the CREATE TABLE. + + Options are declared with HA_*OPTION_* macros (HA_TOPTION_ULL, HA_FOPTION_ENUM, + HA_KOPTION_STRING, etc). + + Every macros takes the option name, and the name of the underlying field of + the appropriate C structure. The "appropriate C structure" is + ha_table_option_struct for table level options, + ha_field_option_struct for field level options, + ha_key_option_struct for key level options. The engine either + defines a structure of this name, or uses #define's to map + these "appropriate" names to the actual structure type name. + + ULL options use a ulonglong as the backing store. + HA_*OPTION_ULL() takes the option name, the structure field name, + the default value for the option, min, max, and blk_siz values. + + STRING options use a char* as a backing store. + HA_*OPTION_STRING takes the option name and the structure field name. + The default value will be 0. + + ENUM options use a uint as a backing store (not enum!!!). + HA_*OPTION_ENUM takes the option name, the structure field name, + the default value for the option as a number, and a string with the + permitted values for this enum - one string with comma separated values, + for example: "gzip,bzip2,lzma" + + BOOL options use a bool as a backing store. + HA_*OPTION_BOOL takes the option name, the structure field name, + and the default value for the option. + From the SQL, BOOL options accept YES/NO, ON/OFF, and 1/0. + + The name of the option is limited to 255 bytes, + the value (for string options) - to the 32767 bytes. + + See ha_example.cc for an example. +*/ +enum ha_option_type { HA_OPTION_TYPE_ULL, /* unsigned long long */ + HA_OPTION_TYPE_STRING, /* char * */ + HA_OPTION_TYPE_ENUM, /* uint */ + HA_OPTION_TYPE_BOOL}; /* bool */ + +#define HA_xOPTION_ULL(name, struc, field, def, min, max, blk_siz) \ + { HA_OPTION_TYPE_ULL, name, sizeof(name)-1, \ + offsetof(struc, field), def, min, max, blk_siz, 0 } +#define HA_xOPTION_STRING(name, struc, field) \ + { HA_OPTION_TYPE_STRING, name, sizeof(name)-1, \ + offsetof(struc, field), 0, 0, 0, 0, 0 } +#define HA_xOPTION_ENUM(name, struc, field, values, def) \ + { HA_OPTION_TYPE_ENUM, name, sizeof(name)-1, \ + offsetof(struc, field), def, 0, \ + sizeof(values)-1, 0, values } +#define HA_xOPTION_BOOL(name, struc, field, def) \ + { HA_OPTION_TYPE_BOOL, name, sizeof(name)-1, \ + offsetof(struc, field), def, 0, 1, 0, 0 } +#define HA_xOPTION_END { HA_OPTION_TYPE_ULL, 0, 0, 0, 0, 0, 0, 0, 0 } + +#define HA_TOPTION_ULL(name, field, def, min, max, blk_siz) \ + HA_xOPTION_ULL(name, ha_table_option_struct, field, def, min, max, blk_siz) +#define HA_TOPTION_STRING(name, field) \ + HA_xOPTION_STRING(name, ha_table_option_struct, field) +#define HA_TOPTION_ENUM(name, field, values, def) \ + HA_xOPTION_ENUM(name, ha_table_option_struct, field, values, def) +#define HA_TOPTION_BOOL(name, field, def) \ + HA_xOPTION_BOOL(name, ha_table_option_struct, field, def) +#define HA_TOPTION_END HA_xOPTION_END + +#define HA_FOPTION_ULL(name, field, def, min, max, blk_siz) \ + HA_xOPTION_ULL(name, ha_field_option_struct, field, def, min, max, blk_siz) +#define HA_FOPTION_STRING(name, field) \ + HA_xOPTION_STRING(name, ha_field_option_struct, field) +#define HA_FOPTION_ENUM(name, field, values, def) \ + HA_xOPTION_ENUM(name, ha_field_option_struct, field, values, def) +#define HA_FOPTION_BOOL(name, field, def) \ + HA_xOPTION_BOOL(name, ha_field_option_struct, field, def) +#define HA_FOPTION_END HA_xOPTION_END + +#define HA_KOPTION_ULL(name, field, def, min, max, blk_siz) \ + HA_xOPTION_ULL(name, ha_key_option_struct, field, def, min, max, blk_siz) +#define HA_KOPTION_STRING(name, field) \ + HA_xOPTION_STRING(name, ha_key_option_struct, field) +#define HA_KOPTION_ENUM(name, field, values, def) \ + HA_xOPTION_ENUM(name, ha_key_option_struct, field, values, def) +#define HA_KOPTION_BOOL(name, field, values, def) \ + HA_xOPTION_BOOL(name, ha_key_option_struct, field, values, def) +#define HA_KOPTION_END HA_xOPTION_END + +typedef struct st_ha_create_table_option { + enum ha_option_type type; + const char *name; + size_t name_length; + ptrdiff_t offset; + ulonglong def_value; + ulonglong min_value, max_value, block_size; + const char *values; +} ha_create_table_option; enum handler_iterator_type { @@ -721,7 +821,13 @@ struct handlerton int (*table_exists_in_engine)(handlerton *hton, THD* thd, const char *db, const char *name); uint32 license; /* Flag for Engine License */ - void *data; /* Location for engines to keep personal structures */ + /* + Optional clauses in the CREATE/ALTER TABLE + */ + ha_create_table_option *table_options; // table level options + ha_create_table_option *field_options; // these are specified per field + ha_create_table_option *index_options; // these are specified per index + }; @@ -945,11 +1051,16 @@ typedef struct st_ha_create_information uint extra_size; /* length of extra data segment */ /** Transactional or not. Unused; reserved for future versions. */ enum ha_choice transactional; - bool table_existed; /* 1 in create if table existed */ - bool frm_only; /* 1 if no ha_create_table() */ - bool varchar; /* 1 if table has a VARCHAR */ - enum ha_storage_media storage_media; /* DEFAULT, DISK or MEMORY */ - enum ha_choice page_checksum; /* If we have page_checksums */ + bool table_existed; ///< 1 in create if table existed + bool frm_only; ///< 1 if no ha_create_table() + bool varchar; ///< 1 if table has a VARCHAR + enum ha_storage_media storage_media; ///< DEFAULT, DISK or MEMORY + enum ha_choice page_checksum; ///< If we have page_checksums + engine_option_value *option_list; ///< list of table create options + /* the following three are only for ALTER TABLE, check_if_incompatible_data() */ + void *option_struct; ///< structure with parsed table options + void **fileds_option_struct; ///< array of field option structures + void **keys_option_struct; ///< array of key option structures } HA_CREATE_INFO; @@ -2201,6 +2312,7 @@ int ha_table_exists_in_engine(THD* thd, extern "C" int ha_init_key_cache(const char *name, KEY_CACHE *key_cache); int ha_resize_key_cache(KEY_CACHE *key_cache); int ha_change_key_cache_param(KEY_CACHE *key_cache); +int ha_repartition_key_cache(KEY_CACHE *key_cache); int ha_change_key_cache(KEY_CACHE *old_key_cache, KEY_CACHE *new_key_cache); int ha_end_key_cache(KEY_CACHE *key_cache); @@ -2254,3 +2366,5 @@ int ha_binlog_end(THD *thd); #define ha_binlog_wait(a) do {} while (0) #define ha_binlog_end(a) do {} while (0) #endif + +#endif === modified file 'sql/log.cc' --- a/sql/log.cc 2010-03-31 08:17:03 +0000 +++ b/sql/log.cc 2010-04-09 09:26:17 +0000 @@ -1238,6 +1238,7 @@ void LOGGER::deactivate_log_handler(THD { my_bool *tmp_opt= 0; MYSQL_LOG *file_log; + LINT_INIT(file_log); switch (log_type) { case QUERY_LOG_SLOW: @@ -6036,3 +6037,20 @@ mysql_declare_plugin(binlog) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(binlog) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &binlog_storage_engine, + "binlog", + "MySQL AB", + "This is a pseudo storage engine to represent the binlog in a transaction", + PLUGIN_LICENSE_GPL, + binlog_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'sql/log_event.cc' --- a/sql/log_event.cc 2010-03-15 11:51:23 +0000 +++ b/sql/log_event.cc 2010-04-06 21:49:01 +0000 @@ -1716,10 +1716,11 @@ beg: case MYSQL_TYPE_DATETIME: { + /* these must be size_t, because it's what my_b_printf expects for %d */ size_t d, t; uint64 i64= uint8korr(ptr); /* YYYYMMDDhhmmss */ - d= i64 / 1000000; - t= i64 % 1000000; + d= (size_t)(i64 / 1000000); + t= (size_t)(i64 % 1000000); my_b_printf(file, "%04d-%02d-%02d %02d:%02d:%02d", d / 10000, (d % 10000) / 100, d % 100, t / 10000, (t % 10000) / 100, t % 100); === modified file 'sql/log_event.h' --- a/sql/log_event.h 2010-03-15 11:51:23 +0000 +++ b/sql/log_event.h 2010-04-08 12:10:05 +0000 @@ -1371,7 +1371,7 @@ protected: MODE_PIPES_AS_CONCAT==0x2 MODE_ANSI_QUOTES==0x4 MODE_IGNORE_SPACE==0x8 - MODE_NOT_USED==0x10 + MODE_IGNORE_BAD_TABLE_OPTIONS==0x10 MODE_ONLY_FULL_GROUP_BY==0x20 MODE_NO_UNSIGNED_SUBTRACTION==0x40 MODE_NO_DIR_IN_CREATE==0x80 === modified file 'sql/mysql_priv.h' --- a/sql/mysql_priv.h 2010-03-29 15:13:53 +0000 +++ b/sql/mysql_priv.h 2010-04-08 12:10:05 +0000 @@ -520,7 +520,7 @@ protected: #define MODE_PIPES_AS_CONCAT 2 #define MODE_ANSI_QUOTES 4 #define MODE_IGNORE_SPACE 8 -#define MODE_NOT_USED 16 +#define MODE_IGNORE_BAD_TABLE_OPTIONS 16 #define MODE_ONLY_FULL_GROUP_BY 32 #define MODE_NO_UNSIGNED_SUBTRACTION 64 #define MODE_NO_DIR_IN_CREATE 128 @@ -1508,7 +1508,8 @@ bool add_field_to_list(THD *thd, LEX_STR char *change, List<String> *interval_list, CHARSET_INFO *cs, uint uint_geom_type, - Virtual_column_info *vcol_info); + Virtual_column_info *vcol_info, + engine_option_value *create_options); Create_field * new_create_field(THD *thd, char *field_name, enum_field_types type, char *length, char *decimals, uint type_modifier, === modified file 'sql/mysqld.cc' --- a/sql/mysqld.cc 2010-03-31 08:17:03 +0000 +++ b/sql/mysqld.cc 2010-04-09 09:26:17 +0000 @@ -1,4 +1,5 @@ /* Copyright (C) 2000-2008 MySQL AB, 2008-2009 Sun Microsystems, Inc. + 2009-2010 Monty Program Ab This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -244,7 +245,7 @@ const char *show_comp_option_name[]= {"Y static const char *sql_mode_names[]= { "REAL_AS_FLOAT", "PIPES_AS_CONCAT", "ANSI_QUOTES", "IGNORE_SPACE", - "?", "ONLY_FULL_GROUP_BY", "NO_UNSIGNED_SUBTRACTION", + "IGNORE_BAD_TABLE_OPTIONS", "ONLY_FULL_GROUP_BY", "NO_UNSIGNED_SUBTRACTION", "NO_DIR_IN_CREATE", "POSTGRESQL", "ORACLE", "MSSQL", "DB2", "MAXDB", "NO_KEY_OPTIONS", "NO_TABLE_OPTIONS", "NO_FIELD_OPTIONS", "MYSQL323", "MYSQL40", "ANSI", @@ -264,7 +265,7 @@ static const unsigned int sql_mode_names /*PIPES_AS_CONCAT*/ 15, /*ANSI_QUOTES*/ 11, /*IGNORE_SPACE*/ 12, - /*?*/ 1, + /*IGNORE_BAD_TABLE_OPTIONS*/ 24, /*ONLY_FULL_GROUP_BY*/ 18, /*NO_UNSIGNED_SUBTRACTION*/ 23, /*NO_DIR_IN_CREATE*/ 16, @@ -383,7 +384,7 @@ static bool volatile select_thread_in_us static bool volatile ready_to_exit; static my_bool opt_debugging= 0, opt_external_locking= 0, opt_console= 0; static my_bool opt_short_log_format= 0; -static my_bool opt_ignore_wrong_options= 0; +static my_bool opt_ignore_wrong_options= 0, opt_expect_abort= 0; static uint kill_cached_threads, wake_thread; ulong thread_created; static ulong max_used_connections; @@ -2549,13 +2550,19 @@ extern "C" sig_handler handle_segfault(i curr_time= my_time(0); localtime_r(&curr_time, &tm); - fprintf(stderr,"\ -%02d%02d%02d %2d:%02d:%02d - mysqld got " SIGNAL_FMT " ;\n\ + fprintf(stderr, "%02d%02d%02d %2d:%02d:%02d ", + tm.tm_year % 100, tm.tm_mon+1, tm.tm_mday, + tm.tm_hour, tm.tm_min, tm.tm_sec); + if (opt_expect_abort && sig == SIGABRT) + { + fprintf(stderr,"[Note] mysqld did an expected abort\n"); + goto end; + } + + fprintf(stderr,"[ERROR] mysqld got " SIGNAL_FMT " ;\n\ This could be because you hit a bug. It is also possible that this binary\n\ or one of the libraries it was linked against is corrupt, improperly built,\n\ or misconfigured. This error can also be caused by malfunctioning hardware.\n", - tm.tm_year % 100, tm.tm_mon+1, tm.tm_mday, - tm.tm_hour, tm.tm_min, tm.tm_sec, sig); fprintf(stderr, "\ We will try our best to scrape up some info that will hopefully help diagnose\n\ @@ -2674,9 +2681,11 @@ bugs.\n"); } #endif +end: #ifndef __WIN__ - /* On Windows, do not terminate, but pass control to exception filter */ exit(1); +#else + /* On Windows, do not terminate, but pass control to the exception filter */; #endif } @@ -5831,6 +5840,7 @@ enum options_mysqld OPT_INTERACTIVE_TIMEOUT, OPT_JOIN_BUFF_SIZE, OPT_KEY_BUFFER_SIZE, OPT_KEY_CACHE_BLOCK_SIZE, OPT_KEY_CACHE_DIVISION_LIMIT, OPT_KEY_CACHE_AGE_THRESHOLD, + OPT_KEY_CACHE_PARTITIONS, OPT_LONG_QUERY_TIME, OPT_LOWER_CASE_TABLE_NAMES, OPT_MAX_ALLOWED_PACKET, OPT_MAX_BINLOG_CACHE_SIZE, OPT_MAX_BINLOG_SIZE, @@ -5924,7 +5934,7 @@ enum options_mysqld OPT_MIN_EXAMINED_ROW_LIMIT, OPT_LOG_SLOW_SLAVE_STATEMENTS, OPT_DEBUG_CRC, OPT_DEBUG_ON, OPT_OLD_MODE, - OPT_TEST_IGNORE_WRONG_OPTIONS, + OPT_TEST_IGNORE_WRONG_OPTIONS, OPT_TEST_RESTART, #if defined(ENABLED_DEBUG_SYNC) OPT_DEBUG_SYNC_TIMEOUT, #endif /* defined(ENABLED_DEBUG_SYNC) */ @@ -6645,8 +6655,6 @@ Can't be set to 1 if --log-slave-updates {"shared-memory", OPT_ENABLE_SHARED_MEMORY, "Enable the shared memory.",(uchar**) &opt_enable_shared_memory, (uchar**) &opt_enable_shared_memory, 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0}, -#endif -#ifdef HAVE_SMEM {"shared-memory-base-name",OPT_SHARED_MEMORY_BASE_NAME, "Base name of shared memory.", (uchar**) &shared_memory_base_name, (uchar**) &shared_memory_base_name, 0, GET_STR, REQUIRED_ARG, 0, 0, 0, 0, 0, 0}, @@ -6774,6 +6782,10 @@ log and this option does nothing anymore "Ignore wrong enums values in command line arguments. Useful only for test scripts", (uchar**) &opt_ignore_wrong_options, (uchar**) &opt_ignore_wrong_options, 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0}, + {"test-expect-abort", OPT_TEST_RESTART, + "Expect that server aborts with 'abort'; Don't write out server variables on 'abort'. Useful only for test scripts", + (uchar**) &opt_expect_abort, (uchar**) &opt_expect_abort, + 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0}, {"timed_mutexes", OPT_TIMED_MUTEXES, "Specify whether to time mutexes (only InnoDB mutexes are currently supported)", (uchar**) &timed_mutexes, (uchar**) &timed_mutexes, 0, GET_BOOL, NO_ARG, 0, @@ -6930,6 +6942,12 @@ log and this option does nothing anymore (uchar**) 0, 0, (GET_ULONG | GET_ASK_ADDR) , REQUIRED_ARG, 100, 1, 100, 0, 1, 0}, + {"key_cache_partitions", OPT_KEY_CACHE_PARTITIONS, + "The number of partitions in key cache", + (uchar**) &dflt_key_cache_var.param_partitions, + (uchar**) 0, + 0, (GET_ULONG | GET_ASK_ADDR), REQUIRED_ARG, DEFAULT_KEY_CACHE_PARTITIONS, + 0, MAX_KEY_CACHE_PARTITIONS, 0, 1, 0}, {"log-slow-filter", OPT_LOG_SLOW_FILTER, "Log only the queries that followed certain execution plan. Multiple flags allowed in a comma-separated string. [admin, filesort, filesort_on_disk, full_join, full_scan, query_cache, query_cache_miss, tmp_table, tmp_table_on_disk]. Sets log-slow-admin-command to ON", 0, 0, 0, GET_STR, REQUIRED_ARG, 0, 0, 0, QPLAN_ALWAYS_SET, 0, 0}, @@ -8912,6 +8930,7 @@ mysql_getopt_value(const char *keyname, case OPT_KEY_CACHE_BLOCK_SIZE: case OPT_KEY_CACHE_DIVISION_LIMIT: case OPT_KEY_CACHE_AGE_THRESHOLD: + case OPT_KEY_CACHE_PARTITIONS: { KEY_CACHE *key_cache; if (!(key_cache= get_or_create_key_cache(keyname, key_length))) @@ -8929,6 +8948,8 @@ mysql_getopt_value(const char *keyname, return (uchar**) &key_cache->param_division_limit; case OPT_KEY_CACHE_AGE_THRESHOLD: return (uchar**) &key_cache->param_age_threshold; + case OPT_KEY_CACHE_PARTITIONS: + return (uchar**) &key_cache->param_partitions; } } } === modified file 'sql/set_var.cc' --- a/sql/set_var.cc 2010-03-31 08:17:03 +0000 +++ b/sql/set_var.cc 2010-04-09 09:26:17 +0000 @@ -320,15 +320,18 @@ static sys_var_thd_ulong sys_interactive static sys_var_thd_ulong sys_join_buffer_size(&vars, "join_buffer_size", &SV::join_buff_size); static sys_var_key_buffer_size sys_key_buffer_size(&vars, "key_buffer_size"); -static sys_var_key_cache_long sys_key_cache_block_size(&vars, "key_cache_block_size", - offsetof(KEY_CACHE, - param_block_size)); -static sys_var_key_cache_long sys_key_cache_division_limit(&vars, "key_cache_division_limit", - offsetof(KEY_CACHE, - param_division_limit)); -static sys_var_key_cache_long sys_key_cache_age_threshold(&vars, "key_cache_age_threshold", - offsetof(KEY_CACHE, - param_age_threshold)); +static sys_var_key_cache_long sys_key_cache_block_size(&vars, + "key_cache_block_size", + offsetof(KEY_CACHE,param_block_size)); +static sys_var_key_cache_long sys_key_cache_division_limit(&vars, + "key_cache_division_limit", + offsetof(KEY_CACHE, param_division_limit)); +static sys_var_key_cache_long sys_key_cache_age_threshold(&vars, + "key_cache_age_threshold", + offsetof(KEY_CACHE, param_age_threshold)); +static sys_var_key_cache_long sys_key_cache_partitions(&vars, + "key_cache_partitions", + offsetof(KEY_CACHE, param_partitions)); static sys_var_const sys_language(&vars, "language", OPT_GLOBAL, SHOW_CHAR, (uchar*) language); @@ -2548,7 +2551,15 @@ bool sys_var_key_cache_long::update(THD pthread_mutex_unlock(&LOCK_global_system_variables); - error= (bool) (ha_resize_key_cache(key_cache)); + if (offset == offsetof(KEY_CACHE, param_block_size)) + error= (bool) (ha_resize_key_cache(key_cache)); + else + if (offset == offsetof(KEY_CACHE, param_division_limit) || + offset == offsetof(KEY_CACHE, param_age_threshold)) + error= (bool) (ha_change_key_cache_param(key_cache)); + else + if (offset == offsetof(KEY_CACHE, param_partitions)) + error= (bool) (ha_repartition_key_cache(key_cache)); pthread_mutex_lock(&LOCK_global_system_variables); key_cache->in_init= 0; @@ -2593,7 +2604,7 @@ static int sys_check_log_path(THD *thd, char path[FN_REFLEN], buff[FN_REFLEN]; MY_STAT f_stat; String str(buff, sizeof(buff), system_charset_info), *res; - const char *log_file_str; + const char *log_file_str= 0; size_t path_length; if (!(res= var->value->val_str(&str))) @@ -2643,7 +2654,7 @@ static int sys_check_log_path(THD *thd, err: my_error(ER_WRONG_VALUE_FOR_VAR, MYF(0), var->var->name, - res ? log_file_str : "NULL"); + log_file_str ? log_file_str : "NULL"); return 1; } @@ -2652,7 +2663,7 @@ bool update_sys_var_str_path(THD *thd, s set_var *var, const char *log_ext, bool log_state, uint log_type) { - MYSQL_QUERY_LOG *file_log; + MYSQL_QUERY_LOG *file_log= 0; char buff[FN_REFLEN]; char *res= 0, *old_value=(char *)(var ? var->value->str_value.ptr() : 0); bool result= 0; @@ -4156,6 +4167,7 @@ static KEY_CACHE *create_key_cache(const key_cache->param_block_size= dflt_key_cache_var.param_block_size; key_cache->param_division_limit= dflt_key_cache_var.param_division_limit; key_cache->param_age_threshold= dflt_key_cache_var.param_age_threshold; + key_cache->param_partitions= dflt_key_cache_var.param_partitions; } } DBUG_RETURN(key_cache); === modified file 'sql/set_var.h' --- a/sql/set_var.h 2009-12-03 11:19:05 +0000 +++ b/sql/set_var.h 2010-02-16 16:41:11 +0000 @@ -1427,6 +1427,7 @@ public: my_free((uchar*) name, MYF(0)); } friend bool process_key_caches(process_key_cache_t func); + friend int fill_key_cache_tables(THD *thd, TABLE_LIST *tables, COND *cond); friend void delete_elements(I_List<NAMED_LIST> *list, void (*free_element)(const char*, uchar*)); }; === modified file 'sql/share/errmsg.txt' --- a/sql/share/errmsg.txt 2010-03-15 11:51:23 +0000 +++ b/sql/share/errmsg.txt 2010-04-08 12:10:05 +0000 @@ -6240,3 +6240,8 @@ ER_DEBUG_SYNC_TIMEOUT ER_DEBUG_SYNC_HIT_LIMIT eng "debug sync point hit limit reached" ger "Debug Sync Point Hit Limit erreicht" + +ER_UNKNOWN_OPTION + eng "Unknown option '%-.64s'" +ER_BAD_OPTION_VALUE + eng "Incorrect value '%-.64s' for option '%-.64s'" === modified file 'sql/slave.cc' --- a/sql/slave.cc 2010-03-04 08:03:07 +0000 +++ b/sql/slave.cc 2010-04-06 22:47:08 +0000 @@ -2255,6 +2255,7 @@ static int exec_relay_log_event(THD* thd if (slave_trans_retries) { int temp_err; + LINT_INIT(temp_err); if (exec_res && (temp_err= has_temporary_error(thd))) { const char *errmsg; === modified file 'sql/sp.cc' --- a/sql/sp.cc 2010-03-15 11:51:23 +0000 +++ b/sql/sp.cc 2010-04-08 12:10:05 +0000 @@ -147,7 +147,8 @@ TABLE_FIELD_TYPE proc_table_fields[MYSQL { { C_STRING_WITH_LEN("sql_mode") }, { C_STRING_WITH_LEN("set('REAL_AS_FLOAT','PIPES_AS_CONCAT','ANSI_QUOTES'," - "'IGNORE_SPACE','NOT_USED','ONLY_FULL_GROUP_BY','NO_UNSIGNED_SUBTRACTION'," + "'IGNORE_SPACE','IGNORE_BAD_TABLE_OPTIONS','ONLY_FULL_GROUP_BY'," + "'NO_UNSIGNED_SUBTRACTION'," "'NO_DIR_IN_CREATE','POSTGRESQL','ORACLE','MSSQL','DB2','MAXDB'," "'NO_KEY_OPTIONS','NO_TABLE_OPTIONS','NO_FIELD_OPTIONS','MYSQL323','MYSQL40'," "'ANSI','NO_AUTO_VALUE_ON_ZERO','NO_BACKSLASH_ESCAPES','STRICT_TRANS_TABLES'," === modified file 'sql/sp_head.cc' --- a/sql/sp_head.cc 2010-03-15 11:51:23 +0000 +++ b/sql/sp_head.cc 2010-04-08 12:10:05 +0000 @@ -2089,6 +2089,8 @@ sp_head::reset_lex(THD *thd) sublex->dec= NULL; sublex->interval_list.empty(); sublex->type= 0; + sublex->uint_geom_type= 0; + sublex->vcol_info= 0; DBUG_RETURN(FALSE); } @@ -2216,7 +2218,7 @@ sp_head::fill_field_definition(THD *thd, lex->charset ? lex->charset : thd->variables.collation_database, lex->uint_geom_type, - lex->vcol_info)) + lex->vcol_info, NULL)) return TRUE; if (field_def->interval_list.elements) === modified file 'sql/sql_acl.cc' --- a/sql/sql_acl.cc 2010-03-29 15:13:53 +0000 +++ b/sql/sql_acl.cc 2010-04-06 11:04:56 +0000 @@ -161,7 +161,10 @@ static LEX_STRING old_password_plugin_na /// @todo make it configurable LEX_STRING *default_auth_plugin_name= &native_password_plugin_name; -static plugin_ref native_password_plugin, old_password_plugin; +static plugin_ref native_password_plugin; +#ifndef EMBEDDED_LIBRARY +static plugin_ref old_password_plugin; +#endif /* Classes */ @@ -284,9 +287,8 @@ static void init_check_host(void); static void rebuild_check_host(void); static ACL_USER *find_acl_user(const char *host, const char *user, my_bool exact); -static bool update_user_table(THD *thd, TABLE *table, - const char *host, const char *user, - const char *new_password, uint new_password_len); +static bool update_user_table(THD *, TABLE *, const char *, const char *, + const char *, uint); static void update_hostname(acl_host_and_ip *host, const char *hostname); static bool compare_hostname(const acl_host_and_ip *host,const char *hostname, const char *ip); @@ -316,6 +318,35 @@ set_user_salt(ACL_USER *acl_user, const acl_user->salt_len= 0; } +/** + Fix ACL::plugin pointer to point to a hard-coded string, if appropriate + + Make sure that if ACL_USER's plugin is a built-in, then it points + to a hard coded string, not to an allocated copy. Run-time, for + authentication, we want to be able to detect built-ins by comparing + pointers, not strings. + + Additionally - update the salt if the plugin is built-in. + + @retval 0 the pointers were fixed + @retval 1 this ACL_USER uses a not built-in plugin +*/ +static bool fix_user_plugin_ptr(ACL_USER *user) +{ + if (my_strcasecmp(system_charset_info, user->plugin.str, + native_password_plugin_name.str) == 0) + user->plugin= native_password_plugin_name; + else + if (my_strcasecmp(system_charset_info, user->plugin.str, + old_password_plugin_name.str) == 0) + user->plugin= old_password_plugin_name; + else + return true; + + set_user_salt(user, user->auth_string.str, user->auth_string.length); + return false; +} + /* This after_update function is used when user.password is less than SCRAMBLE_LENGTH bytes. @@ -533,6 +564,9 @@ static my_bool acl_load(THD *thd, TABLE_ while (!(read_record_info.read_record(&read_record_info))) { ACL_USER user; + char *password; + uint password_len; + bzero(&user, sizeof(user)); update_hostname(&user.host, get_field(&mem, table->field[0])); user.user= get_field(&mem, table->field[1]); @@ -545,8 +579,8 @@ static my_bool acl_load(THD *thd, TABLE_ continue; } - char *password= get_field(thd->mem_root, table->field[2]); - uint password_len= password ? strlen(password) : 0; + password= get_field(&mem, table->field[2]); + password_len= password ? strlen(password) : 0; user.auth_string.str= password ? password : const_cast<char*>(""); user.auth_string.length= password_len; set_user_salt(&user, password, password_len); @@ -656,6 +690,8 @@ static my_bool acl_load(THD *thd, TABLE_ char *tmpstr= get_field(&mem, table->field[next_field++]); if (tmpstr) { + user.plugin.str= tmpstr; + user.plugin.length= strlen(user.plugin.str); if (user.auth_string.length) { sql_print_warning("'user' entry '%s@%s' has both a password " @@ -664,22 +700,12 @@ static my_bool acl_load(THD *thd, TABLE_ user.user ? user.user : "", user.host.hostname ? user.host.hostname : ""); } - if (my_strcasecmp(system_charset_info, tmpstr, - native_password_plugin_name.str) == 0) - user.plugin= native_password_plugin_name; - else - if (my_strcasecmp(system_charset_info, tmpstr, - old_password_plugin_name.str) == 0) - user.plugin= old_password_plugin_name; - else - { - user.plugin.str= tmpstr; - user.plugin.length= strlen(tmpstr); - } user.auth_string.str= get_field(&mem, table->field[next_field++]); if (!user.auth_string.str) user.auth_string.str= const_cast<char*>(""); user.auth_string.length= strlen(user.auth_string.str); + + fix_user_plugin_ptr(&user); } } } @@ -1126,12 +1152,20 @@ static void acl_update_user(const char * { if (plugin->str[0]) { - acl_user->plugin.str= strmake_root(&mem, plugin->str, plugin->length); - acl_user->plugin.length= plugin->length; + acl_user->plugin= *plugin; acl_user->auth_string.str= auth->str ? strmake_root(&mem, auth->str, auth->length) : const_cast<char*>(""); acl_user->auth_string.length= auth->length; + if (fix_user_plugin_ptr(acl_user)) + acl_user->plugin.str= strmake_root(&mem, plugin->str, plugin->length); } + else + if (password) + { + acl_user->auth_string.str= strmake_root(&mem, password, password_len); + acl_user->auth_string.length= password_len; + set_user_salt(acl_user, password, password_len); + } acl_user->access=privileges; if (mqh->specified_limits & USER_RESOURCES::QUERIES_PER_HOUR) acl_user->user_resource.questions=mqh->questions; @@ -1151,8 +1185,6 @@ static void acl_update_user(const char * acl_user->x509_subject= (x509_subject ? strdup_root(&mem,x509_subject) : 0); } - if (password) - set_user_salt(acl_user, password, password_len); /* search complete: */ break; } @@ -1180,11 +1212,12 @@ static void acl_insert_user(const char * update_hostname(&acl_user.host, *host ? strdup_root(&mem, host): 0); if (plugin->str[0]) { - acl_user.plugin.str= strmake_root(&mem, plugin->str, plugin->length); - acl_user.plugin.length= plugin->length; + acl_user.plugin= *plugin; acl_user.auth_string.str= auth->str ? strmake_root(&mem, auth->str, auth->length) : const_cast<char*>(""); acl_user.auth_string.length= auth->length; + if (fix_user_plugin_ptr(&acl_user)) + acl_user.plugin.str= strmake_root(&mem, plugin->str, plugin->length); } else { @@ -1192,6 +1225,7 @@ static void acl_insert_user(const char * old_password_plugin_name : native_password_plugin_name; acl_user.auth_string.str= strmake_root(&mem, password, password_len); acl_user.auth_string.length= password_len; + set_user_salt(&acl_user, password, password_len); } acl_user.access=privileges; @@ -1204,8 +1238,6 @@ static void acl_insert_user(const char * acl_user.x509_issuer= x509_issuer ? strdup_root(&mem,x509_issuer) : 0; acl_user.x509_subject=x509_subject ? strdup_root(&mem,x509_subject) : 0; - set_user_salt(&acl_user, password, password_len); - VOID(push_dynamic(&acl_users,(uchar*) &acl_user)); if (!acl_user.host.hostname || (acl_user.host.hostname[0] == wild_many && !acl_user.host.hostname[1])) @@ -1583,7 +1615,13 @@ bool change_password(THD *thd, const cha goto end; } /* update loaded acl entry: */ - set_user_salt(acl_user, new_password, new_password_len); + if (acl_user->plugin.str == native_password_plugin_name.str || + acl_user->plugin.str == old_password_plugin_name.str) + { + acl_user->auth_string.str= strmake_root(&mem, new_password, new_password_len); + acl_user->auth_string.length= new_password_len; + set_user_salt(acl_user, new_password, new_password_len); + } if (update_user_table(thd, table, acl_user->host.hostname ? acl_user->host.hostname : "", @@ -4614,16 +4652,27 @@ bool mysql_show_grants(THD *thd,LEX_USER global.append(lex_user->host.str,lex_user->host.length, system_charset_info); global.append ('\''); - if (acl_user->salt_len) + if (acl_user->plugin.str == native_password_plugin_name.str || + acl_user->plugin.str == old_password_plugin_name.str) { - char passwd_buff[SCRAMBLED_PASSWORD_CHAR_LENGTH+1]; - if (acl_user->salt_len == SCRAMBLE_LENGTH) - make_password_from_salt(passwd_buff, acl_user->salt); - else - make_password_from_salt_323(passwd_buff, (ulong *) acl_user->salt); - global.append(STRING_WITH_LEN(" IDENTIFIED BY PASSWORD '")); - global.append(passwd_buff); - global.append('\''); + if (acl_user->auth_string.length) + { + DBUG_ASSERT(acl_user->salt_len); + global.append(STRING_WITH_LEN(" IDENTIFIED BY PASSWORD '")); + global.append(acl_user->auth_string.str, acl_user->auth_string.length); + global.append('\''); + } + } + else + { + global.append(STRING_WITH_LEN(" IDENTIFIED VIA ")); + global.append(acl_user->plugin.str, acl_user->plugin.length); + if (acl_user->auth_string.length) + { + global.append(STRING_WITH_LEN(" USING '")); + global.append(acl_user->auth_string.str, acl_user->auth_string.length); + global.append('\''); + } } /* "show grants" SSL related stuff */ if (acl_user->ssl_type == SSL_TYPE_ANY) @@ -6843,6 +6892,7 @@ bool check_routine_level_acl(THD *thd, c #define initialized 0 #define decrease_user_connections(X) /* nothing */ #define check_for_max_user_connections(X,Y) 0 +#define get_or_create_user_conn(A,B,C,D) 0 #endif #endif #ifndef HAVE_OPENSSL @@ -7823,15 +7873,15 @@ static int do_auth_once(THD *thd, LEX_ST bool unlock_plugin= false; plugin_ref plugin; +#ifdef EMBEDDED_LIBRARY + plugin= native_password_plugin; +#else if (auth_plugin_name->str == native_password_plugin_name.str) plugin= native_password_plugin; - else -#ifndef EMBEDDED_LIBRARY - if (auth_plugin_name->str == old_password_plugin_name.str) + else if (auth_plugin_name->str == old_password_plugin_name.str) plugin= old_password_plugin; - else - if ((plugin= my_plugin_lock_by_name(thd, auth_plugin_name, - MYSQL_AUTHENTICATION_PLUGIN))) + else if ((plugin= my_plugin_lock_by_name(thd, auth_plugin_name, + MYSQL_AUTHENTICATION_PLUGIN))) unlock_plugin= true; #endif @@ -8254,3 +8304,35 @@ mysql_declare_plugin(mysql_password) } mysql_declare_plugin_end; +maria_declare_plugin(mysql_password) +{ + MYSQL_AUTHENTICATION_PLUGIN, /* type constant */ + &native_password_handler, /* type descriptor */ + native_password_plugin_name.str, /* Name */ + "R.J.Silk, Sergei Golubchik", /* Author */ + "Native MySQL authentication", /* Description */ + PLUGIN_LICENSE_GPL, /* License */ + NULL, /* Init function */ + NULL, /* Deinit function */ + 0x0100, /* Version (1.0) */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* String version */ + MariaDB_PLUGIN_MATURITY_BETA /* Maturity */ +}, +{ + MYSQL_AUTHENTICATION_PLUGIN, /* type constant */ + &old_password_handler, /* type descriptor */ + old_password_plugin_name.str, /* Name */ + "R.J.Silk, Sergei Golubchik", /* Author */ + "Old MySQL-4.0 authentication", /* Description */ + PLUGIN_LICENSE_GPL, /* License */ + NULL, /* Init function */ + NULL, /* Deinit function */ + 0x0100, /* Version (1.0) */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* String version */ + MariaDB_PLUGIN_MATURITY_BETA /* Maturity */ +} +maria_declare_plugin_end; === modified file 'sql/sql_builtin.cc.in' --- a/sql/sql_builtin.cc.in 2010-03-29 15:13:53 +0000 +++ b/sql/sql_builtin.cc.in 2010-04-01 14:34:51 +0000 @@ -16,13 +16,12 @@ #include <my_global.h> #include <mysql/plugin.h> -typedef struct st_mysql_plugin builtin_plugin[]; +typedef struct st_maria_plugin builtin_maria_plugin[]; -extern builtin_plugin - builtin_binlog_plugin, builtin_mysql_password_plugin@mysql_plugin_defs@; +extern builtin_maria_plugin + builtin_maria_binlog_plugin, builtin_maria_mysql_password_plugin@maria_plugin_defs@; -struct st_mysql_plugin *mysqld_builtins[]= +struct st_maria_plugin *mariadb_builtins[]= { - builtin_binlog_plugin, builtin_mysql_password_plugin@mysql_plugin_defs@,(struct st_mysql_plugin *)0 + builtin_maria_binlog_plugin, builtin_maria_mysql_password_plugin@maria_plugin_defs@,(struct st_maria_plugin *)0 }; - === modified file 'sql/sql_class.cc' --- a/sql/sql_class.cc 2010-03-29 15:13:53 +0000 +++ b/sql/sql_class.cc 2010-04-08 12:10:05 +0000 @@ -106,6 +106,7 @@ Key::Key(const Key &rhs, MEM_ROOT *mem_r key_create_info(rhs.key_create_info), columns(rhs.columns, mem_root), name(rhs.name), + option_list(rhs.option_list), generated(rhs.generated) { list_copy_and_replace_each_value(columns, mem_root); @@ -775,6 +776,7 @@ THD::THD() void THD::push_internal_handler(Internal_error_handler *handler) { + DBUG_ENTER("THD::push_internal_handler"); if (m_internal_handler) { handler->m_prev_internal_handler= m_internal_handler; @@ -784,6 +786,7 @@ void THD::push_internal_handler(Internal { m_internal_handler= handler; } + DBUG_VOID_RETURN; } @@ -803,8 +806,10 @@ bool THD::handle_error(uint sql_errno, c void THD::pop_internal_handler() { + DBUG_ENTER("THD::pop_internal_handler"); DBUG_ASSERT(m_internal_handler != NULL); m_internal_handler= m_internal_handler->m_prev_internal_handler; + DBUG_VOID_RETURN; } extern "C" @@ -924,7 +929,7 @@ void THD::update_stats(void) /* A SQL query. */ if (lex->sql_command == SQLCOM_SELECT) select_commands++; - else if (! sql_command_flags[lex->sql_command] & CF_STATUS_COMMAND) + else if (sql_command_flags[lex->sql_command] & CF_STATUS_COMMAND) { /* Ignore 'SHOW ' commands */ } === modified file 'sql/sql_class.h' --- a/sql/sql_class.h 2010-03-31 08:17:03 +0000 +++ b/sql/sql_class.h 2010-04-09 09:26:17 +0000 @@ -205,13 +205,15 @@ public: KEY_CREATE_INFO key_create_info; List<Key_part_spec> columns; const char *name; + engine_option_value *option_list; bool generated; Key(enum Keytype type_par, const char *name_arg, KEY_CREATE_INFO *key_info_arg, - bool generated_arg, List<Key_part_spec> &cols) + bool generated_arg, List<Key_part_spec> &cols, + engine_option_value *create_opt) :type(type_par), key_create_info(*key_info_arg), columns(cols), - name(name_arg), generated(generated_arg) + name(name_arg), option_list(create_opt), generated(generated_arg) {} Key(const Key &rhs, MEM_ROOT *mem_root); virtual ~Key() {} @@ -240,7 +242,7 @@ public: Foreign_key(const char *name_arg, List<Key_part_spec> &cols, Table_ident *table, List<Key_part_spec> &ref_cols, uint delete_opt_arg, uint update_opt_arg, uint match_opt_arg) - :Key(FOREIGN_KEY, name_arg, &default_key_create_info, 0, cols), + :Key(FOREIGN_KEY, name_arg, &default_key_create_info, 0, cols, NULL), ref_table(table), ref_columns(ref_cols), delete_opt(delete_opt_arg), update_opt(update_opt_arg), match_opt(match_opt_arg) === modified file 'sql/sql_insert.cc' --- a/sql/sql_insert.cc 2010-03-29 15:13:53 +0000 +++ b/sql/sql_insert.cc 2010-04-08 12:10:05 +0000 @@ -72,15 +72,6 @@ static void unlink_blobs(register TABLE #endif static bool check_view_insertability(THD *thd, TABLE_LIST *view); -/* Define to force use of my_malloc() if the allocated memory block is big */ - -#ifndef HAVE_ALLOCA -#define my_safe_alloca(size, min_length) my_alloca(size) -#define my_safe_afree(ptr, size, min_length) my_afree(ptr) -#else -#define my_safe_alloca(size, min_length) ((size <= min_length) ? my_alloca(size) : my_malloc(size,MYF(0))) -#define my_safe_afree(ptr, size, min_length) if (size > min_length) my_free(ptr,MYF(0)) -#endif /* Check that insert/update fields are from the same single table of a view. === modified file 'sql/sql_lex.h' --- a/sql/sql_lex.h 2010-03-29 15:13:53 +0000 +++ b/sql/sql_lex.h 2010-04-08 12:10:05 +0000 @@ -1749,6 +1749,11 @@ typedef struct st_lex : public Query_tab const char *stmt_definition_end; /** + Collects create options for Field and KEY + */ + engine_option_value *option_list, *option_list_last; + + /** During name resolution search only in the table list given by Name_resolution_context::first_name_resolution_table and Name_resolution_context::last_name_resolution_table === modified file 'sql/sql_parse.cc' --- a/sql/sql_parse.cc 2010-03-29 15:13:53 +0000 +++ b/sql/sql_parse.cc 2010-04-08 12:10:05 +0000 @@ -6086,7 +6086,8 @@ bool add_field_to_list(THD *thd, LEX_STR char *change, List<String> *interval_list, CHARSET_INFO *cs, uint uint_geom_type, - Virtual_column_info *vcol_info) + Virtual_column_info *vcol_info, + engine_option_value *create_options) { register Create_field *new_field; LEX *lex= thd->lex; @@ -6104,7 +6105,7 @@ bool add_field_to_list(THD *thd, LEX_STR lex->col_list.push_back(new Key_part_spec(field_name->str, 0)); key= new Key(Key::PRIMARY, NullS, &default_key_create_info, - 0, lex->col_list); + 0, lex->col_list, NULL); lex->alter_info.key_list.push_back(key); lex->col_list.empty(); } @@ -6114,7 +6115,7 @@ bool add_field_to_list(THD *thd, LEX_STR lex->col_list.push_back(new Key_part_spec(field_name->str, 0)); key= new Key(Key::UNIQUE, NullS, &default_key_create_info, 0, - lex->col_list); + lex->col_list, NULL); lex->alter_info.key_list.push_back(key); lex->col_list.empty(); } @@ -6172,7 +6173,8 @@ bool add_field_to_list(THD *thd, LEX_STR if (!(new_field= new Create_field()) || new_field->init(thd, field_name->str, type, length, decimals, type_modifier, default_value, on_update_value, comment, change, - interval_list, cs, uint_geom_type, vcol_info)) + interval_list, cs, uint_geom_type, vcol_info, + create_options)) DBUG_RETURN(1); lex->alter_info.create_list.push_back(new_field); === modified file 'sql/sql_plugin.cc' --- a/sql/sql_plugin.cc 2010-03-29 15:13:53 +0000 +++ b/sql/sql_plugin.cc 2010-04-08 17:19:01 +0000 @@ -20,7 +20,7 @@ #define REPORT_TO_LOG 1 #define REPORT_TO_USER 2 -extern struct st_mysql_plugin *mysqld_builtins[]; +extern struct st_maria_plugin *mariadb_builtins[]; /** @note The order of the enumeration is critical. @@ -78,6 +78,14 @@ static const char *sizeof_st_plugin_sym= "_mysql_sizeof_struct_st_plugin_"; static const char *plugin_declarations_sym= "_mysql_plugin_declarations_"; static int min_plugin_interface_version= MYSQL_PLUGIN_INTERFACE_VERSION & ~0xFF; +static const char *maria_plugin_interface_version_sym= + "_maria_plugin_interface_version_"; +static const char *maria_sizeof_st_plugin_sym= + "_maria_sizeof_struct_st_plugin_"; +static const char *maria_plugin_declarations_sym= + "_maria_plugin_declarations_"; +static int min_maria_plugin_interface_version= + MARIA_PLUGIN_INTERFACE_VERSION & ~0xFF; #endif /* Note that 'int version' must be the first field of every plugin @@ -211,7 +219,7 @@ static bool plugin_load_list(MEM_ROOT *t const char *list); static int test_plugin_options(MEM_ROOT *, struct st_plugin_int *, int *, char **); -static bool register_builtin(struct st_mysql_plugin *, struct st_plugin_int *, +static bool register_builtin(struct st_maria_plugin *, struct st_plugin_int *, struct st_plugin_int **); static void unlock_variables(THD *thd, struct system_variables *vars); static void cleanup_variables(THD *thd, struct system_variables *vars); @@ -368,6 +376,225 @@ static inline void free_plugin_mem(struc } +/** + Reads data from mysql plugin interface + + @param plugin_dl Structure where the data should be put + @param sym Reverence on version info + @param dlpath Path to the module + @param report What errors should be reported + + @retval FALSE OK + @retval TRUE ERROR +*/ + +#ifdef HAVE_DLOPEN +static my_bool read_mysql_plugin_info(struct st_plugin_dl *plugin_dl, + void *sym, char *dlpath, + int report) +{ + DBUG_ENTER("read_maria_plugin_info"); + /* Determine interface version */ + if (!sym) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_FIND_DL_ENTRY, plugin_interface_version_sym); + DBUG_RETURN(TRUE); + } + plugin_dl->mariaversion= 0; + plugin_dl->mysqlversion= *(int *)sym; + /* Versioning */ + if (plugin_dl->mysqlversion < min_plugin_interface_version || + (plugin_dl->mysqlversion >> 8) > (MYSQL_PLUGIN_INTERFACE_VERSION >> 8)) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_OPEN_LIBRARY, dlpath, 0, + "plugin interface version mismatch"); + DBUG_RETURN(TRUE); + } + /* Find plugin declarations */ + if (!(sym= dlsym(plugin_dl->handle, plugin_declarations_sym))) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_FIND_DL_ENTRY, plugin_declarations_sym); + DBUG_RETURN(TRUE); + } + + /* convert mysql declaration to maria one */ + { + int i; + uint sizeof_st_plugin; + struct st_mysql_plugin *old; + struct st_maria_plugin *cur; + char *ptr= (char *)sym; + + if ((sym= dlsym(plugin_dl->handle, sizeof_st_plugin_sym))) + sizeof_st_plugin= *(int *)sym; + else + { + DBUG_ASSERT(min_plugin_interface_version == 0); + sizeof_st_plugin= (int)offsetof(struct st_mysql_plugin, version); + } + + for (i= 0; + ((struct st_mysql_plugin *)(ptr + i * sizeof_st_plugin))->info; + i++) + /* no op */; + + cur= (struct st_maria_plugin*) + my_malloc((i + 1) * sizeof(struct st_maria_plugin), + MYF(MY_ZEROFILL|MY_WME)); + if (!cur) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_OUTOFMEMORY, plugin_dl->dl.length); + DBUG_RETURN(TRUE); + } + /* + All st_plugin fields not initialized in the plugin explicitly, are + set to 0. It matches C standard behaviour for struct initializers that + have less values than the struct definition. + */ + for (i=0; + (old= (struct st_mysql_plugin *)(ptr + i * sizeof_st_plugin))->info; + i++) + { + + cur->type= old->type; + cur->info= old->info; + cur->name= old->name; + cur->author= old->author; + cur->descr= old->descr; + cur->license= old->license; + cur->init= old->init; + cur->deinit= old->deinit; + cur->version= old->version; + cur->status_vars= old->status_vars; + cur->system_vars= old->system_vars; + /* + Something like this should be added to process + new mysql plugin versions: + if (plugin_dl->mysqlversion > 0x0101) + { + cur->newfield= CONSTANT_MEANS_UNKNOWN; + } + else + { + cur->newfield= old->newfield; + } + */ + /* Maria only fields */ + cur->version_info= "Unknown"; + cur->maturity= MariaDB_PLUGIN_MATURITY_UNKNOWN; + } + plugin_dl->allocated= true; + plugin_dl->plugins= (struct st_maria_plugin *)cur; + } + + DBUG_RETURN(FALSE); +} + + +/** + Reads data from maria plugin interface + + @param plugin_dl Structure where the data should be put + @param sym Reverence on version info + @param dlpath Path to the module + @param report what errors should be reported + + @retval FALSE OK + @retval TRUE ERROR +*/ + +static my_bool read_maria_plugin_info(struct st_plugin_dl *plugin_dl, + void *sym, char *dlpath, + int report) +{ + DBUG_ENTER("read_maria_plugin_info"); + + /* Determine interface version */ + if (!(sym)) + { + /* + Actually this branch impossible because in case of absence of maria + version we try mysql version. + */ + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_FIND_DL_ENTRY, + maria_plugin_interface_version_sym); + DBUG_RETURN(TRUE); + } + plugin_dl->mariaversion= *(int *)sym; + plugin_dl->mysqlversion= 0; + /* Versioning */ + if (plugin_dl->mariaversion < min_maria_plugin_interface_version || + (plugin_dl->mariaversion >> 8) > (MARIA_PLUGIN_INTERFACE_VERSION >> 8)) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_OPEN_LIBRARY, dlpath, 0, + "plugin interface version mismatch"); + DBUG_RETURN(TRUE); + } + /* Find plugin declarations */ + if (!(sym= dlsym(plugin_dl->handle, maria_plugin_declarations_sym))) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_FIND_DL_ENTRY, maria_plugin_declarations_sym); + DBUG_RETURN(TRUE); + } + if (plugin_dl->mariaversion != MARIA_PLUGIN_INTERFACE_VERSION) + { + uint sizeof_st_plugin; + struct st_maria_plugin *old, *cur; + char *ptr= (char *)sym; + + if ((sym= dlsym(plugin_dl->handle, maria_sizeof_st_plugin_sym))) + sizeof_st_plugin= *(int *)sym; + else + { + free_plugin_mem(plugin_dl); + report_error(report, ER_CANT_FIND_DL_ENTRY, maria_sizeof_st_plugin_sym); + DBUG_RETURN(TRUE); + } + + if (sizeof_st_plugin != sizeof(st_mysql_plugin)) + { + int i; + for (i= 0; + ((struct st_maria_plugin *)(ptr + i * sizeof_st_plugin))->info; + i++) + /* no op */; + + cur= (struct st_maria_plugin*) + my_malloc((i + 1) * sizeof(struct st_maria_plugin), + MYF(MY_ZEROFILL|MY_WME)); + if (!cur) + { + free_plugin_mem(plugin_dl); + report_error(report, ER_OUTOFMEMORY, plugin_dl->dl.length); + DBUG_RETURN(TRUE); + } + /* + All st_plugin fields not initialized in the plugin explicitly, are + set to 0. It matches C standard behaviour for struct initializers that + have less values than the struct definition. + */ + for (i=0; + (old= (struct st_maria_plugin *)(ptr + i * sizeof_st_plugin))->info; + i++) + memcpy(cur + i, old, min(sizeof(cur[i]), sizeof_st_plugin)); + + sym= cur; + plugin_dl->allocated= true; + } + } + plugin_dl->plugins= (struct st_maria_plugin *)sym; + + DBUG_RETURN(FALSE); +} +#endif /* HAVE_DLOPEN */ + static st_plugin_dl *plugin_dl_add(const LEX_STRING *dl, int report) { #ifdef HAVE_DLOPEN @@ -415,22 +642,21 @@ static st_plugin_dl *plugin_dl_add(const report_error(report, ER_CANT_OPEN_LIBRARY, dlpath, errno, errmsg); DBUG_RETURN(0); } - /* Determine interface version */ - if (!(sym= dlsym(plugin_dl.handle, plugin_interface_version_sym))) + + /* Checks which plugin interface present and reads info */ + if (!(sym= dlsym(plugin_dl.handle, maria_plugin_interface_version_sym))) { - free_plugin_mem(&plugin_dl); - report_error(report, ER_CANT_FIND_DL_ENTRY, plugin_interface_version_sym); - DBUG_RETURN(0); + if (read_mysql_plugin_info(&plugin_dl, + dlsym(plugin_dl.handle, + plugin_interface_version_sym), + dlpath, + report)) + DBUG_RETURN(0); } - plugin_dl.version= *(int *)sym; - /* Versioning */ - if (plugin_dl.version < min_plugin_interface_version || - (plugin_dl.version >> 8) > (MYSQL_PLUGIN_INTERFACE_VERSION >> 8)) + else { - free_plugin_mem(&plugin_dl); - report_error(report, ER_CANT_OPEN_LIBRARY, dlpath, 0, - "plugin interface version mismatch"); - DBUG_RETURN(0); + if (read_maria_plugin_info(&plugin_dl, sym, dlpath, report)) + DBUG_RETURN(0); } /* link the services in */ @@ -438,7 +664,7 @@ static st_plugin_dl *plugin_dl_add(const { if ((sym= dlsym(plugin_dl.handle, list_of_services[i].name))) { - uint ver= (uint)(intptr)*(void**)sym; + uint ver= (uint)(intptr) *(void **)sym; if (ver > list_of_services[i].version || (ver >> 8) < (list_of_services[i].version >> 8)) { @@ -449,73 +675,10 @@ static st_plugin_dl *plugin_dl_add(const report_error(report, ER_CANT_OPEN_LIBRARY, dlpath, 0, buf); DBUG_RETURN(0); } - *(void**)sym= list_of_services[i].service; + *(void **)sym= list_of_services[i].service; } } - /* Find plugin declarations */ - if (!(sym= dlsym(plugin_dl.handle, plugin_declarations_sym))) - { - free_plugin_mem(&plugin_dl); - report_error(report, ER_CANT_FIND_DL_ENTRY, plugin_declarations_sym); - DBUG_RETURN(0); - } - - if (plugin_dl.version != MYSQL_PLUGIN_INTERFACE_VERSION) - { - uint sizeof_st_plugin; - struct st_mysql_plugin *old, *cur; - char *ptr= (char *)sym; - - if ((sym= dlsym(plugin_dl.handle, sizeof_st_plugin_sym))) - sizeof_st_plugin= *(int *)sym; - else - { -#ifdef ERROR_ON_NO_SIZEOF_PLUGIN_SYMBOL - free_plugin_mem(&plugin_dl); - report_error(report, ER_CANT_FIND_DL_ENTRY, sizeof_st_plugin_sym); - DBUG_RETURN(0); -#else - /* - When the following assert starts failing, we'll have to switch - to the upper branch of the #ifdef - */ - DBUG_ASSERT(min_plugin_interface_version == 0); - sizeof_st_plugin= (int)offsetof(struct st_mysql_plugin, version); -#endif - } - - if (sizeof_st_plugin != sizeof(st_mysql_plugin)) - { - for (i= 0; - ((struct st_mysql_plugin *)(ptr+i*sizeof_st_plugin))->info; - i++) - /* no op */; - - cur= (struct st_mysql_plugin*) - my_malloc((i+1)*sizeof(struct st_mysql_plugin), MYF(MY_ZEROFILL|MY_WME)); - if (!cur) - { - free_plugin_mem(&plugin_dl); - report_error(report, ER_OUTOFMEMORY, plugin_dl.dl.length); - DBUG_RETURN(0); - } - /* - All st_plugin fields not initialized in the plugin explicitly, are - set to 0. It matches C standard behaviour for struct initializers that - have less values than the struct definition. - */ - for (i=0; - (old=(struct st_mysql_plugin *)(ptr+i*sizeof_st_plugin))->info; - i++) - memcpy(cur+i, old, min(sizeof(cur[i]), sizeof_st_plugin)); - - sym= cur; - plugin_dl.allocated= true; - } - } - plugin_dl.plugins= (struct st_mysql_plugin *)sym; - /* Duplicate and convert dll name */ plugin_dl.dl.length= dl->length * files_charset_info->mbmaxlen + 1; if (! (plugin_dl.dl.str= (char*) my_malloc(plugin_dl.dl.length, MYF(0)))) @@ -665,7 +828,7 @@ static plugin_ref intern_plugin_lock(LEX *plugin= pi; #endif pi->ref_count++; - DBUG_PRINT("info",("thd: 0x%lx plugin: \"%s\" ref_count: %d", + DBUG_PRINT("lock",("thd: 0x%lx plugin: \"%s\" LOCK ref_count: %d", (long) current_thd, pi->name.str, pi->ref_count)); if (lex) @@ -756,7 +919,7 @@ static bool plugin_add(MEM_ROOT *tmp_roo int *argc, char **argv, int report) { struct st_plugin_int tmp; - struct st_mysql_plugin *plugin; + struct st_maria_plugin *plugin; DBUG_ENTER("plugin_add"); if (plugin_find_internal(name, MYSQL_ANY_PLUGIN)) { @@ -972,8 +1135,6 @@ static void intern_plugin_unlock(LEX *le my_free((uchar*) plugin, MYF(MY_WME)); #endif - DBUG_PRINT("info",("unlocking plugin, name= %s, ref_count= %d", - pi->name.str, pi->ref_count)); if (lex) { /* @@ -993,6 +1154,9 @@ static void intern_plugin_unlock(LEX *le DBUG_ASSERT(pi->ref_count); pi->ref_count--; + DBUG_PRINT("lock",("thd: 0x%lx plugin: \"%s\" UNLOCK ref_count: %d", + (long) current_thd, pi->name.str, pi->ref_count)); + if (pi->state == PLUGIN_IS_DELETED && !pi->ref_count) reap_needed= true; @@ -1161,8 +1325,8 @@ int plugin_init(int *argc, char **argv, { uint i; bool is_myisam; - struct st_mysql_plugin **builtins; - struct st_mysql_plugin *plugin; + struct st_maria_plugin **builtins; + struct st_maria_plugin *plugin; struct st_plugin_int tmp, *plugin_ptr, **reap; MEM_ROOT tmp_root; bool reaped_mandatory_plugin= FALSE; @@ -1201,7 +1365,7 @@ int plugin_init(int *argc, char **argv, /* First we register builtin plugins */ - for (builtins= mysqld_builtins; *builtins; builtins++) + for (builtins= mariadb_builtins; *builtins; builtins++) { for (plugin= *builtins; plugin->info; plugin++) { @@ -1316,7 +1480,7 @@ err: } -static bool register_builtin(struct st_mysql_plugin *plugin, +static bool register_builtin(struct st_maria_plugin *plugin, struct st_plugin_int *tmp, struct st_plugin_int **ptr) { @@ -1352,7 +1516,7 @@ static bool register_builtin(struct st_m RETURN false - plugin registered successfully */ -bool plugin_register_builtin(THD *thd, struct st_mysql_plugin *plugin) +bool plugin_register_builtin(THD *thd, struct st_maria_plugin *plugin) { struct st_plugin_int tmp, *ptr; bool result= true; @@ -1481,7 +1645,7 @@ static bool plugin_load_list(MEM_ROOT *t char buffer[FN_REFLEN]; LEX_STRING name= {buffer, 0}, dl= {NULL, 0}, *str= &name; struct st_plugin_dl *plugin_dl; - struct st_mysql_plugin *plugin; + struct st_maria_plugin *plugin; char *p= buffer; DBUG_ENTER("plugin_load_list"); while (list) @@ -1723,9 +1887,10 @@ bool mysql_install_plugin(THD *thd, cons if (tmp->state == PLUGIN_IS_DISABLED) { - push_warning_printf(thd, MYSQL_ERROR::WARN_LEVEL_WARN, - ER_CANT_INITIALIZE_UDF, ER(ER_CANT_INITIALIZE_UDF), - name->str, "Plugin is disabled"); + if (global_system_variables.log_warnings) + push_warning_printf(thd, MYSQL_ERROR::WARN_LEVEL_WARN, + ER_CANT_INITIALIZE_UDF, ER(ER_CANT_INITIALIZE_UDF), + name->str, "Plugin is disabled"); } else { === modified file 'sql/sql_plugin.h' --- a/sql/sql_plugin.h 2010-03-15 11:51:23 +0000 +++ b/sql/sql_plugin.h 2010-04-01 14:34:51 +0000 @@ -52,8 +52,9 @@ struct st_plugin_dl { LEX_STRING dl; void *handle; - struct st_mysql_plugin *plugins; - int version; + struct st_maria_plugin *plugins; + int mysqlversion; + int mariaversion; bool allocated; uint ref_count; /* number of plugins loaded from the library */ }; @@ -63,7 +64,7 @@ struct st_plugin_dl struct st_plugin_int { LEX_STRING name; - struct st_mysql_plugin *plugin; + struct st_maria_plugin *plugin; struct st_plugin_dl *plugin_dl; uint state; uint ref_count; /* number of threads using the plugin */ === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2010-03-15 11:51:23 +0000 +++ b/sql/sql_select.cc 2010-04-08 21:03:07 +0000 @@ -2335,7 +2335,13 @@ JOIN::destroy() tab->cleanup(); } tmp_join->tmp_join= 0; + /* + We need to clean up tmp_table_param for reusable JOINs (having non-zero + and different from self tmp_join) because it's not being cleaned up + anywhere else (as we need to keep the join is reusable). + */ tmp_table_param.cleanup(); + tmp_join->tmp_table_param.copy_field= 0; DBUG_RETURN(tmp_join->destroy()); } cond_equal= 0; === modified file 'sql/sql_show.cc' --- a/sql/sql_show.cc 2010-03-15 11:51:23 +0000 +++ b/sql/sql_show.cc 2010-04-08 12:10:05 +0000 @@ -18,6 +18,7 @@ #include "mysql_priv.h" #include "sql_select.h" // For select_describe +#include "create_options.h" #include "sql_show.h" #include "repl_failsafe.h" #include "sp.h" @@ -94,11 +95,21 @@ static int make_version_string(char *buf return my_snprintf(buf, buf_length, "%d.%d", version>>8,version&0xff); } + +static const LEX_STRING maturity_name[]={ + { C_STRING_WITH_LEN("Unknown") }, + { C_STRING_WITH_LEN("Experimental") }, + { C_STRING_WITH_LEN("Alpha") }, + { C_STRING_WITH_LEN("Beta") }, + { C_STRING_WITH_LEN("Gamma") }, + { C_STRING_WITH_LEN("Stable") }}; + + static my_bool show_plugins(THD *thd, plugin_ref plugin, void *arg) { TABLE *table= (TABLE*) arg; - struct st_mysql_plugin *plug= plugin_decl(plugin); + struct st_maria_plugin *plug= plugin_decl(plugin); struct st_plugin_dl *plugin_dl= plugin_dlib(plugin); CHARSET_INFO *cs= system_charset_info; char version_buf[20]; @@ -143,7 +154,7 @@ static my_bool show_plugins(THD *thd, pl table->field[5]->set_notnull(); table->field[6]->store(version_buf, make_version_string(version_buf, sizeof(version_buf), - plugin_dl->version), + plugin_dl->mariaversion), cs); table->field[6]->set_notnull(); } @@ -186,6 +197,26 @@ static my_bool show_plugins(THD *thd, pl } table->field[9]->set_notnull(); + if ((uint) plug->maturity <= MariaDB_PLUGIN_MATURITY_STABLE) + table->field[10]->store(maturity_name[plug->maturity].str, + maturity_name[plug->maturity].length, + cs); + else + { + DBUG_ASSERT(0); + table->field[10]->store("Unknown", 7, cs); + } + table->field[10]->set_notnull(); + + if (plug->version_info) + { + table->field[11]->store(plug->version_info, + strlen(plug->version_info), cs); + table->field[11]->set_notnull(); + } + else + table->field[11]->set_null(); + return schema_table_store_record(thd, table); } @@ -1013,7 +1044,7 @@ append_identifier(THD *thd, String *pack /* The identifier must be quoted as it includes a quote character or - it's a keyword + it's a keyword */ VOID(packet->reserve(length*2 + 2)); @@ -1173,6 +1204,31 @@ static bool get_field_default_value(THD return has_default; } + +/** + Appends list of options to string + + @param thd thread handler + @param packet string to append + @param opt list of options +*/ + +static void append_create_options(THD *thd, String *packet, + engine_option_value *opt) +{ + for(; opt; opt= opt->next) + { + DBUG_ASSERT(opt->value.str); + packet->append(' '); + append_identifier(thd, packet, opt->name.str, opt->name.length); + packet->append('='); + if (opt->quoted_value) + append_unescaped(packet, opt->value.str, opt->value.length); + else + packet->append(opt->value.str, opt->value.length); + } +} + /* Build a CREATE TABLE statement for a table. @@ -1355,6 +1411,7 @@ int store_create_info(THD *thd, TABLE_LI packet->append(STRING_WITH_LEN(" COMMENT ")); append_unescaped(packet, field->comment.str, field->comment.length); } + append_create_options(thd, packet, field->option_list); } key_info= table->key_info; @@ -1426,6 +1483,7 @@ int store_create_info(THD *thd, TABLE_LI append_identifier(thd, packet, parser_name->str, parser_name->length); packet->append(STRING_WITH_LEN(" */ ")); } + append_create_options(thd, packet, key_info->option_list); } /* @@ -1585,6 +1643,7 @@ int store_create_info(THD *thd, TABLE_LI packet->append(STRING_WITH_LEN(" CONNECTION=")); append_unescaped(packet, share->connect_string.str, share->connect_string.length); } + append_create_options(thd, packet, share->option_list); append_directory(thd, packet, "DATA", create_info.data_file_name); append_directory(thd, packet, "INDEX", create_info.index_file_name); } @@ -2219,6 +2278,35 @@ void remove_status_vars(SHOW_VAR *list) } + +static void update_key_cache_stat_var(KEY_CACHE *key_cache, size_t ofs) +{ + uint var_no; + if (ofs == offsetof(KEY_CACHE, blocks_used) || + ofs == offsetof(KEY_CACHE, blocks_unused) || + ofs == offsetof(KEY_CACHE, global_blocks_changed)) + { + var_no= (ofs-offsetof(KEY_CACHE, blocks_used))/sizeof(ulong); + *(ulong *)((char *) key_cache + ofs)= + (ulong) get_key_cache_stat_value(key_cache, var_no); + return; + } + + if (ofs == offsetof(KEY_CACHE, global_cache_r_requests) || + ofs == offsetof(KEY_CACHE, global_cache_read) || + ofs == offsetof(KEY_CACHE, global_cache_w_requests) || + ofs == offsetof(KEY_CACHE, global_cache_write)) + { + var_no= NUM_LONG_KEY_CACHE_STAT_VARIABLES + + (ofs-offsetof(KEY_CACHE, global_cache_w_requests))/ + sizeof(ulonglong); + *(ulonglong *)((char *) key_cache + ofs)= + get_key_cache_stat_value(key_cache, var_no); + return; + } +} + + static bool show_status_array(THD *thd, const char *wild, SHOW_VAR *variables, enum enum_var_type value_type, @@ -2351,10 +2439,12 @@ static bool show_status_array(THD *thd, break; } case SHOW_KEY_CACHE_LONG: + update_key_cache_stat_var(dflt_key_cache, (size_t) value); value= (char*) dflt_key_cache + (ulong)value; end= int10_to_str(*(long*) value, buff, 10); break; case SHOW_KEY_CACHE_LONGLONG: + update_key_cache_stat_var(dflt_key_cache, (size_t) value); value= (char*) dflt_key_cache + (ulong)value; end= longlong10_to_str(*(longlong*) value, buff, 10); break; @@ -2499,8 +2589,8 @@ int send_user_stats(THD* thd, HASH *all_ table->field[j++]->store(user_stats->user, user_stats->user_name_length, system_charset_info); table->field[j++]->store((longlong)user_stats->total_connections,TRUE); - table->field[j++]->store((longlong)user_stats->concurrent_connections); - table->field[j++]->store((longlong)user_stats->connected_time); + table->field[j++]->store((longlong)user_stats->concurrent_connections, TRUE); + table->field[j++]->store((longlong)user_stats->connected_time, TRUE); table->field[j++]->store((double)user_stats->busy_time); table->field[j++]->store((double)user_stats->cpu_time); table->field[j++]->store((longlong)user_stats->bytes_received, TRUE); @@ -4398,7 +4488,7 @@ static my_bool iter_schema_engines(THD * if (plugin_state(plugin) != PLUGIN_IS_READY) { - struct st_mysql_plugin *plug= plugin_decl(plugin); + struct st_maria_plugin *plug= plugin_decl(plugin); if (!(wild && wild[0] && wild_case_compare(scs, plug->name,wild))) { @@ -6610,6 +6700,90 @@ int fill_schema_files(THD *thd, TABLE_LI } +static +int store_key_cache_table_record(THD *thd, TABLE *table, + const char *name, uint name_length, + KEY_CACHE *key_cache, + uint partitions, uint partition_no) +{ + KEY_CACHE_STATISTICS keycache_stats; + uint err; + DBUG_ENTER("store_key_cache_table_record"); + + get_key_cache_statistics(key_cache, partition_no, &keycache_stats); + + if (!key_cache->key_cache_inited || keycache_stats.mem_size == 0) + DBUG_RETURN(0); + + restore_record(table, s->default_values); + table->field[0]->store(name, name_length, system_charset_info); + if (partitions == 0) + table->field[1]->set_null(); + else + { + table->field[1]->set_notnull(); + table->field[1]->store((long) partitions, TRUE); + } + + if (partition_no == 0) + table->field[2]->set_null(); + else + { + table->field[2]->set_notnull(); + table->field[2]->store((long) partition_no, TRUE); + } + table->field[3]->store(keycache_stats.mem_size, TRUE); + table->field[4]->store(keycache_stats.block_size, TRUE); + table->field[5]->store(keycache_stats.blocks_used, TRUE); + table->field[6]->store(keycache_stats.blocks_unused, TRUE); + table->field[7]->store(keycache_stats.blocks_changed, TRUE); + table->field[8]->store(keycache_stats.read_requests, TRUE); + table->field[9]->store(keycache_stats.reads, TRUE); + table->field[10]->store(keycache_stats.write_requests, TRUE); + table->field[11]->store(keycache_stats.writes, TRUE); + + err= schema_table_store_record(thd, table); + DBUG_RETURN(err); +} + + +int fill_key_cache_tables(THD *thd, TABLE_LIST *tables, COND *cond) +{ + TABLE *table= tables->table; + I_List_iterator<NAMED_LIST> it(key_caches); + NAMED_LIST *element; + DBUG_ENTER("fill_key_cache_tables"); + + while ((element= it++)) + { + KEY_CACHE *key_cache= (KEY_CACHE *) element->data; + + if (!key_cache->key_cache_inited) + continue; + + uint partitions= key_cache->partitions; + DBUG_ASSERT(partitions <= MAX_KEY_CACHE_PARTITIONS); + + if (partitions) + { + for (uint i= 0; i < partitions; i++) + { + if (store_key_cache_table_record(thd, table, + element->name, element->name_length, + key_cache, partitions, i+1)) + DBUG_RETURN(1); + } + } + + if (store_key_cache_table_record(thd, table, + element->name, element->name_length, + key_cache, partitions, 0)) + DBUG_RETURN(1); + } + DBUG_RETURN(0); +} + + ST_FIELD_INFO schema_fields_info[]= { {"CATALOG_NAME", FN_REFLEN, MYSQL_TYPE_STRING, 0, 1, 0, SKIP_OPEN_TABLE}, @@ -7095,6 +7269,8 @@ ST_FIELD_INFO plugin_fields_info[]= {"PLUGIN_AUTHOR", NAME_CHAR_LEN, MYSQL_TYPE_STRING, 0, 1, 0, SKIP_OPEN_TABLE}, {"PLUGIN_DESCRIPTION", 65535, MYSQL_TYPE_STRING, 0, 1, 0, SKIP_OPEN_TABLE}, {"PLUGIN_LICENSE", 80, MYSQL_TYPE_STRING, 0, 1, "License", SKIP_OPEN_TABLE}, + {"PLUGIN_MATURITY", 12, MYSQL_TYPE_STRING, 0, 1, 0, SKIP_OPEN_TABLE}, + {"PLUGIN_AUTH_VERSION", 80, MYSQL_TYPE_STRING, 0, 1, 0, SKIP_OPEN_TABLE}, {0, 0, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE} }; @@ -7187,6 +7363,35 @@ ST_FIELD_INFO referential_constraints_fi }; +ST_FIELD_INFO keycache_fields_info[]= +{ + {"KEY_CACHE_NAME", NAME_LEN, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE}, + {"PARTITIONS", 3, MYSQL_TYPE_LONG, 0, + (MY_I_S_MAYBE_NULL | MY_I_S_UNSIGNED) , 0, SKIP_OPEN_TABLE}, + {"PARTITION_NUMBER", 3, MYSQL_TYPE_LONG, 0, + (MY_I_S_MAYBE_NULL | MY_I_S_UNSIGNED), 0, SKIP_OPEN_TABLE}, + {"FULL_SIZE", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), 0, SKIP_OPEN_TABLE}, + {"BLOCK_SIZE", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), 0, SKIP_OPEN_TABLE }, + {"USED_BLOCKS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_blocks_used", SKIP_OPEN_TABLE}, + {"UNUSED_BLOCKS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_blocks_unused", SKIP_OPEN_TABLE}, + {"DIRTY_BLOCKS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_blocks_not_flushed", SKIP_OPEN_TABLE}, + {"READ_REQUESTS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_read_requests", SKIP_OPEN_TABLE}, + {"READS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_reads", SKIP_OPEN_TABLE}, + {"WRITE_REQUESTS", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_write_requests", SKIP_OPEN_TABLE}, + {"WRITES", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, + (MY_I_S_UNSIGNED), "Key_writes", SKIP_OPEN_TABLE}, + {0, 0, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE} +}; + + /* Description of ST_FIELD_INFO in table.h @@ -7226,6 +7431,8 @@ ST_SCHEMA_TABLE schema_tables[]= fill_variables, make_old_format, 0, 0, -1, 0, 0}, {"INDEX_STATISTICS", index_stats_fields_info, create_schema_table, fill_schema_index_stats, make_old_format, 0, -1, -1, 0, 0}, + {"KEY_CACHES", keycache_fields_info, create_schema_table, + fill_key_cache_tables, make_old_format, 0, -1,-1, 0, 0}, {"KEY_COLUMN_USAGE", key_column_usage_fields_info, create_schema_table, get_all_tables, 0, get_schema_key_column_usage_record, 4, 5, 0, OPEN_TABLE_ONLY}, === modified file 'sql/sql_table.cc' --- a/sql/sql_table.cc 2010-03-31 08:17:03 +0000 +++ b/sql/sql_table.cc 2010-04-09 09:26:17 +0000 @@ -19,6 +19,7 @@ #include <hash.h> #include <myisam.h> #include <my_dir.h> +#include "create_options.h" #include "sp_head.h" #include "sql_trigger.h" #include "sql_show.h" @@ -42,17 +43,10 @@ static int copy_data_between_tables(TABL static bool prepare_blob_field(THD *thd, Create_field *sql_field); static bool check_engine(THD *, const char *, HA_CREATE_INFO *); -static int -mysql_prepare_create_table(THD *thd, HA_CREATE_INFO *create_info, - Alter_info *alter_info, - bool tmp_table, - uint *db_options, - handler *file, KEY **key_info_buffer, - uint *key_count, int select_field_count); -static bool -mysql_prepare_alter_table(THD *thd, TABLE *table, - HA_CREATE_INFO *create_info, - Alter_info *alter_info); +static int mysql_prepare_create_table(THD *, HA_CREATE_INFO *, Alter_info *, + bool, uint *, handler *, KEY **, uint *, int); +static bool mysql_prepare_alter_table(THD *, TABLE *, HA_CREATE_INFO *, + Alter_info *); #ifndef DBUG_OFF @@ -2863,6 +2857,11 @@ mysql_prepare_create_table(THD *thd, HA_ sql_field->offset= record_offset; if (MTYP_TYPENR(sql_field->unireg_check) == Field::NEXT_NUMBER) auto_increment++; + if (parse_option_list(thd, &sql_field->option_struct, + sql_field->option_list, + create_info->db_type->field_options, FALSE, + thd->mem_root)) + DBUG_RETURN(TRUE); /* For now skip fields that are not physically stored in the database (virtual fields) and update their offset later @@ -3061,6 +3060,12 @@ mysql_prepare_create_table(THD *thd, HA_ key_info->key_part=key_part_info; key_info->usable_key_parts= key_number; key_info->algorithm= key->key_create_info.algorithm; + key_info->option_list= key->option_list; + if (parse_option_list(thd, &key_info->option_struct, + key_info->option_list, + create_info->db_type->index_options, FALSE, + thd->mem_root)) + DBUG_RETURN(TRUE); if (key->type == Key::FULLTEXT) { @@ -3438,6 +3443,12 @@ mysql_prepare_create_table(THD *thd, HA_ } } + if (parse_option_list(thd, &create_info->option_struct, + create_info->option_list, + create_info->db_type->table_options, FALSE, + thd->mem_root)) + DBUG_RETURN(TRUE); + DBUG_RETURN(FALSE); } @@ -5679,6 +5690,7 @@ compare_tables(TABLE *table, KEY_PART_INFO *key_part; KEY_PART_INFO *end; THD *thd= table->in_use; + uint i; /* Remember if the new definition has new VARCHAR column; create_info->varchar will be reset in mysql_prepare_create_table. @@ -5769,6 +5781,12 @@ compare_tables(TABLE *table, DBUG_RETURN(0); } + if ((create_info->fileds_option_struct= + (void**)thd->calloc(sizeof(void*) * table->s->fields)) == NULL || + (create_info->keys_option_struct= + (void**)thd->calloc(sizeof(void*) * table->s->keys)) == NULL) + DBUG_RETURN(1); + /* Use transformed info to evaluate possibility of fast ALTER TABLE but use the preserved field to persist modifications. @@ -5776,15 +5794,19 @@ compare_tables(TABLE *table, new_field_it.init(alter_info->create_list); tmp_new_field_it.init(tmp_alter_info.create_list); - /* Go through fields and check if the original ones are compatible + /* + Go through fields and check if the original ones are compatible with new table. */ - for (f_ptr= table->field, new_field= new_field_it++, + for (i= 0, f_ptr= table->field, new_field= new_field_it++, tmp_new_field= tmp_new_field_it++; (field= *f_ptr); - f_ptr++, new_field= new_field_it++, + i++, f_ptr++, new_field= new_field_it++, tmp_new_field= tmp_new_field_it++) { + DBUG_ASSERT(i < table->s->fields); + create_info->fileds_option_struct[i]= tmp_new_field->option_struct; + /* Make sure we have at least the default charset in use. */ if (!new_field->charset) new_field->charset= create_info->default_table_charset; @@ -5938,7 +5960,9 @@ compare_tables(TABLE *table, for (new_key= *key_info_buffer; new_key < new_key_end; new_key++) { /* Search an old key with the same name. */ - for (table_key= table->key_info; table_key < table_key_end; table_key++) + for (i= 0, table_key= table->key_info; + table_key < table_key_end; + i++, table_key++) { if (! strcmp(table_key->name, new_key->name)) break; @@ -5957,6 +5981,11 @@ compare_tables(TABLE *table, } DBUG_PRINT("info", ("index added: '%s'", new_key->name)); } + else + { + DBUG_ASSERT(i < table->s->keys); + create_info->keys_option_struct[i]= new_key->option_struct; + } } /* Check if changes are compatible with current handler without a copy */ @@ -6132,6 +6161,8 @@ mysql_prepare_alter_table(THD *thd, TABL } restore_record(table, s->default_values); // Empty record for DEFAULT + create_info->option_list= merge_engine_table_options(table->s->option_list, + create_info->option_list, thd->mem_root); /* First collect all fields from table which isn't in drop_list */ @@ -6384,7 +6415,7 @@ mysql_prepare_alter_table(THD *thd, TABL key= new Key(key_type, key_name, &key_create_info, test(key_info->flags & HA_GENERATED_KEY), - key_parts); + key_parts, key_info->option_list); new_key_list.push_back(key); } } === modified file 'sql/sql_test.cc' --- a/sql/sql_test.cc 2010-03-04 08:03:07 +0000 +++ b/sql/sql_test.cc 2010-03-29 21:16:12 +0000 @@ -445,7 +445,8 @@ static int print_key_cache_status(const Buffer_size: %10lu\n\ Block_size: %10lu\n\ Division_limit: %10lu\n\ -Age_limit: %10lu\n\ +Age_threshold: %10lu\n\ +Partitions: %10lu\n\ blocks used: %10lu\n\ not flushed: %10lu\n\ w_requests: %10s\n\ @@ -455,6 +456,7 @@ reads: %10s\n\n", name, (ulong) key_cache->param_buff_size, key_cache->param_block_size, key_cache->param_division_limit, key_cache->param_age_threshold, + key_cache->param_partitions, key_cache->blocks_used,key_cache->global_blocks_changed, llstr(key_cache->global_cache_w_requests,llbuff1), llstr(key_cache->global_cache_write,llbuff2), === modified file 'sql/sql_view.cc' --- a/sql/sql_view.cc 2010-03-30 12:36:49 +0000 +++ b/sql/sql_view.cc 2010-04-08 12:10:05 +0000 @@ -1185,7 +1185,7 @@ bool mysql_make_view(THD *thd, File_pars + MODE_PIPES_AS_CONCAT affect expression parsing + MODE_ANSI_QUOTES affect expression parsing + MODE_IGNORE_SPACE affect expression parsing - - MODE_NOT_USED not used :) + - MODE_IGNORE_BAD_TABLE_OPTIONS affect only CREATE/ALTER TABLE parsing * MODE_ONLY_FULL_GROUP_BY affect execution * MODE_NO_UNSIGNED_SUBTRACTION affect execution - MODE_NO_DIR_IN_CREATE affect table creation only === modified file 'sql/sql_yacc.yy' --- a/sql/sql_yacc.yy 2010-03-31 08:17:03 +0000 +++ b/sql/sql_yacc.yy 2010-04-09 09:26:17 +0000 @@ -44,6 +44,7 @@ #include "sp_rcontext.h" #include "sp.h" #include "event_parse_data.h" +#include "create_options.h" #include <myisam.h> #include <myisammrg.h> @@ -608,6 +609,7 @@ static bool add_create_index_prepare (LE lex->alter_info.flags= ALTER_ADD_INDEX; lex->col_list.empty(); lex->change= NullS; + lex->option_list= NULL; return FALSE; } @@ -617,7 +619,7 @@ static bool add_create_index (LEX *lex, { Key *key; key= new Key(type, name, info ? info : &lex->key_create_info, generated, - lex->col_list); + lex->col_list, lex->option_list); if (key == NULL) return TRUE; @@ -2342,6 +2344,7 @@ sp_init_param: lex->interval_list.empty(); lex->uint_geom_type= 0; + lex->vcol_info= 0; } ; @@ -3897,7 +3900,11 @@ create2: ; create2a: - field_list ')' opt_create_table_options + field_list ')' + { + Lex->create_info.option_list= NULL; + } + opt_create_table_options opt_partitioning create3 {} | opt_partitioning @@ -4750,6 +4757,30 @@ create_table_option: Lex->create_info.used_fields|= HA_CREATE_USED_TRANSACTIONAL; Lex->create_info.transactional= $3; } + | IDENT_sys equal TEXT_STRING_sys + { + new (YYTHD->mem_root) + engine_option_value($1, $3, true, &Lex->create_info.option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ident + { + new (YYTHD->mem_root) + engine_option_value($1, $3, false, &Lex->create_info.option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ulonglong_num + { + new (YYTHD->mem_root) + engine_option_value($1, $3, &Lex->create_info.option_list, + &Lex->option_list_last, YYTHD->mem_root); + } + | IDENT_sys equal DEFAULT + { + new (YYTHD->mem_root) + engine_option_value($1, &Lex->create_info.option_list, + &Lex->option_list_last); + } ; default_charset: @@ -4871,25 +4902,33 @@ column_def: ; key_def: - normal_key_type opt_ident key_alg '(' key_list ')' normal_key_options + normal_key_type opt_ident key_alg '(' key_list ')' + { Lex->option_list= NULL; } + normal_key_options { if (add_create_index (Lex, $1, $2)) MYSQL_YYABORT; } | fulltext opt_key_or_index opt_ident init_key_options - '(' key_list ')' fulltext_key_options + '(' key_list ')' + { Lex->option_list= NULL; } + fulltext_key_options { if (add_create_index (Lex, $1, $3)) MYSQL_YYABORT; } | spatial opt_key_or_index opt_ident init_key_options - '(' key_list ')' spatial_key_options + '(' key_list ')' + { Lex->option_list= NULL; } + spatial_key_options { if (add_create_index (Lex, $1, $3)) MYSQL_YYABORT; } | opt_constraint constraint_key_type opt_ident key_alg - '(' key_list ')' normal_key_options + '(' key_list ')' + { Lex->option_list= NULL; } + normal_key_options { if (add_create_index (Lex, $2, $3 ? $3 : $1)) MYSQL_YYABORT; @@ -4952,6 +4991,7 @@ field_spec: lex->comment=null_lex_str; lex->charset=NULL; lex->vcol_info= 0; + lex->option_list= NULL; } field_def { @@ -4962,7 +5002,7 @@ field_spec: &lex->comment, lex->change,&lex->interval_list,lex->charset, lex->uint_geom_type, - lex->vcol_info)) + lex->vcol_info, lex->option_list)) MYSQL_YYABORT; } ; @@ -5382,6 +5422,29 @@ attribute: Lex->charset=$2; } } + | IDENT_sys equal TEXT_STRING_sys + { + new (YYTHD->mem_root) + engine_option_value($1, $3, true, &Lex->option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ident + { + new (YYTHD->mem_root) + engine_option_value($1, $3, false, &Lex->option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ulonglong_num + { + new (YYTHD->mem_root) + engine_option_value($1, $3, &Lex->option_list, + &Lex->option_list_last, YYTHD->mem_root); + } + | IDENT_sys equal DEFAULT + { + new (YYTHD->mem_root) + engine_option_value($1, &Lex->option_list, &Lex->option_list_last); + } ; now_or_signed_literal: @@ -5671,6 +5734,29 @@ key_using_alg: all_key_opt: KEY_BLOCK_SIZE opt_equal ulong_num { Lex->key_create_info.block_size= $3; } + | IDENT_sys equal TEXT_STRING_sys + { + new (YYTHD->mem_root) + engine_option_value($1, $3, true, &Lex->option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ident + { + new (YYTHD->mem_root) + engine_option_value($1, $3, false, &Lex->option_list, + &Lex->option_list_last); + } + | IDENT_sys equal ulonglong_num + { + new (YYTHD->mem_root) + engine_option_value($1, $3, &Lex->option_list, + &Lex->option_list_last, YYTHD->mem_root); + } + | IDENT_sys equal DEFAULT + { + new (YYTHD->mem_root) + engine_option_value($1, &Lex->option_list, &Lex->option_list_last); + } ; normal_key_opt: @@ -6160,6 +6246,7 @@ alter_list_item: LEX *lex=Lex; lex->change= $3.str; lex->alter_info.flags|= ALTER_CHANGE_COLUMN; + lex->option_list= NULL; } field_spec opt_place | MODIFY_SYM opt_column field_ident @@ -6171,6 +6258,7 @@ alter_list_item: lex->charset= NULL; lex->alter_info.flags|= ALTER_CHANGE_COLUMN; lex->vcol_info= 0; + lex->option_list= NULL; } field_def { @@ -6182,7 +6270,7 @@ alter_list_item: &lex->comment, $3.str, &lex->interval_list, lex->charset, lex->uint_geom_type, - lex->vcol_info)) + lex->vcol_info, lex->option_list)) MYSQL_YYABORT; } opt_place @@ -6289,8 +6377,7 @@ alter_list_item: } | create_table_options_space_separated { - LEX *lex=Lex; - lex->alter_info.flags|= ALTER_OPTIONS; + Lex->alter_info.flags|= ALTER_OPTIONS; } | FORCE_SYM { === modified file 'sql/structs.h' --- a/sql/structs.h 2010-03-29 15:13:53 +0000 +++ b/sql/structs.h 2010-04-08 12:10:05 +0000 @@ -68,6 +68,7 @@ typedef struct st_key_part_info { /* Inf uint8 null_bit; /* Position to null_bit */ } KEY_PART_INFO ; +class engine_option_value; typedef struct st_key { uint key_length; /* Tot length of key */ @@ -101,6 +102,9 @@ typedef struct st_key { int bdb_return_if_eq; } handler; struct st_table *table; + /** reference to the list of options or NULL */ + engine_option_value *option_list; + void *option_struct; /* structure with parsed options */ } KEY; === modified file 'sql/table.cc' --- a/sql/table.cc 2010-03-29 15:13:53 +0000 +++ b/sql/table.cc 2010-04-08 12:10:05 +0000 @@ -18,6 +18,7 @@ #include "mysql_priv.h" #include "sql_trigger.h" +#include "create_options.h" #include <m_ctype.h> #include "my_md5.h" @@ -667,12 +668,13 @@ static int open_binary_frm(THD *thd, TAB uint db_create_options, keys, key_parts, n_length; uint key_info_length, com_length, null_bit_pos; uint vcol_screen_length; - uint extra_rec_buf_length; + uint extra_rec_buf_length, options_len; uint i,j; bool use_hash; char *keynames, *names, *comment_pos, *vcol_screen_pos; uchar *record; - uchar *disk_buff, *strpos, *null_flags, *null_pos; + uchar *disk_buff, *strpos, *null_flags, *null_pos, *options; + uchar *buff= 0; ulong pos, record_offset, *rec_per_key, rec_buff_length; handler *handler_file= 0; KEY *keyinfo; @@ -788,7 +790,6 @@ static int open_binary_frm(THD *thd, TAB for (i=0 ; i < keys ; i++, keyinfo++) { - keyinfo->table= 0; // Updated in open_frm if (new_frm_ver >= 3) { keyinfo->flags= (uint) uint2korr(strpos) ^ HA_NOSAME; @@ -858,15 +859,14 @@ static int open_binary_frm(THD *thd, TAB if ((n_length= uint4korr(head+55))) { /* Read extra data segment */ - uchar *buff, *next_chunk, *buff_end; + uchar *next_chunk, *buff_end; DBUG_PRINT("info", ("extra segment size is %u bytes", n_length)); if (!(next_chunk= buff= (uchar*) my_malloc(n_length, MYF(MY_WME)))) goto err; if (my_pread(file, buff, n_length, record_offset + share->reclength, MYF(MY_NABP))) { - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } share->connect_string.length= uint2korr(buff); if (!(share->connect_string.str= strmake_root(&share->mem_root, @@ -874,8 +874,7 @@ static int open_binary_frm(THD *thd, TAB share->connect_string. length))) { - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } next_chunk+= share->connect_string.length + 2; buff_end= buff + n_length; @@ -895,8 +894,7 @@ static int open_binary_frm(THD *thd, TAB plugin_data(tmp_plugin, handlerton *))) { /* bad file, legacy_db_type did not match the name */ - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } /* tmp_plugin is locked with a local lock. @@ -925,8 +923,7 @@ static int open_binary_frm(THD *thd, TAB error= 8; my_error(ER_OPTION_PREVENTS_STATEMENT, MYF(0), "--skip-partition"); - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } plugin_unlock(NULL, share->db_plugin); share->db_plugin= ha_lock_engine(NULL, partition_hton); @@ -940,8 +937,7 @@ static int open_binary_frm(THD *thd, TAB /* purecov: begin inspected */ error= 8; my_error(ER_UNKNOWN_STORAGE_ENGINE, MYF(0), name.str); - my_free(buff, MYF(0)); - goto err; + goto free_and_err; /* purecov: end */ } next_chunk+= str_db_type_length + 2; @@ -957,16 +953,14 @@ static int open_binary_frm(THD *thd, TAB memdup_root(&share->mem_root, next_chunk + 4, partition_info_len + 1))) { - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } } #else if (partition_info_len) { DBUG_PRINT("info", ("WITH_PARTITION_STORAGE_ENGINE is not defined")); - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } #endif next_chunk+= 5 + partition_info_len; @@ -1002,8 +996,7 @@ static int open_binary_frm(THD *thd, TAB { DBUG_PRINT("error", ("fulltext key uses parser that is not defined in .frm")); - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } parser_name.str= (char*) next_chunk; parser_name.length= strlen((char*) next_chunk); @@ -1013,12 +1006,20 @@ static int open_binary_frm(THD *thd, TAB if (! keyinfo->parser) { my_error(ER_PLUGIN_IS_NOT_LOADED, MYF(0), parser_name.str); - my_free(buff, MYF(0)); - goto err; + goto free_and_err; } } } - my_free(buff, MYF(0)); + if (share->db_create_options & HA_OPTION_TEXT_CREATE_OPTIONS) + { + /* + store options position, but skip till the time we will + know number of fields + */ + options_len= uint4korr(next_chunk); + options= next_chunk + 4; + next_chunk+= options_len + 4; + } } share->key_block_size= uint2korr(head+62); @@ -1028,21 +1029,21 @@ static int open_binary_frm(THD *thd, TAB share->rec_buff_length= rec_buff_length; if (!(record= (uchar *) alloc_root(&share->mem_root, rec_buff_length))) - goto err; /* purecov: inspected */ + goto free_and_err; /* purecov: inspected */ share->default_values= record; if (my_pread(file, record, (size_t) share->reclength, record_offset, MYF(MY_NABP))) - goto err; /* purecov: inspected */ + goto free_and_err; /* purecov: inspected */ VOID(my_seek(file,pos,MY_SEEK_SET,MYF(0))); if (my_read(file, head,288,MYF(MY_NABP))) - goto err; + goto free_and_err; #ifdef HAVE_CRYPTED_FRM if (crypted) { crypted->decode((char*) head+256,288-256); if (sint2korr(head+284) != 0) // Should be 0 - goto err; // Wrong password + goto free_and_err; // Wrong password } #endif @@ -1062,6 +1063,8 @@ static int open_binary_frm(THD *thd, TAB share->comment.length); DBUG_PRINT("info",("i_count: %d i_parts: %d index: %d n_length: %d int_length: %d com_length: %d vcol_screen_length: %d", interval_count,interval_parts, share->keys,n_length,int_length, com_length, vcol_screen_length)); + + if (!(field_ptr = (Field **) alloc_root(&share->mem_root, (uint) ((share->fields+1)*sizeof(Field*)+ @@ -1070,14 +1073,14 @@ static int open_binary_frm(THD *thd, TAB keys+3)*sizeof(char *)+ (n_length+int_length+com_length+ vcol_screen_length))))) - goto err; /* purecov: inspected */ + goto free_and_err; /* purecov: inspected */ share->field= field_ptr; read_length=(uint) (share->fields * field_pack_length + pos+ (uint) (n_length+int_length+com_length+ vcol_screen_length)); if (read_string(file,(uchar**) &disk_buff,read_length)) - goto err; /* purecov: inspected */ + goto free_and_err; /* purecov: inspected */ #ifdef HAVE_CRYPTED_FRM if (crypted) { @@ -1104,7 +1107,7 @@ static int open_binary_frm(THD *thd, TAB fix_type_pointers(&interval_array, &share->fieldnames, 1, &names); if (share->fieldnames.count != share->fields) - goto err; + goto free_and_err; fix_type_pointers(&interval_array, share->intervals, interval_count, &names); @@ -1118,7 +1121,7 @@ static int open_binary_frm(THD *thd, TAB uint count= (uint) (interval->count + 1) * sizeof(uint); if (!(interval->type_lengths= (uint *) alloc_root(&share->mem_root, count))) - goto err; + goto free_and_err; for (count= 0; count < interval->count; count++) { char *val= (char*) interval->type_names[count]; @@ -1134,7 +1137,7 @@ static int open_binary_frm(THD *thd, TAB /* Allocate handler */ if (!(handler_file= get_new_handler(share, thd->mem_root, share->db_type()))) - goto err; + goto free_and_err; record= share->default_values-1; /* Fieldstart = 1 */ if (share->null_field_first) @@ -1196,7 +1199,7 @@ static int open_binary_frm(THD *thd, TAB charset= &my_charset_bin; #else error= 4; // unsupported field type - goto err; + goto free_and_err; #endif } else @@ -1207,7 +1210,7 @@ static int open_binary_frm(THD *thd, TAB { error= 5; // Unknown or unavailable charset errarg= (int) strpos[14]; - goto err; + goto free_and_err; } } @@ -1247,7 +1250,7 @@ static int open_binary_frm(THD *thd, TAB if ((uint)vcol_screen_pos[0] != 1) { error= 4; - goto err; + goto free_and_err; } field_type= (enum_field_types) (uchar) vcol_screen_pos[1]; fld_stored_in_db= (bool) (uint) vcol_screen_pos[2]; @@ -1256,7 +1259,7 @@ static int open_binary_frm(THD *thd, TAB (char *)memdup_root(&share->mem_root, vcol_screen_pos+(uint)FRM_VCOL_HEADER_SIZE, vcol_expr_length))) - goto err; + goto free_and_err; vcol_info->expr_str.length= vcol_expr_length; vcol_screen_pos+= vcol_info_length; share->vfields++; @@ -1346,7 +1349,7 @@ static int open_binary_frm(THD *thd, TAB if (!reg_field) // Not supported field type { error= 4; - goto err; /* purecov: inspected */ + goto free_and_err; /* purecov: inspected */ } reg_field->field_index= i; @@ -1385,7 +1388,7 @@ static int open_binary_frm(THD *thd, TAB sent (OOM). */ error= 8; - goto err; + goto free_and_err; } } if (!reg_field->stored_in_db) @@ -1462,7 +1465,7 @@ static int open_binary_frm(THD *thd, TAB if (!key_part->fieldnr) { error= 4; // Wrong file - goto err; + goto free_and_err; } field= key_part->field= share->field[key_part->fieldnr-1]; key_part->type= field->key_type(); @@ -1627,6 +1630,15 @@ static int open_binary_frm(THD *thd, TAB null_length, 255); } + if (share->db_create_options & HA_OPTION_TEXT_CREATE_OPTIONS) + { + DBUG_ASSERT(options_len); + if (engine_table_options_frm_read(options, options_len, share) || + parse_engine_table_options(thd, handler_file->ht, share)) + goto free_and_err; + } + my_free(buff, MYF(MY_ALLOW_ZERO_PTR)); + if (share->found_next_number_field) { reg_field= *share->found_next_number_field; @@ -1685,6 +1697,8 @@ static int open_binary_frm(THD *thd, TAB #endif DBUG_RETURN (0); + free_and_err: + my_free(buff, MYF(MY_ALLOW_ZERO_PTR)); err: share->error= error; share->open_errno= my_errno; @@ -2883,6 +2897,7 @@ File create_frm(THD *thd, const char *na ulong length; uchar fill[IO_SIZE]; int create_flags= O_RDWR | O_TRUNC; + DBUG_ENTER("create_frm"); if (create_info->options & HA_LEX_CREATE_TMP_TABLE) create_flags|= O_EXCL | O_NOFOLLOW; @@ -2964,7 +2979,7 @@ File create_frm(THD *thd, const char *na { VOID(my_close(file,MYF(0))); VOID(my_delete(name,MYF(0))); - return(-1); + DBUG_RETURN(-1); } } } @@ -2975,7 +2990,7 @@ File create_frm(THD *thd, const char *na else my_error(ER_CANT_CREATE_TABLE,MYF(0),table,my_errno); } - return (file); + DBUG_RETURN(file); } /* create_frm */ @@ -2994,6 +3009,7 @@ void update_create_info_from_table(HA_CR create_info->comment= share->comment; create_info->transactional= share->transactional; create_info->page_checksum= share->page_checksum; + create_info->option_list= share->option_list; DBUG_VOID_RETURN; } === modified file 'sql/table.h' --- a/sql/table.h 2010-02-12 08:47:31 +0000 +++ b/sql/table.h 2010-04-08 12:10:05 +0000 @@ -340,6 +340,8 @@ typedef struct st_table_share #ifdef NOT_YET struct st_table *open_tables; /* link to open tables */ #endif + engine_option_value *option_list; /* text options for table */ + void *option_struct; /* structure with parsed options */ /* The following is copied to each TABLE on OPEN */ Field **field; @@ -953,6 +955,7 @@ enum enum_schema_tables SCH_GLOBAL_STATUS, SCH_GLOBAL_VARIABLES, SCH_INDEX_STATS, + SCH_KEY_CACHES, SCH_KEY_COLUMN_USAGE, SCH_OPEN_TABLES, SCH_PARTITIONS, === modified file 'sql/unireg.cc' --- a/sql/unireg.cc 2010-01-04 17:54:42 +0000 +++ b/sql/unireg.cc 2010-04-08 12:10:05 +0000 @@ -24,11 +24,15 @@ */ #include "mysql_priv.h" +#include "create_options.h" #include <m_ctype.h> #include <assert.h> #define FCOMP 17 /* Bytes for a packed field */ +/* threshold for safe_alloca */ +#define ALLOCA_THRESHOLD 2048 + static uchar * pack_screens(List<Create_field> &create_fields, uint *info_length, uint *screens, bool small_file); static uint pack_keys(uchar *keybuff,uint key_count, KEY *key_info, @@ -107,6 +111,7 @@ bool mysql_create_frm(THD *thd, const ch ulong key_buff_length; File file; ulong filepos, data_offset; + uint options_len; uchar fileinfo[64],forminfo[288],*keybuff; TYPELIB formnames; uchar *screen_buff; @@ -183,6 +188,16 @@ bool mysql_create_frm(THD *thd, const ch create_info->extra_size+= key_info[i].parser_name->length + 1; } + options_len= engine_table_options_frm_length(create_info->option_list, + create_fields, + keys, key_info); + DBUG_PRINT("info", ("Options length: %u", options_len)); + if (options_len) + { + create_info->table_options|= HA_OPTION_TEXT_CREATE_OPTIONS; + create_info->extra_size+= (options_len + 4); + } + if ((file=create_frm(thd, file_name, db, table, reclength, fileinfo, create_info, keys)) < 0) { @@ -294,6 +309,7 @@ bool mysql_create_frm(THD *thd, const ch if (my_write(file, (uchar*) buff, 6, MYF_RW)) goto err; } + for (i= 0; i < keys; i++) { if (key_info[i].parser_name) @@ -304,6 +320,24 @@ bool mysql_create_frm(THD *thd, const ch } } + if (options_len) + { + uchar *optbuff= (uchar *)my_safe_alloca(options_len + 4, ALLOCA_THRESHOLD); + my_bool error; + DBUG_PRINT("info", ("Create options length: %u", options_len)); + if (!optbuff) + goto err; + int4store(optbuff, options_len); + engine_table_options_frm_image(optbuff + 4, + create_info->option_list, + create_fields, + keys, key_info); + error= my_write(file, optbuff, options_len + 4, MYF_RW); + my_safe_afree(optbuff, options_len + 4, ALLOCA_THRESHOLD); + if (error) + goto err; + } + VOID(my_seek(file,filepos,MY_SEEK_SET,MYF(0))); if (my_write(file, forminfo, 288, MYF_RW) || my_write(file, screen_buff, info_length, MYF_RW) || === modified file 'storage/archive/ha_archive.cc' --- a/storage/archive/ha_archive.cc 2010-03-04 08:03:07 +0000 +++ b/storage/archive/ha_archive.cc 2010-04-01 14:34:51 +0000 @@ -1655,4 +1655,21 @@ mysql_declare_plugin(archive) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(archive) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &archive_storage_engine, + "ARCHIVE", + "Brian Aker, MySQL AB", + "Archive storage engine", + PLUGIN_LICENSE_GPL, + archive_db_init, /* Plugin Init */ + archive_db_done, /* Plugin Deinit */ + 0x0300 /* 3.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/blackhole/ha_blackhole.cc' --- a/storage/blackhole/ha_blackhole.cc 2009-10-16 10:29:42 +0000 +++ b/storage/blackhole/ha_blackhole.cc 2010-04-01 14:34:51 +0000 @@ -369,3 +369,20 @@ mysql_declare_plugin(blackhole) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(blackhole) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &blackhole_storage_engine, + "BLACKHOLE", + "MySQL AB", + "/dev/null storage engine (anything you write to it disappears)", + PLUGIN_LICENSE_GPL, + blackhole_init, /* Plugin Init */ + blackhole_fini, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/csv/ha_tina.cc' --- a/storage/csv/ha_tina.cc 2009-12-03 11:19:05 +0000 +++ b/storage/csv/ha_tina.cc 2010-04-01 14:34:51 +0000 @@ -1636,4 +1636,20 @@ mysql_declare_plugin(csv) NULL /* config options */ } mysql_declare_plugin_end; - +maria_declare_plugin(csv) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &csv_storage_engine, + "CSV", + "Brian Aker, MySQL AB", + "CSV storage engine", + PLUGIN_LICENSE_GPL, + tina_init_func, /* Plugin Init */ + tina_done_func, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/example/ha_example.cc' --- a/storage/example/ha_example.cc 2010-03-03 14:44:14 +0000 +++ b/storage/example/ha_example.cc 2010-04-08 12:10:05 +0000 @@ -113,6 +113,76 @@ static HASH example_open_tables; /* The mutex used to init the hash; variable for example share methods */ pthread_mutex_t example_mutex; + +/** + structure for CREATE TABLE options (table options) + + These can be specified in the CREATE TABLE: + CREATE TABLE ( ... ) {...here...} +*/ + +struct example_table_options_struct +{ + const char *strparam; + ulonglong ullparam; + uint enumparam; + bool boolparam; +}; + + +/** + structure for CREATE TABLE options (field options) + + These can be specified in the CREATE TABLE per field: + CREATE TABLE ( field ... {...here...}, ... ) +*/ + +struct example_field_options_struct +{ + const char *compex_param_to_parse_it_in_engine; +}; + +/* HA_TOPTION_* macros expect the structure called ha_table_option_struct */ +#define ha_table_option_struct example_table_options_struct +ha_create_table_option example_table_option_list[]= +{ + /* + one numeric option, with the default of UINT_MAX32, valid + range of values 0..UINT_MAX32, and a "block size" of 10 + (any value must be divisible by 10). + */ + HA_TOPTION_ULL("ULL", ullparam, UINT_MAX32, 0, UINT_MAX32, 10), + /* + one option that takes an arbitrary string + */ + HA_TOPTION_STRING("STR", strparam), + /* + one enum option. a valid values are strings ONE and TWO. + A default value is 0, that is "one". + */ + HA_TOPTION_ENUM("one_or_two", enumparam, "one,two", 0), + /* + one boolean option, the valid values are YES/NO, ON/OFF, 1/0. + The default is 1, that is true, yes, on. + */ + HA_TOPTION_BOOL("YESNO", boolparam, 1), + HA_TOPTION_END +}; + +/* HA_FOPTION_* macros expect the structure called ha_field_option_struct */ +#define ha_field_option_struct example_field_options_struct +ha_create_table_option example_field_option_list[]= +{ + /* + If the engine wants something more complex than a string, number, enum, + or boolean - for example a list - it needs to specify the option + as a string and parse it internally. + */ + HA_FOPTION_STRING("COMPLEX", compex_param_to_parse_it_in_engine), + HA_FOPTION_END +}; + + /** @brief Function we use in the creation of our hash to get key. @@ -138,6 +208,8 @@ static int example_init_func(void *p) example_hton->state= SHOW_OPTION_YES; example_hton->create= example_create_handler; example_hton->flags= HTON_CAN_RECREATE; + example_hton->table_options= example_table_option_list; + example_hton->field_options= example_field_option_list; DBUG_RETURN(0); } @@ -774,27 +846,6 @@ int ha_example::delete_table(const char /** @brief - Renames a table from one name to another via an alter table call. - - @details - If you do not implement this, the default rename_table() is called from - handler.cc and it will delete all files with the file extensions returned - by bas_ext(). - - Called from sql_table.cc by mysql_rename_table(). - - @see - mysql_rename_table() in sql_table.cc -*/ -int ha_example::rename_table(const char * from, const char * to) -{ - DBUG_ENTER("ha_example::rename_table "); - DBUG_RETURN(HA_ERR_WRONG_COMMAND); -} - - -/** - @brief Given a starting key and an ending key, estimate the number of rows that will exist between the two keys. @@ -836,15 +887,105 @@ ha_rows ha_example::records_in_range(uin int ha_example::create(const char *name, TABLE *table_arg, HA_CREATE_INFO *create_info) { + example_table_options_struct *options= + (example_table_options_struct *)table_arg->s->option_struct; DBUG_ENTER("ha_example::create"); /* - This is not implemented but we want someone to be able to see that it - works. + This example shows how to support custom engine specific table and field + options. */ + DBUG_ASSERT(options); + DBUG_PRINT("info", ("strparam: '%-.64s' ullparam: %llu enumparam: %u "\ + "boolparam: %u", + (options->strparam ? options->strparam : "<NULL>"), + options->ullparam, options->enumparam, options->boolparam)); + for (Field **field= table_arg->s->field; *field; field++) + { + example_field_options_struct *field_options= + (example_field_options_struct *)(*field)->option_struct; + DBUG_ASSERT(field_options); + DBUG_PRINT("info", ("field: %s complex: '%-.64s'", + (*field)->field_name, + (field_options->compex_param_to_parse_it_in_engine ? + field_options->compex_param_to_parse_it_in_engine : + "<NULL>"))); + + } + DBUG_RETURN(0); } +/** + check_if_incompatible_data() called if ALTER TABLE can't detect otherwise + if new and old definition are compatible + + @details If there are no other explicit signs like changed number of + fields this function will be called by compare_tables() + (sql/sql_tables.cc) to decide should we rewrite whole table or only .frm + file. + +*/ + +bool ha_example::check_if_incompatible_data(HA_CREATE_INFO *info, + uint table_changes) +{ + example_table_options_struct *param_old, *param_new; + uint i; + DBUG_ENTER("ha_example::check_if_incompatible_data"); + /* + This example shows how custom engine specific table and field + options can be accessed from this function to be compared. + */ + param_new= (example_table_options_struct *)info->option_struct; + DBUG_PRINT("info", ("new strparam: '%-.64s' ullparam: %llu enumparam: %u " + "boolparam: %u", + (param_new->strparam ? param_new->strparam : "<NULL>"), + param_new->ullparam, param_new->enumparam, + param_new->boolparam)); + + param_old= (example_table_options_struct *)table->s->option_struct; + DBUG_PRINT("info", ("old strparam: '%-.64s' ullparam: %llu enumparam: %u " + "boolparam: %u", + (param_old->strparam ? param_old->strparam : "<NULL>"), + param_old->ullparam, param_old->enumparam, + param_old->boolparam)); + + /* + check important parameters: + for this example engine, we'll assume that changing ullparam or + boolparam requires a table to be rebuilt, while changing strparam + or enumparam - does not. + */ + if (param_new->ullparam != param_old->ullparam || + param_new->boolparam != param_old->boolparam) + DBUG_RETURN(COMPATIBLE_DATA_NO); + + for (i= 0; i < table->s->fields; i++) + { + example_field_options_struct *f_old, *f_new; + f_old= (example_field_options_struct *)table->s->field[i]->option_struct; + DBUG_ASSERT(f_old); + DBUG_PRINT("info", ("old field: %u old complex: '%-.64s'", i, + (f_old->compex_param_to_parse_it_in_engine ? + f_old->compex_param_to_parse_it_in_engine : + "<NULL>"))); + if (info->fileds_option_struct[i]) + { + f_new= (example_field_options_struct *)info->fileds_option_struct[i]; + DBUG_PRINT("info", ("old field: %u new complex: '%-.64s'", i, + (f_new->compex_param_to_parse_it_in_engine ? + f_new->compex_param_to_parse_it_in_engine : + "<NULL>"))); + } + else + DBUG_PRINT("info", ("old field %i did not changed", i)); + } + + DBUG_RETURN(COMPATIBLE_DATA_YES); +} + + struct st_mysql_storage_engine example_storage_engine= { MYSQL_HANDLERTON_INTERFACE_VERSION }; @@ -924,3 +1065,20 @@ mysql_declare_plugin(example) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(example) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &example_storage_engine, + "EXAMPLE", + "Brian Aker, MySQL AB", + "Example storage engine", + PLUGIN_LICENSE_GPL, + example_init_func, /* Plugin Init */ + example_done_func, /* Plugin Deinit */ + 0x0001 /* 0.1 */, + func_status, /* status variables */ + example_system_variables, /* system variables */ + "0.1", /* string version */ + MariaDB_PLUGIN_MATURITY_EXPERIMENTAL /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/example/ha_example.h' --- a/storage/example/ha_example.h 2010-03-28 18:10:00 +0000 +++ b/storage/example/ha_example.h 2010-04-08 12:10:05 +0000 @@ -243,9 +243,10 @@ public: ha_rows records_in_range(uint inx, key_range *min_key, key_range *max_key); int delete_table(const char *from); - int rename_table(const char * from, const char * to); int create(const char *name, TABLE *form, HA_CREATE_INFO *create_info); ///< required + bool check_if_incompatible_data(HA_CREATE_INFO *info, + uint table_changes); THR_LOCK_DATA **store_lock(THD *thd, THR_LOCK_DATA **to, enum thr_lock_type lock_type); ///< required === modified file 'storage/federated/ha_federated.cc' --- a/storage/federated/ha_federated.cc 2009-09-07 20:50:10 +0000 +++ b/storage/federated/ha_federated.cc 2010-04-01 14:34:51 +0000 @@ -3379,3 +3379,20 @@ mysql_declare_plugin(federated) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(federated) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &federated_storage_engine, + "FEDERATED", + "Patrick Galbraith and Brian Aker, MySQL AB", + "Federated MySQL storage engine", + PLUGIN_LICENSE_GPL, + federated_db_init, /* Plugin Init */ + federated_done, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_BETA /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/federatedx/ha_federatedx.cc' --- a/storage/federatedx/ha_federatedx.cc 2010-03-10 10:32:14 +0000 +++ b/storage/federatedx/ha_federatedx.cc 2010-04-01 14:34:51 +0000 @@ -3505,9 +3505,26 @@ mysql_declare_plugin(federated) PLUGIN_LICENSE_GPL, federatedx_db_init, /* Plugin Init */ federatedx_done, /* Plugin Deinit */ - 0x0100 /* 1.0 */, + 0x0200 /* 2.0 */, NULL, /* status variables */ NULL, /* system variables */ NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(federated) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &federatedx_storage_engine, + "FEDERATED", + "Patrick Galbraith", + "FederatedX pluggable storage engine", + PLUGIN_LICENSE_GPL, + federatedx_db_init, /* Plugin Init */ + federatedx_done, /* Plugin Deinit */ + 0x0200 /* 2.0 */, + NULL, /* status variables */ + NULL, /* system variables */ + "2.0", /* string version */ + MariaDB_PLUGIN_MATURITY_BETA /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/heap/ha_heap.cc' --- a/storage/heap/ha_heap.cc 2009-09-07 20:50:10 +0000 +++ b/storage/heap/ha_heap.cc 2010-04-01 14:34:51 +0000 @@ -767,3 +767,20 @@ mysql_declare_plugin(heap) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(heap) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &heap_storage_engine, + "MEMORY", + "MySQL AB", + "Hash based, stored in memory, useful for temporary tables", + PLUGIN_LICENSE_GPL, + heap_init, + NULL, + 0x0100, /* 1.0 */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/ibmdb2i/ha_ibmdb2i.cc' --- a/storage/ibmdb2i/ha_ibmdb2i.cc 2009-12-11 07:01:16 +0000 +++ b/storage/ibmdb2i/ha_ibmdb2i.cc 2010-04-01 14:34:51 +0000 @@ -3357,3 +3357,20 @@ mysql_declare_plugin(ibmdb2i) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(ibmdb2i) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &ibmdb2i_storage_engine, + "IBMDB2I", + "The IBM development team in Rochester, Minnesota", + "IBM DB2 for i Storage Engine", + PLUGIN_LICENSE_GPL, + ibmdb2i_init_func, /* Plugin Init */ + ibmdb2i_done_func, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + NULL, /* status variables */ + ibmdb2i_system_variables, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_UNKNOWN /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/innobase/handler/ha_innodb.cc' --- a/storage/innobase/handler/ha_innodb.cc 2010-03-15 11:51:23 +0000 +++ b/storage/innobase/handler/ha_innodb.cc 2010-04-01 14:34:51 +0000 @@ -8930,6 +8930,23 @@ mysql_declare_plugin(innobase) NULL /* reserved */ } mysql_declare_plugin_end; +maria_declare_plugin(innobase) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &innobase_storage_engine, + innobase_hton_name, + "Innobase OY", + "Supports transactions, row-level locking, and foreign keys", + PLUGIN_LICENSE_GPL, + innobase_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100 /* 1.0 */, + innodb_status_variables_export, /* status variables */ + innobase_system_variables, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE, /* maturity */ +} +maria_declare_plugin_end; /** @brief Initialize the default value of innodb_commit_concurrency. === modified file 'storage/innodb_plugin/handler/i_s.cc' --- a/storage/innodb_plugin/handler/i_s.cc 2009-10-12 12:00:56 +0000 +++ b/storage/innodb_plugin/handler/i_s.cc 2010-04-01 14:34:51 +0000 @@ -455,6 +455,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_trx_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_TRX"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB transactions"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_trx_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table INFORMATION_SCHEMA.innodb_locks */ static ST_FIELD_INFO innodb_locks_fields_info[] = { @@ -730,6 +783,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_locks_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_LOCKS"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB conflicting locks"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_locks_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table INFORMATION_SCHEMA.innodb_lock_waits */ static ST_FIELD_INFO innodb_lock_waits_fields_info[] = { @@ -913,6 +1019,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_lock_waits_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_LOCK_WAITS"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, "Innobase Oy"), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB which lock is blocking which"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_lock_waits_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*******************************************************************//** Common function to fill any of the dynamic tables: INFORMATION_SCHEMA.innodb_trx @@ -1245,6 +1404,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_mysql_plugin i_s_innodb_cmp_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMP"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compression"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmp_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_cmp_reset = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -1295,6 +1507,60 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmp_reset_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMP_RESET"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compression;" + " reset cumulated counts"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmp_reset_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table information_schema.innodb_cmpmem. */ static ST_FIELD_INFO i_s_cmpmem_fields_info[] = { @@ -1511,6 +1777,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmpmem_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMPMEM"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compressed buffer pool"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmpmem_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_cmpmem_reset = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -1561,6 +1880,60 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmpmem_reset_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMPMEM_RESET"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compressed buffer pool;" + " reset cumulated counts"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmpmem_reset_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*******************************************************************//** Unbind a dynamic INFORMATION_SCHEMA table. @return 0 on success */ === modified file 'storage/maria/ha_maria.cc' --- a/storage/maria/ha_maria.cc 2010-03-15 11:51:23 +0000 +++ b/storage/maria/ha_maria.cc 2010-04-01 14:34:51 +0000 @@ -3471,9 +3471,26 @@ mysql_declare_plugin(maria) PLUGIN_LICENSE_GPL, ha_maria_init, /* Plugin Init */ NULL, /* Plugin Deinit */ - 0x0100, /* 1.0 */ + 0x0105, /* 1.5 */ status_variables, /* status variables */ system_variables, /* system variables */ NULL } mysql_declare_plugin_end; +maria_declare_plugin(maria) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &maria_storage_engine, + "MARIA", + "MySQL AB", + "Crash-safe tables with MyISAM heritage", + PLUGIN_LICENSE_GPL, + ha_maria_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0105, /* 1.5 */ + status_variables, /* status variables */ + system_variables, /* system variables */ + "1.5", /* string version */ + MariaDB_PLUGIN_MATURITY_GAMMA /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/maria/ma_loghandler.c' --- a/storage/maria/ma_loghandler.c 2010-03-31 08:17:03 +0000 +++ b/storage/maria/ma_loghandler.c 2010-04-09 09:26:17 +0000 @@ -1274,10 +1274,10 @@ static my_bool translog_set_lsn_for_file for (file= from_file; file <= to_file; file++) { LOGHANDLER_FILE_INFO info; + File fd; LINT_INIT(info.max_lsn); - File fd= open_logfile_by_number_no_cache(file); - LINT_INIT_STRUCT(info); + fd= open_logfile_by_number_no_cache(file); if ((fd < 0) || ((translog_read_file_header(&info, fd) || (cmp_translog_addr(lsn, info.max_lsn) > 0 && @@ -3969,7 +3969,7 @@ my_bool translog_init_with_table(const c if (!old_log_was_recovered && old_flags == flags) { LOGHANDLER_FILE_INFO info; - LINT_INIT(info.maria_version); + LINT_INIT_STRUCT(info); /* Accessing &log_descriptor.open_files without mutex is safe === modified file 'storage/myisam/ha_myisam.cc' --- a/storage/myisam/ha_myisam.cc 2010-01-04 17:54:42 +0000 +++ b/storage/myisam/ha_myisam.cc 2010-04-01 14:34:51 +0000 @@ -2174,6 +2174,23 @@ mysql_declare_plugin(myisam) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(myisam) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &myisam_storage_engine, + "MyISAM", + "MySQL AB", + "Default engine as of MySQL 3.23 with great performance", + PLUGIN_LICENSE_GPL, + myisam_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100, /* 1.0 */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; #ifdef HAVE_QUERY_CACHE === modified file 'storage/myisam/mi_check.c' --- a/storage/myisam/mi_check.c 2010-01-14 16:51:00 +0000 +++ b/storage/myisam/mi_check.c 2010-02-16 16:41:11 +0000 @@ -332,7 +332,8 @@ int chk_size(HA_CHECK *param, register M /* The following is needed if called externally (not from myisamchk) */ flush_key_blocks(info->s->key_cache, - info->s->kfile, FLUSH_FORCE_WRITE); + info->s->kfile, &info->s->dirty_part_map, + FLUSH_FORCE_WRITE); size= my_seek(info->s->kfile, 0L, MY_SEEK_END, MYF(MY_THREADSAFE)); if ((skr=(my_off_t) info->state->key_file_length) != size) @@ -1474,6 +1475,7 @@ static int mi_drop_all_indexes(HA_CHECK */ DBUG_PRINT("repair", ("all disabled are empty: create missing")); error= flush_key_blocks(share->key_cache, share->kfile, + &share->dirty_part_map, FLUSH_FORCE_WRITE); goto end; } @@ -1488,6 +1490,7 @@ static int mi_drop_all_indexes(HA_CHECK /* Remove all key blocks of this index file from key cache. */ if ((error= flush_key_blocks(share->key_cache, share->kfile, + &share->dirty_part_map, FLUSH_IGNORE_CHANGED))) goto end; /* purecov: inspected */ @@ -1549,7 +1552,7 @@ int mi_repair(HA_CHECK *param, register if (!param->using_global_keycache) VOID(init_key_cache(dflt_key_cache, param->key_cache_block_size, - (size_t) param->use_buffers, 0, 0)); + (size_t) param->use_buffers, 0, 0, 0)); if (init_io_cache(&param->read_cache,info->dfile, (uint) param->read_buffer_length, @@ -1762,7 +1765,8 @@ err: VOID(end_io_cache(&param->read_cache)); info->opt_flag&= ~(READ_CACHE_USED | WRITE_CACHE_USED); VOID(end_io_cache(&info->rec_cache)); - got_error|=flush_blocks(param, share->key_cache, share->kfile); + got_error|=flush_blocks(param, share->key_cache, share->kfile, + &share->dirty_part_map); if (!got_error && param->testflag & T_UNPACK) { share->state.header.options[0]&= (uchar) ~HA_OPTION_COMPRESS_RECORD; @@ -1908,9 +1912,10 @@ void lock_memory(HA_CHECK *param __attri /* Flush all changed blocks to disk */ -int flush_blocks(HA_CHECK *param, KEY_CACHE *key_cache, File file) +int flush_blocks(HA_CHECK *param, KEY_CACHE *key_cache, File file, + ulonglong *dirty_part_map) { - if (flush_key_blocks(key_cache, file, FLUSH_RELEASE)) + if (flush_key_blocks(key_cache, file, dirty_part_map, FLUSH_RELEASE)) { mi_check_print_error(param,"%d when trying to write bufferts",my_errno); return(1); @@ -1977,7 +1982,8 @@ int mi_sort_index(HA_CHECK *param, regis } /* Flush key cache for this file if we are calling this outside myisamchk */ - flush_key_blocks(share->key_cache,share->kfile, FLUSH_IGNORE_CHANGED); + flush_key_blocks(share->key_cache, share->kfile, &share->dirty_part_map, + FLUSH_IGNORE_CHANGED); share->state.version=(ulong) time((time_t*) 0); old_state= share->state; /* save state if not stored */ @@ -2535,7 +2541,8 @@ int mi_repair_by_sort(HA_CHECK *param, r memcpy( &share->state.state, info->state, sizeof(*info->state)); err: - got_error|= flush_blocks(param, share->key_cache, share->kfile); + got_error|= flush_blocks(param, share->key_cache, share->kfile, + &share->dirty_part_map); VOID(end_io_cache(&info->rec_cache)); if (!got_error) { @@ -3059,7 +3066,8 @@ int mi_repair_parallel(HA_CHECK *param, memcpy(&share->state.state, info->state, sizeof(*info->state)); err: - got_error|= flush_blocks(param, share->key_cache, share->kfile); + got_error|= flush_blocks(param, share->key_cache, share->kfile, + &share->dirty_part_map); /* Destroy the write cache. The master thread did already detach from the share by remove_io_thread() or it was not yet started (if the === modified file 'storage/myisam/mi_close.c' --- a/storage/myisam/mi_close.c 2010-02-10 19:06:24 +0000 +++ b/storage/myisam/mi_close.c 2010-02-16 16:41:11 +0000 @@ -64,6 +64,7 @@ int mi_close(register MI_INFO *info) if (share->kfile >= 0) abort();); if (share->kfile >= 0 && flush_key_blocks(share->key_cache, share->kfile, + &share->dirty_part_map, ((share->temporary || share->deleting) ? FLUSH_IGNORE_CHANGED : FLUSH_RELEASE))) === modified file 'storage/myisam/mi_delete_all.c' --- a/storage/myisam/mi_delete_all.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/mi_delete_all.c 2010-02-16 16:41:11 +0000 @@ -52,7 +52,8 @@ int mi_delete_all_rows(MI_INFO *info) If we are using delayed keys or if the user has done changes to the tables since it was locked then there may be key blocks in the key cache */ - flush_key_blocks(share->key_cache, share->kfile, FLUSH_IGNORE_CHANGED); + flush_key_blocks(share->key_cache, share->kfile, &share->dirty_part_map, + FLUSH_IGNORE_CHANGED); #ifdef HAVE_MMAP if (share->file_map) _mi_unmap_file(info); === modified file 'storage/myisam/mi_extra.c' --- a/storage/myisam/mi_extra.c 2010-02-10 19:06:24 +0000 +++ b/storage/myisam/mi_extra.c 2010-02-16 16:41:11 +0000 @@ -268,6 +268,7 @@ int mi_extra(MI_INFO *info, enum ha_extr pthread_mutex_lock(&share->intern_lock); /* Flush pages that we don't need anymore */ if (flush_key_blocks(share->key_cache, share->kfile, + &share->dirty_part_map, (function == HA_EXTRA_PREPARE_FOR_DROP ? FLUSH_IGNORE_CHANGED : FLUSH_RELEASE))) { @@ -326,7 +327,8 @@ int mi_extra(MI_INFO *info, enum ha_extr break; case HA_EXTRA_FLUSH: if (!share->temporary) - flush_key_blocks(share->key_cache, share->kfile, FLUSH_KEEP); + flush_key_blocks(share->key_cache, share->kfile, &share->dirty_part_map, + FLUSH_KEEP); #ifdef HAVE_PWRITE _mi_decrement_open_count(info); #endif === modified file 'storage/myisam/mi_keycache.c' --- a/storage/myisam/mi_keycache.c 2008-03-29 15:56:33 +0000 +++ b/storage/myisam/mi_keycache.c 2010-02-16 16:41:11 +0000 @@ -75,7 +75,8 @@ int mi_assign_to_key_cache(MI_INFO *info in the old key cache. */ - if (flush_key_blocks(share->key_cache, share->kfile, FLUSH_RELEASE)) + if (flush_key_blocks(share->key_cache, share->kfile, &share->dirty_part_map, + FLUSH_RELEASE)) { error= my_errno; mi_print_error(info->s, HA_ERR_CRASHED); @@ -90,7 +91,8 @@ int mi_assign_to_key_cache(MI_INFO *info (This can never fail as there is never any not written data in the new key cache) */ - (void) flush_key_blocks(key_cache, share->kfile, FLUSH_RELEASE); + (void) flush_key_blocks(key_cache, share->kfile, &share->dirty_part_map, + FLUSH_RELEASE); /* ensure that setting the key cache and changing the multi_key_cache @@ -102,6 +104,7 @@ int mi_assign_to_key_cache(MI_INFO *info This should be seen at the lastes for the next call to an myisam function. */ share->key_cache= key_cache; + share->dirty_part_map= 0; /* store the key cache in the global hash structure for future opens */ if (multi_key_cache_set((uchar*) share->unique_file_name, === modified file 'storage/myisam/mi_locking.c' --- a/storage/myisam/mi_locking.c 2009-10-06 06:57:22 +0000 +++ b/storage/myisam/mi_locking.c 2010-02-16 16:41:11 +0000 @@ -68,7 +68,9 @@ int mi_lock_database(MI_INFO *info, int --share->tot_locks; if (info->lock_type == F_WRLCK && !share->w_locks && !share->delay_key_write && flush_key_blocks(share->key_cache, - share->kfile,FLUSH_KEEP)) + share->kfile, + &share->dirty_part_map, + FLUSH_KEEP)) { error=my_errno; mi_print_error(info->s, HA_ERR_CRASHED); @@ -513,7 +515,8 @@ int _mi_test_if_changed(register MI_INFO { /* Keyfile has changed */ DBUG_PRINT("info",("index file changed")); if (share->state.process != share->this_process) - VOID(flush_key_blocks(share->key_cache, share->kfile, FLUSH_RELEASE)); + VOID(flush_key_blocks(share->key_cache, share->kfile, + &share->dirty_part_map, FLUSH_RELEASE)); share->last_process=share->state.process; info->last_unique= share->state.unique; info->last_loop= share->state.update_count; === modified file 'storage/myisam/mi_page.c' --- a/storage/myisam/mi_page.c 2009-05-06 12:03:24 +0000 +++ b/storage/myisam/mi_page.c 2010-02-16 16:41:11 +0000 @@ -94,10 +94,11 @@ int _mi_write_keypage(register MI_INFO * } #endif DBUG_RETURN((key_cache_write(info->s->key_cache, - info->s->kfile,page, level, (uchar*) buff,length, - (uint) keyinfo->block_length, - (int) ((info->lock_type != F_UNLCK) || - info->s->delay_key_write)))); + info->s->kfile, &info->s->dirty_part_map, + page, level, (uchar*) buff, length, + (uint) keyinfo->block_length, + (int) ((info->lock_type != F_UNLCK) || + info->s->delay_key_write)))); } /* mi_write_keypage */ @@ -116,7 +117,8 @@ int _mi_dispose(register MI_INFO *info, mi_sizestore(buff,old_link); info->s->state.changed|= STATE_NOT_SORTED_PAGES; DBUG_RETURN(key_cache_write(info->s->key_cache, - info->s->kfile, pos , level, buff, + info->s->kfile, &info->s->dirty_part_map, + pos , level, buff, sizeof(buff), (uint) keyinfo->block_length, (int) (info->lock_type != F_UNLCK))); === modified file 'storage/myisam/mi_panic.c' --- a/storage/myisam/mi_panic.c 2006-12-31 00:32:21 +0000 +++ b/storage/myisam/mi_panic.c 2010-02-16 16:41:11 +0000 @@ -47,7 +47,8 @@ int mi_panic(enum ha_panic_function flag if (info->s->options & HA_OPTION_READ_ONLY_DATA) break; #endif - if (flush_key_blocks(info->s->key_cache, info->s->kfile, FLUSH_RELEASE)) + if (flush_key_blocks(info->s->key_cache, info->s->kfile, + &info->s->dirty_part_map, FLUSH_RELEASE)) error=my_errno; if (info->opt_flag & WRITE_CACHE_USED) if (flush_io_cache(&info->rec_cache)) === modified file 'storage/myisam/mi_preload.c' --- a/storage/myisam/mi_preload.c 2009-09-09 15:13:13 +0000 +++ b/storage/myisam/mi_preload.c 2010-02-16 16:41:11 +0000 @@ -65,7 +65,7 @@ int mi_preload(MI_INFO *info, ulonglong } } else - block_length= share->key_cache->key_cache_block_size; + block_length= share->key_cache->param_block_size; length= info->preload_buff_size/block_length * block_length; set_if_bigger(length, block_length); @@ -73,7 +73,8 @@ int mi_preload(MI_INFO *info, ulonglong if (!(buff= (uchar *) my_malloc(length, MYF(MY_WME)))) DBUG_RETURN(my_errno= HA_ERR_OUT_OF_MEM); - if (flush_key_blocks(share->key_cache,share->kfile, FLUSH_RELEASE)) + if (flush_key_blocks(share->key_cache, share->kfile, &share->dirty_part_map, + FLUSH_RELEASE)) goto err; do === modified file 'storage/myisam/mi_test1.c' --- a/storage/myisam/mi_test1.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/mi_test1.c 2010-02-16 16:41:11 +0000 @@ -49,7 +49,8 @@ int main(int argc,char *argv[]) MY_INIT(argv[0]); my_init(); if (key_cacheing) - init_key_cache(dflt_key_cache,KEY_CACHE_BLOCK_SIZE,IO_SIZE*16,0,0); + init_key_cache(dflt_key_cache,KEY_CACHE_BLOCK_SIZE,IO_SIZE*16,0,0, + DEFAULT_KEY_CACHE_PARTITIONS); get_options(argc,argv); exit(run_test("test1")); === modified file 'storage/myisam/mi_test2.c' --- a/storage/myisam/mi_test2.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/mi_test2.c 2010-02-16 16:41:11 +0000 @@ -215,7 +215,8 @@ int main(int argc, char *argv[]) if (!silent) printf("- Writing key:s\n"); if (key_cacheing) - init_key_cache(dflt_key_cache,key_cache_block_size,key_cache_size,0,0); + init_key_cache(dflt_key_cache,key_cache_block_size,key_cache_size,0,0, + DEFAULT_KEY_CACHE_PARTITIONS); if (do_locking) mi_lock_database(file,F_WRLCK); if (write_cacheing) === modified file 'storage/myisam/mi_test3.c' --- a/storage/myisam/mi_test3.c 2010-01-06 21:27:53 +0000 +++ b/storage/myisam/mi_test3.c 2010-02-16 16:41:11 +0000 @@ -177,7 +177,8 @@ void start_test(int id) exit(1); } if (key_cacheing && rnd(2) == 0) - init_key_cache(dflt_key_cache, KEY_CACHE_BLOCK_SIZE, 65536L, 0, 0); + init_key_cache(dflt_key_cache, KEY_CACHE_BLOCK_SIZE, 65536L, 0, 0, + DEFAULT_KEY_CACHE_PARTITIONS); printf("Process %d, pid: %ld\n", id, (long) getpid()); fflush(stdout); === modified file 'storage/myisam/myisam_ftdump.c' --- a/storage/myisam/myisam_ftdump.c 2009-11-30 13:36:06 +0000 +++ b/storage/myisam/myisam_ftdump.c 2010-02-16 16:41:11 +0000 @@ -83,7 +83,7 @@ int main(int argc,char *argv[]) usage(); } - init_key_cache(dflt_key_cache,MI_KEY_BLOCK_LENGTH,USE_BUFFER_INIT, 0, 0); + init_key_cache(dflt_key_cache,MI_KEY_BLOCK_LENGTH,USE_BUFFER_INIT, 0, 0, 0); if (!(info=mi_open(argv[0], O_RDONLY, HA_OPEN_ABORT_IF_LOCKED|HA_OPEN_FROM_SQL_LAYER))) === modified file 'storage/myisam/myisamchk.c' --- a/storage/myisam/myisamchk.c 2009-12-03 11:34:11 +0000 +++ b/storage/myisam/myisamchk.c 2010-02-16 16:41:11 +0000 @@ -1102,7 +1102,7 @@ static int myisamchk(HA_CHECK *param, ch { if (param->testflag & (T_EXTEND | T_MEDIUM)) VOID(init_key_cache(dflt_key_cache,opt_key_cache_block_size, - (size_t) param->use_buffers, 0, 0)); + (size_t) param->use_buffers, 0, 0, 0)); VOID(init_io_cache(&param->read_cache,datafile, (uint) param->read_buffer_length, READ_CACHE, @@ -1116,7 +1116,8 @@ static int myisamchk(HA_CHECK *param, ch HA_OPTION_COMPRESS_RECORD)) || (param->testflag & (T_EXTEND | T_MEDIUM))) error|=chk_data_link(param, info, test(param->testflag & T_EXTEND)); - error|=flush_blocks(param, share->key_cache, share->kfile); + error|=flush_blocks(param, share->key_cache, share->kfile, + &share->dirty_part_map); VOID(end_io_cache(&param->read_cache)); } if (!error) @@ -1526,7 +1527,7 @@ static int mi_sort_records(HA_CHECK *par DBUG_RETURN(0); /* Nothing to do */ init_key_cache(dflt_key_cache, opt_key_cache_block_size, - (size_t) param->use_buffers, 0, 0); + (size_t) param->use_buffers, 0, 0, 0); if (init_io_cache(&info->rec_cache,-1,(uint) param->write_buffer_length, WRITE_CACHE,share->pack.header_length,1, MYF(MY_WME | MY_WAIT_IF_FULL))) @@ -1641,8 +1642,8 @@ err: my_free(sort_info.buff,MYF(MY_ALLOW_ZERO_PTR)); sort_info.buff=0; share->state.sortkey=sort_key; - DBUG_RETURN(flush_blocks(param, share->key_cache, share->kfile) | - got_error); + DBUG_RETURN(flush_blocks(param, share->key_cache, share->kfile, + &share->dirty_part_map) | got_error); } /* sort_records */ === modified file 'storage/myisam/myisamdef.h' --- a/storage/myisam/myisamdef.h 2010-03-04 08:03:07 +0000 +++ b/storage/myisam/myisamdef.h 2010-03-29 21:16:12 +0000 @@ -174,6 +174,8 @@ typedef struct st_mi_isam_share *index_file_name; uchar *file_map; /* mem-map of file if possible */ KEY_CACHE *key_cache; /* ref to the current key cache */ + /* To mark the key cache partitions containing dirty pages for this file */ + ulonglong dirty_part_map; MI_DECODE_TREE *decode_trees; uint16 *decode_tables; /* Function to use for a row checksum. */ @@ -732,7 +734,8 @@ void mi_check_print_info _VARARGS((HA_CH #ifdef THREAD pthread_handler_t thr_find_all_keys(void *arg); #endif -int flush_blocks(HA_CHECK *param, KEY_CACHE *key_cache, File file); +int flush_blocks(HA_CHECK *param, KEY_CACHE *key_cache, File file, + ulonglong *dirty_part_map); #ifdef __cplusplus } #endif === modified file 'storage/myisam/myisamlog.c' --- a/storage/myisam/myisamlog.c 2009-11-29 23:08:56 +0000 +++ b/storage/myisam/myisamlog.c 2010-02-16 16:41:11 +0000 @@ -333,7 +333,7 @@ static int examine_log(char * file_name, init_tree(&tree,0,0,sizeof(file_info),(qsort_cmp2) file_info_compare,1, (tree_element_free) file_info_free, NULL); VOID(init_key_cache(dflt_key_cache,KEY_CACHE_BLOCK_SIZE,KEY_CACHE_SIZE, - 0, 0)); + 0, 0, 0)); files_open=0; access_time=0; while (access_time++ != number_of_commands && === modified file 'storage/myisammrg/ha_myisammrg.cc' --- a/storage/myisammrg/ha_myisammrg.cc 2010-03-30 12:36:49 +0000 +++ b/storage/myisammrg/ha_myisammrg.cc 2010-04-01 14:34:51 +0000 @@ -1289,3 +1289,20 @@ mysql_declare_plugin(myisammrg) NULL /* config options */ } mysql_declare_plugin_end; +maria_declare_plugin(myisammrg) +{ + MYSQL_STORAGE_ENGINE_PLUGIN, + &myisammrg_storage_engine, + "MRG_MYISAM", + "MySQL AB", + "Collection of identical MyISAM tables", + PLUGIN_LICENSE_GPL, + myisammrg_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + 0x0100, /* 1.0 */ + NULL, /* status variables */ + NULL, /* system variables */ + "1.0", /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +} +maria_declare_plugin_end; === modified file 'storage/pbxt/src/discover_xt.cc' --- a/storage/pbxt/src/discover_xt.cc 2010-02-01 06:14:12 +0000 +++ b/storage/pbxt/src/discover_xt.cc 2010-04-08 12:10:05 +0000 @@ -1623,7 +1623,7 @@ int xt_create_table_frm(handlerton *hton #endif NULL /*default_value*/, NULL /*on_update_value*/, &comment, NULL /*change*/, NULL /*interval_list*/, info->field_charset, 0 /*uint_geom_type*/, - NULL /*vcol_info*/)) + NULL /*vcol_info*/, NULL /* create options */)) #endif goto error; === modified file 'storage/pbxt/src/ha_pbxt.cc' --- a/storage/pbxt/src/ha_pbxt.cc 2010-03-03 14:44:14 +0000 +++ b/storage/pbxt/src/ha_pbxt.cc 2010-04-01 14:34:51 +0000 @@ -1215,8 +1215,6 @@ static int pbxt_init(void *p) THD *thd = NULL; #ifndef DRIZZLED - extern myxt_mutex_t LOCK_plugin; - /* {MYSQL QUIRK} * I have to release this lock for PBXT recovery to * work, because it needs to open .frm files. @@ -5921,6 +5919,40 @@ mysql_declare_plugin(pbxt) drizzle_declare_plugin_end; #else mysql_declare_plugin_end; +#ifdef MARIADB_BASE_VERSION +maria_declare_plugin(pbxt) +{ /* PBXT */ + MYSQL_STORAGE_ENGINE_PLUGIN, + &pbxt_storage_engine, + "PBXT", + "Paul McCullagh, PrimeBase Technologies GmbH", + "High performance, multi-versioning transactional engine", + PLUGIN_LICENSE_GPL, + pbxt_init, /* Plugin Init */ + pbxt_end, /* Plugin Deinit */ + 0x0001 /* 0.1 */, + NULL, /* status variables */ + pbxt_system_variables, /* system variables */ + "1.0.09g RC3", /* string version */ + MariaDB_PLUGIN_MATURITY_GAMMA /* maturity */ +}, +{ /* PBXT_STATISTICS */ + MYSQL_INFORMATION_SCHEMA_PLUGIN, + &pbxt_statitics, + "PBXT_STATISTICS", + "Paul McCullagh, PrimeBase Technologies GmbH", + "PBXT internal system statitics", + PLUGIN_LICENSE_GPL, + pbxt_init_statistics, /* plugin init */ + pbxt_exit_statistics, /* plugin deinit */ + 0x0005, + NULL, /* status variables */ + NULL, /* system variables */ + "1.0.09g RC3", /* string version */ + MariaDB_PLUGIN_MATURITY_GAMMA /* maturity */ +} +maria_declare_plugin_end; +#endif #endif #if defined(XT_WIN) && defined(XT_COREDUMP) === modified file 'storage/xtradb/handler/ha_innodb.cc' --- a/storage/xtradb/handler/ha_innodb.cc 2010-02-01 06:14:12 +0000 +++ b/storage/xtradb/handler/ha_innodb.cc 2010-04-01 14:34:51 +0000 @@ -10818,6 +10818,39 @@ i_s_innodb_index_stats, i_s_innodb_admin_command, i_s_innodb_patches mysql_declare_plugin_end; +maria_declare_plugin(innobase) +{ /* InnoDB */ + MYSQL_STORAGE_ENGINE_PLUGIN, + &innobase_storage_engine, + innobase_hton_name, + "Innobase Oy", + "Supports transactions, row-level locking, and foreign keys", + PLUGIN_LICENSE_GPL, + innobase_init, /* Plugin Init */ + NULL, /* Plugin Deinit */ + INNODB_VERSION_SHORT, + innodb_status_variables_export,/* status variables */ + innobase_system_variables, /* system variables */ + INNODB_VERSION_STR, /* string version */ + MariaDB_PLUGIN_MATURITY_STABLE /* maturity */ +}, +i_s_innodb_rseg_maria, +i_s_innodb_buffer_pool_pages_maria, +i_s_innodb_buffer_pool_pages_index_maria, +i_s_innodb_buffer_pool_pages_blob_maria, +i_s_innodb_trx_maria, +i_s_innodb_locks_maria, +i_s_innodb_lock_waits_maria, +i_s_innodb_cmp_maria, +i_s_innodb_cmp_reset_maria, +i_s_innodb_cmpmem_maria, +i_s_innodb_cmpmem_reset_maria, +i_s_innodb_table_stats_maria, +i_s_innodb_index_stats_maria, +i_s_innodb_admin_command_maria, +i_s_innodb_patches_maria +maria_declare_plugin_end; + /** @brief Initialize the default value of innodb_commit_concurrency. === modified file 'storage/xtradb/handler/i_s.cc' --- a/storage/xtradb/handler/i_s.cc 2010-01-28 11:35:10 +0000 +++ b/storage/xtradb/handler/i_s.cc 2010-04-01 14:34:51 +0000 @@ -393,6 +393,59 @@ UNIV_INTERN struct st_mysql_plugin STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_patches_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "XTRADB_ENHANCEMENTS"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, "Percona"), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Enhancements applied to InnoDB plugin"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_patches_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + static ST_FIELD_INFO i_s_innodb_buffer_pool_pages_fields_info[] = { @@ -1040,6 +1093,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_buffer_pool_pages_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_BUFFER_POOL_PAGES"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB buffer pool pages"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_innodb_buffer_pool_pages_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, 0x0100 /* 1.0 */), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_buffer_pool_pages_index = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -1089,6 +1195,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_buffer_pool_pages_index_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_BUFFER_POOL_PAGES_INDEX"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB buffer pool index pages"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_innodb_buffer_pool_pages_index_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, 0x0100 /* 1.0 */), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_buffer_pool_pages_blob = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -1138,6 +1297,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_buffer_pool_pages_blob_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_BUFFER_POOL_PAGES_BLOB"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB buffer pool blob pages"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_innodb_buffer_pool_pages_blob_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, 0x0100 /* 1.0 */), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table INFORMATION_SCHEMA.innodb_trx */ static ST_FIELD_INFO innodb_trx_fields_info[] = @@ -1373,6 +1585,60 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; + +UNIV_INTERN struct st_maria_plugin i_s_innodb_trx_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_TRX"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB transactions"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_trx_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table INFORMATION_SCHEMA.innodb_locks */ static ST_FIELD_INFO innodb_locks_fields_info[] = { @@ -1648,6 +1914,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_locks_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_LOCKS"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB conflicting locks"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_locks_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /* Fields of the dynamic table INFORMATION_SCHEMA.innodb_lock_waits */ static ST_FIELD_INFO innodb_lock_waits_fields_info[] = { @@ -1831,6 +2150,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_lock_waits_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_LOCK_WAITS"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, "Innobase Oy"), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB which lock is blocking which"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, innodb_lock_waits_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*******************************************************************//** Common function to fill any of the dynamic tables: INFORMATION_SCHEMA.innodb_trx @@ -2163,6 +2535,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmp_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMP"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compression"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmp_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_cmp_reset = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -2213,6 +2638,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmp_reset_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMP_RESET"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compression;" + " reset cumulated counts"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmp_reset_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; /* Fields of the dynamic table information_schema.innodb_cmpmem. */ static ST_FIELD_INFO i_s_cmpmem_fields_info[] = { @@ -2431,6 +2909,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmpmem_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMPMEM"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compressed buffer pool"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmpmem_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_cmpmem_reset = { /* the plugin type (a MYSQL_XXX_PLUGIN value) */ @@ -2481,6 +3012,60 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_cmpmem_reset_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_CMPMEM_RESET"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "Statistics for the InnoDB compressed buffer pool;" + " reset cumulated counts"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_cmpmem_reset_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, INNODB_VERSION_SHORT), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*******************************************************************//** Unbind a dynamic INFORMATION_SCHEMA table. @return 0 on success */ @@ -2660,6 +3245,59 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_rseg_maria = +{ + /* the plugin type (a MYSQL_XXX_PLUGIN value) */ + /* int */ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + + /* pointer to type-specific plugin descriptor */ + /* void* */ + STRUCT_FLD(info, &i_s_info), + + /* plugin name */ + /* const char* */ + STRUCT_FLD(name, "INNODB_RSEG"), + + /* plugin author (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(author, plugin_author), + + /* general descriptive text (for SHOW PLUGINS) */ + /* const char* */ + STRUCT_FLD(descr, "InnoDB rollback segment information"), + + /* the plugin license (PLUGIN_LICENSE_XXX) */ + /* int */ + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + + /* the function to invoke when plugin is loaded */ + /* int (*)(void*); */ + STRUCT_FLD(init, i_s_innodb_rseg_init), + + /* the function to invoke when plugin is unloaded */ + /* int (*)(void*); */ + STRUCT_FLD(deinit, i_s_common_deinit), + + /* plugin version (for SHOW PLUGINS) */ + /* unsigned int */ + STRUCT_FLD(version, 0x0100 /* 1.0 */), + + /* struct st_mysql_show_var* */ + STRUCT_FLD(status_vars, NULL), + + /* struct st_mysql_sys_var** */ + STRUCT_FLD(system_vars, NULL), + + /* string version */ + /* const char * */ + STRUCT_FLD(version_info, "1.0"), + + /* Maturity */ + /* int */ + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*********************************************************************** */ static ST_FIELD_INFO i_s_innodb_table_stats_info[] = @@ -2940,6 +3578,23 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_table_stats_maria = +{ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + STRUCT_FLD(info, &i_s_info), + STRUCT_FLD(name, "INNODB_TABLE_STATS"), + STRUCT_FLD(author, plugin_author), + STRUCT_FLD(descr, "InnoDB table statistics in memory"), + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + STRUCT_FLD(init, i_s_innodb_table_stats_init), + STRUCT_FLD(deinit, i_s_common_deinit), + STRUCT_FLD(version, 0x0100 /* 1.0 */), + STRUCT_FLD(status_vars, NULL), + STRUCT_FLD(system_vars, NULL), + STRUCT_FLD(version_info, "1.0"), + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + UNIV_INTERN struct st_mysql_plugin i_s_innodb_index_stats = { STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), @@ -2956,6 +3611,23 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(__reserved1, NULL) }; +UNIV_INTERN struct st_maria_plugin i_s_innodb_index_stats_maria = +{ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + STRUCT_FLD(info, &i_s_info), + STRUCT_FLD(name, "INNODB_INDEX_STATS"), + STRUCT_FLD(author, plugin_author), + STRUCT_FLD(descr, "InnoDB index statistics in memory"), + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + STRUCT_FLD(init, i_s_innodb_index_stats_init), + STRUCT_FLD(deinit, i_s_common_deinit), + STRUCT_FLD(version, 0x0100 /* 1.0 */), + STRUCT_FLD(status_vars, NULL), + STRUCT_FLD(system_vars, NULL), + STRUCT_FLD(version_info, "1.0"), + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; + /*********************************************************************** */ static ST_FIELD_INFO i_s_innodb_admin_command_info[] = @@ -2988,7 +3660,6 @@ i_s_innodb_admin_command_fill( COND* cond) { TABLE* i_s_table = (TABLE *) tables->table; - CHARSET_INFO *cs= system_charset_info; char** query_str; char* ptr; char quote = '\0'; @@ -3122,3 +3793,20 @@ UNIV_INTERN struct st_mysql_plugin i_s_i STRUCT_FLD(system_vars, NULL), STRUCT_FLD(__reserved1, NULL) }; + +UNIV_INTERN struct st_maria_plugin i_s_innodb_admin_command_maria = +{ + STRUCT_FLD(type, MYSQL_INFORMATION_SCHEMA_PLUGIN), + STRUCT_FLD(info, &i_s_info), + STRUCT_FLD(name, "XTRADB_ADMIN_COMMAND"), + STRUCT_FLD(author, plugin_author), + STRUCT_FLD(descr, "XtraDB specific command acceptor"), + STRUCT_FLD(license, PLUGIN_LICENSE_GPL), + STRUCT_FLD(init, i_s_innodb_admin_command_init), + STRUCT_FLD(deinit, i_s_common_deinit), + STRUCT_FLD(version, 0x0100 /* 1.0 */), + STRUCT_FLD(status_vars, NULL), + STRUCT_FLD(system_vars, NULL), + STRUCT_FLD(version_info, "1.0"), + STRUCT_FLD(maturity, MariaDB_PLUGIN_MATURITY_STABLE) +}; === modified file 'storage/xtradb/handler/i_s.h' --- a/storage/xtradb/handler/i_s.h 2009-11-04 20:11:12 +0000 +++ b/storage/xtradb/handler/i_s.h 2010-04-01 14:34:51 +0000 @@ -42,4 +42,20 @@ extern struct st_mysql_plugin i_s_innodb extern struct st_mysql_plugin i_s_innodb_index_stats; extern struct st_mysql_plugin i_s_innodb_admin_command; +extern struct st_maria_plugin i_s_innodb_buffer_pool_pages_maria; +extern struct st_maria_plugin i_s_innodb_buffer_pool_pages_index_maria; +extern struct st_maria_plugin i_s_innodb_buffer_pool_pages_blob_maria; +extern struct st_maria_plugin i_s_innodb_trx_maria; +extern struct st_maria_plugin i_s_innodb_locks_maria; +extern struct st_maria_plugin i_s_innodb_lock_waits_maria; +extern struct st_maria_plugin i_s_innodb_cmp_maria; +extern struct st_maria_plugin i_s_innodb_cmp_reset_maria; +extern struct st_maria_plugin i_s_innodb_cmpmem_maria; +extern struct st_maria_plugin i_s_innodb_cmpmem_reset_maria; +extern struct st_maria_plugin i_s_innodb_patches_maria; +extern struct st_maria_plugin i_s_innodb_rseg_maria; +extern struct st_maria_plugin i_s_innodb_table_stats_maria; +extern struct st_maria_plugin i_s_innodb_index_stats_maria; +extern struct st_maria_plugin i_s_innodb_admin_command_maria; + #endif /* i_s_h */ === modified file 'storage/xtradb/include/ut0lst.h' --- a/storage/xtradb/include/ut0lst.h 2010-03-30 12:36:49 +0000 +++ b/storage/xtradb/include/ut0lst.h 2010-04-06 22:47:08 +0000 @@ -158,7 +158,7 @@ Inserts a NODE2 after NODE1 in a list. /** Invalidate the pointers in a list node. @param NAME list name @param N pointer to the node that was removed */ -# define UT_LIST_REMOVE_CLEAR(NAME, N) {} while (0) +# define UT_LIST_REMOVE_CLEAR(NAME, N) do {} while (0) #endif /*******************************************************************//** === modified file 'storage/xtradb/log/log0recv.c' --- a/storage/xtradb/log/log0recv.c 2010-01-15 15:58:25 +0000 +++ b/storage/xtradb/log/log0recv.c 2010-03-31 20:50:54 +0000 @@ -3280,7 +3280,6 @@ recv_recovery_from_checkpoint_finish(voi #endif /* UNIV_DEBUG */ if (recv_needed_recovery && srv_recovery_stats) { - FILE* file = stderr; ulint i; fprintf(stderr, === modified file 'strings/my_vsnprintf.c' --- a/strings/my_vsnprintf.c 2010-03-03 14:44:14 +0000 +++ b/strings/my_vsnprintf.c 2010-04-01 09:04:26 +0000 @@ -400,16 +400,18 @@ start: case 's': { char *par= args_arr[print_arr[i].arg_idx].str_arg; - width= (print_arr[i].flags & WIDTH_ARG) ? - args_arr[print_arr[i].width].longlong_arg : print_arr[i].width; + width= ((print_arr[i].flags & WIDTH_ARG) ? + (uint) args_arr[print_arr[i].width].longlong_arg : + (uint) print_arr[i].width); to= process_str_arg(cs, to, end, width, par, print_arr[i].flags); break; } case 'b': { char *par = args_arr[print_arr[i].arg_idx].str_arg; - width= (print_arr[i].flags & WIDTH_ARG) ? - args_arr[print_arr[i].width].longlong_arg : print_arr[i].width; + width= ((print_arr[i].flags & WIDTH_ARG) ? + (uint) args_arr[print_arr[i].width].longlong_arg : + (uint) print_arr[i].width); to= process_bin_arg(to, end, width, par); break; } @@ -428,8 +430,9 @@ start: { /* Integer parameter */ longlong larg; - length= (print_arr[i].flags & LENGTH_ARG) ? - args_arr[print_arr[i].length].longlong_arg : print_arr[i].length; + length= ((print_arr[i].flags & LENGTH_ARG) ? + args_arr[print_arr[i].length].longlong_arg : + (longlong) print_arr[i].length); if (args_arr[print_arr[i].arg_idx].have_longlong) larg = args_arr[print_arr[i].arg_idx].longlong_arg; === modified file 'support-files/compiler_warnings.supp' --- a/support-files/compiler_warnings.supp 2010-03-30 12:36:49 +0000 +++ b/support-files/compiler_warnings.supp 2010-04-07 15:26:30 +0000 @@ -27,7 +27,7 @@ pars0grm.tab.c: .*'yyerrorlab' : unrefer _flex_tmp.c: .*not enough actual parameters for macro 'yywrap'.* pars0lex.l: .*conversion from 'ulint' to 'int', possible loss of data.* btr/btr0cur\.c: .*value computed is not used.*: 3175-3375 -include/buf0buf\.ic: unused parameter ‘mtr’ +include/buf0buf\.ic: unused parameter .*mtr.* fil/fil0fil\.c: pointer targets in passing argument.*differ in signedness fil/fil0fil\.c: comparison between signed and unsigned : 3100-3199 fil/fil0fil\.c: unused parameter @@ -40,6 +40,7 @@ sync/sync0sync\.c: unused parameter sync/sync0sync\.c: unused variable ut/ut0ut\.c: ignoring return value of srv/srv0srv\.c: value computed is not used +buf/buf0buf\.c: .*block_mutex.* might be used uninitialized # # bdb is not critical to keep up to date @@ -54,6 +55,12 @@ db_vrfy.c : .*comparison is always false # readline is not maintained by us # .*/cmd-line-utils/readline/.* : .* +readline\.c : unused parameter +term\.c : unused parameter +vi\.c : unused parameter +common\.c : unused parameter +term\.c : .* + # # Ignore some warnings in libevent, which is not maintained by us. @@ -154,4 +161,4 @@ ctype-simple\.c : .*unary minus operator # Wrong warning due to GCC bug: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29478 regexec\.c : passing argument 3 of.*matcher.* discards qualifiers from pointer target type libmysql\.c: passing argument 2 of .*memcpy.* discards qualifiers from pointer target type : 3000-4000 -storage/xtradb/dict/dict0dict\.c: passing argument 1 of .*strcpy.* discards qualifiers from pointer target type : 2500-3500 +storage/xtradb/dict/dict0dict\.c : passing argument 1 of .*strcpy.* discards qualifiers from pointer target type : 2500-3500

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2846: Fix Windows compile
by noreply＠launchpad.net 09 Apr '10

09 Apr '10

------------------------------------------------------------ revno: 2846 committer: bo.thorsen <bo(a)sonofthor.dk> branch nick: trunk timestamp: Fri 2010-04-09 11:05:23 +0200 message: Fix Windows compile modified: sql/mysqld.cc storage/maria/ma_loghandler.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription

1 0

[Maria-developers] Updated (by Guest): Table elimination (17)
by worklog-noreply＠askmonty.org 09 Apr '10

09 Apr '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Table elimination CREATION DATE..: Sun, 10 May 2009, 19:57 SUPERVISOR.....: Monty IMPLEMENTOR....: Igor COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 17 (http://askmonty.org/worklog/?tid=17) VERSION........: Server-9.x STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 1 ESTIMATE.......: 3 (hours remain) ORIG. ESTIMATE.: 3 PROGRESS NOTES: -=-=(Guest - Thu, 01 Apr 2010, 01:56)=-=- New attachment: '16_bitMICRO-PROCESSOR.htm' -=-=(Guest - Wed, 24 Mar 2010, 05:58)=-=- Status updated. --- /tmp/wklog.17.old.13223 2010-03-24 05:58:26.000000000 +0000 +++ /tmp/wklog.17.new.13223 2010-03-24 05:58:26.000000000 +0000 @@ -1 +1 @@ -In-Progress +Assigned -=-=(Guest - Wed, 24 Mar 2010, 05:58)=-=- Privacy level updated. --- /tmp/wklog.17.old.13223 2010-03-24 05:58:26.000000000 +0000 +++ /tmp/wklog.17.new.13223 2010-03-24 05:58:26.000000000 +0000 @@ -1 +1 @@ -n +y -=-=(Guest - Wed, 28 Oct 2009, 02:01)=-=- Version updated. --- /tmp/wklog.17.old.24041 2009-10-28 02:01:58.000000000 +0200 +++ /tmp/wklog.17.new.24041 2009-10-28 02:01:58.000000000 +0200 @@ -1 +1 @@ -9.x +Server-9.x -=-=(Guest - Sun, 16 Aug 2009, 16:16)=-=- Category updated. --- /tmp/wklog.17.old.24882 2009-08-16 16:16:49.000000000 +0300 +++ /tmp/wklog.17.new.24882 2009-08-16 16:16:49.000000000 +0300 @@ -1 +1 @@ -Client-BackLog +Server-Sprint -=-=(Guest - Sun, 16 Aug 2009, 16:16)=-=- Version updated. --- /tmp/wklog.17.old.24882 2009-08-16 16:16:49.000000000 +0300 +++ /tmp/wklog.17.new.24882 2009-08-16 16:16:49.000000000 +0300 @@ -1 +1 @@ -Server-5.1 +9.x -=-=(Guest - Wed, 29 Jul 2009, 21:41)=-=- Low Level Design modified. --- /tmp/wklog.17.old.26011 2009-07-29 21:41:04.000000000 +0300 +++ /tmp/wklog.17.new.26011 2009-07-29 21:41:04.000000000 +0300 @@ -2,163 +2,146 @@ ~maria-captains/maria/maria-5.1-table-elimination tree. <contents> -1. Conditions for removal -1.1 Quick check if there are candidates -2. Removal operation properties -3. Removal operation -4. User interface -5. Tests and benchmarks -6. Todo, issues to resolve -6.1 To resolve -6.2 Resolved -7. Additional issues +1. Elimination criteria +2. No outside references check +2.1 Quick check if there are tables with no outside references +3. One-match check +3.1 Functional dependency source #1: Potential eq_ref access +3.2 Functional dependency source #2: col2=func(col1) +3.3 Functional dependency source #3: One or zero records in the table +3.4 Functional dependency check implementation +3.4.1 Equality collection: Option1 +3.4.2 Equality collection: Option2 +3.4.3 Functional dependency propagation - option 1 +3.4.4 Functional dependency propagation - option 2 +4. Removal operation properties +5. Removal operation +6. User interface +6.1 @@optimizer_switch flag +6.2 EXPLAIN [EXTENDED] +7. Miscellaneous adjustments +7.1 Fix used_tables() of aggregate functions +7.2 Make subquery predicates collect their outer references +8. Other concerns +8.1 Relationship with outer->inner joins converter +8.2 Relationship with prepared statements +8.3 Relationship with constant table detection +9. Tests and benchmarks </contents> It's not really about elimination of tables, it's about elimination of inner sides of outer joins. -1. Conditions for removal -------------------------- -We can eliminate an inner side of outer join if: -1. For each record combination of outer tables, it will always produce - exactly one record. -2. There are no references to columns of the inner tables anywhere else in +1. Elimination criteria +======================= +We can eliminate inner side of an outer join nest if: + +1. There are no references to columns of the inner tables anywhere else in the query. +2. For each record combination of outer tables, it will always produce + exactly one matching record combination. + +Most of effort in this WL entry is checking these two conditions. -#1 means that every table inside the outer join nest is: - - is a constant table: - = because it can be accessed via eq_ref(const) access, or - = it is a zero-rows or one-row MyISAM-like table [MARK1] - - has an eq_ref access method candidate. - -#2 means that WHERE clause, ON clauses of embedding outer joins, ORDER BY, - GROUP BY and HAVING do not refer to the inner tables of the outer join - nest. - -1.1 Quick check if there are candidates -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -Before we start to enumerate join nests, here is a quick way to check if -there *can be* something to be removed: +2. No outside references check +============================== +Criterion #1 means that the WHERE clause, ON clauses of embedding/subsequent +outer joins, ORDER BY, GROUP BY and HAVING must have no references to inner +tables of the outer join nest we're trying to remove. + +For multi-table UPDATE/DELETE we also must not remove tables that we're +updating/deleting from or tables that are used in UPDATE's SET clause. + +2.1 Quick check if there are tables with no outside references +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Before we start searching for outer join nests that could be eliminated, +we'll do a quick and cheap check if there possibly could be something that +could be eliminated: - if ((tables used in select_list | + if (there are outer joins && + (tables used in select_list | tables used in group/order by UNION | - tables used in where) != bitmap_of_all_tables) + tables used in where) != bitmap_of_all_join_tables) { attempt table elimination; } -2. Removal operation properties -------------------------------- -* There is always one way to remove (no choice to remove either this or that) -* It is always better to remove as much tables as possible (at least within - our cost model). -Thus, no need for any cost calculations/etc. It's an unconditional rewrite. -3. Removal operation --------------------- -* Remove the outer join nest's nested join structure (i.e. get the - outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, - $OJ->embedding->nested_join. Update table_map's of all ancestor nested - joins). [MARK2] +3. One-match check +================== +We can eliminate inner side of outer join if it will always generate exactly +one matching record combination. -* Move the tables and their JOIN_TABs to front like it is done with const - tables, with exception that if eliminated outer join nest was within - another outer join nest, that shouldn't prevent us from moving away the - eliminated tables. +By definition of OUTER JOIN, a NULL-complemented record combination will be +generated when the inner side of outer join has not produced any matches. -* Update join->table_count and all-join-tables bitmap. +What remains to be checked is that there is no possiblity that inner side of +the outer join could produce more than one matching record combination. -* That's it. Nothing else? +We'll refer to one-match property as "functional dependency": -4. User interface ------------------ -* We'll add an @@optimizer switch flag for table elimination. Tentative - name: 'table_elimination'. - (Note ^^ utility of the above questioned ^, as table elimination can never - be worse than no elimination. We're leaning towards not adding the flag) - -* EXPLAIN will not show the removed tables at all. This will allow to check - if tables were removed, and also will behave nicely with anchor model and - VIEWs: stuff that user doesn't care about just won't be there. +- A outer join nest is functionally dependent [wrt outer tables] if it will + produce one matching record combination per each record combination of + outer tables -5. Tests and benchmarks ------------------------ -Create a benchmark in sql-bench which checks if the DBMS has table -elimination. -[According to Monty] Run - - queries that would use elimination - - queries that are very similar to one above (so that they would have same - QEP, execution cost, etc) but cannot use table elimination. -then compare run times and make a conclusion about whether dbms supports table -elimination. +- A table is functionally dependent wrt certain set of dependency tables, if + record combination of dependency tables uniquely identifies zero or one + matching record in the table -6. Todo, issues to resolve --------------------------- +- Definitions of functional dependency of keys (=column tuples) and columns are + apparent. -6.1 To resolve -~~~~~~~~~~~~~~ -- Relationship with prepared statements. - On one hand, it's natural to desire to make table elimination a - once-per-statement operation, like outer->inner join conversion. We'll have - to limit the applicability by removing [MARK1] as that can change during - lifetime of the statement. - - The other option is to do table elimination every time. This will require to - rework operation [MARK2] to be undoable. - - I'm leaning towards doing the former. With anchor modeling, it is unlikely - that we'll meet outer joins which have N inner tables of which some are 1-row - MyISAM tables that do not have primary key. - -6.2 Resolved -~~~~~~~~~~~~ -* outer->inner join conversion is not a problem for table elimination. - We make outer->inner conversions based on predicates in WHERE. If the WHERE - referred to an inner table (requirement for OJ->IJ conversion) then table - elimination would not be applicable anyway. - -* For Multi-table UPDATEs/DELETEs, need to also analyze the SET clause: - - affected tables must not be eliminated - - tables that are used on the right side of the SET x=y assignments must - not be eliminated either. +Our goal is to prove that the entire join nest is functionally-dependent. -* Aggregate functions used to report that they depend on all tables, that is, +Join nest is functionally dependent (on the otside tables) if each of its +elements (those can be either base tables or join nests) is functionally +dependent. - item_agg_func->used_tables() == (1ULL << join->tables) - 1 +Functional dependency is transitive: if table A is f-dependent on the outer +tables and table B is f.dependent on {A, outer_tables} then B is functionally +dependent on the outer tables. + +Subsequent sections list cases when we can declare a table to be +functionally-dependent. + +3.1 Functional dependency source #1: Potential eq_ref access +------------------------------------------------------------ +This is the most practically-important case. Taking the example from the HLD +of this WL entry: + + select + A.colA + from + tableA A + left outer join + tableB B + on + B.id = A.id; - always. Fixed it, now aggregate function reports it depends on - tables that its arguments depend on. In particular, COUNT(*) reports - that it depends on no tables (item_count_star->used_tables()==0). - One consequence of that is that "item->used_tables()==0" is not - equivalent to "item->const_item()==true" anymore (not sure if it's - "anymore" or this has been already happening). - -* EXPLAIN EXTENDED warning text was generated after the JOIN object has - been discarded. This didn't allow to use information about join plan - when printing the warning. Fixed this by keeping the JOIN objects until - we've printed the warning (have also an intent to remove the const - tables from the join output). - -7. Additional issues --------------------- -* We remove ON clauses within outer join nests. If these clauses contain - subqueries, they probably should be gone from EXPLAIN output also? - Yes. Current approach: when removing an outer join nest, walk the ON clause - and mark subselects as eliminated. Then let EXPLAIN code check if the - SELECT was eliminated before the printing (EXPLAIN is generated by doing - a recursive descent, so the check will also cause children of eliminated - selects not to be printed) - -* Table elimination is performed after constant table detection (but before - the range analysis). Constant tables are technically different from - eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). - Considering we've already done the join_read_const_table() call, is there any - real difference between constant table and eliminated one? If there is, should - we mark const tables also as eliminated? - from user/EXPLAIN point of view: no. constant table is the one that we read - one record from. eliminated table is the one that we don't acccess at all. +and generalizing it: a table TBL is functionally-dependent if the ON +expression allows to construct a potential eq_ref access to table TBL that +uses only outer or functionally-dependent tables. + +In other words: table TBL will have one match if the ON expression can be +converted into this form + + TBL.unique_key=func(one_match_tables) AND .. remainder ... + +(with appropriate extension for multi-part keys), where + + one_match_tables= { + tables that are not on the inner side of the outer join in question, and + functionally dependent tables + } + +Note that this will cover constant tables, except those that are constant because +they have 0/1 record or are partitioned and have no used partitions. + + +3.2 Functional dependency source #2: col2=func(col1) +---------------------------------------------------- +This comes from the second example in the HLS: -* What is described above will not be able to eliminate this outer join create unique index idx on tableB (id, fromDate); ... left outer join @@ -169,32 +152,331 @@ B.fromDate = (select max(sub.fromDate) from tableB sub where sub.id = A.id); - This is because condition "B.fromDate= func(tableB)" cannot be used. - Reason#1: update_ref_and_keys() does not consider such conditions to - be of any use (and indeed they are not usable for ref access) - so they are not put into KEYUSE array. - Reason#2: even if they were put there, we would need to be able to tell - between predicates like - B.fromDate= func(B.id) // guarantees only one matching row as - // B.id is already bound by B.id=A.id - // hence B.fromDate becomes bound too. - and - "B.fromDate= func(B.*)" // Can potentially have many matching - // records. - We need to - - Have update_ref_and_keys() create KEYUSE elements for such equalities - - Have eliminate_tables() and friends make a more accurate check. - The right check is to check whether all parts of a unique key are bound. - If we have keypartX to be bound, then t.keypartY=func(keypartX) makes - keypartY to be bound. - The difficulty here is that correlated subquery predicate cannot tell what - columns it depends on (it only remembers tables). - Traversing the predicate is expensive and complicated. - We're leaning towards making each subquery predicate have a List<Item> with - items that - - are in the current select - - and it depends on. - This list will be useful in certain other subquery optimizations as well, - it is cheap to collect it in fix_fields() phase, so it will be collected - for every subquery predicate. +Here it is apparent that tableB can be eliminated. It is not possible to +construct eq_ref access to tableB, though, because for the second part of the +primary key (fromDate column) we only got a condition in this form: + + B.fromDate= func(tableB) + +(we write "func(tableB)" because ref optimizer can only determine which tables +the right part of the equality depends on). + +In general case, equality like this doesn't guarantee functional dependency. +For example, if func() == { return fromDate;}, i.e the ON expression is + + ... ON B.id = A.id and B.fromDate = B.fromDate + +then that would allow table B to have multiple matches per record of table A. + +In order to be able to distinguish between these two cases, we'll need to go +down to column level: + +- A table is functionally dependent if it has a unique key that's functionally + dependent + +- A unique key is functionally dependent when all of its columns are + functionally dependent + +- A table column is functionally dependent if the ON clause allows to extract + an AND-part in this form: + + tbl.column = f(functionally-dependent columns or columns of outer tables) + +3.3 Functional dependency source #3: One or zero records in the table +--------------------------------------------------------------------- +A table with one or zero records cannot generate more than one matching +record. This source is of lesser importance as one/zero-record tables are only +MyISAM tables. + +3.4 Functional dependency check implementation +---------------------------------------------- +As shown above, we need something similar to KEYUSE structures, but not +exactly that (we need things that current ref optimizer considers unusable and +don't need things that it considers usable). + +3.4.1 Equality collection: Option1 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +We could +- extend KEYUSE structures to store all kinds of equalities we need +- change update_ref_and_keys() and co. to collect equalities both for ref + access and for table elimination + = [possibly] Improve [eq_]ref access to be able to use equalities in + form keypart2=func(keypart1) +- process the KEYUSE array both by table elimination and by ref access + optimizer. + ++ This requires less effort. +- Code will have to be changed all over sql_select.cc +- update_ref_and_keys() and co. already do several unrelated things. Hooking + up table elimination will make it even worse. + +3.4.2 Equality collection: Option2 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Alternatively, we could process the WHERE clause totally on our own. ++ Table elimination is standalone and easy to detach module. +- Some code duplication with update_ref_and_keys() and co. + +Having got the equalities, we'll to propagate functional dependency property +to unique keys, tables and, ultimately, join nests. + +3.4.3 Functional dependency propagation - option 1 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Borrow the approach used in constant table detection code: + + do + { + converted= FALSE; + for each table T in join nest + { + if (check_if_functionally_dependent(T)) + converted= TRUE; + } + } while (converted == TRUE); + + check_if_functionally_dependent(T) + { + if (T has eq_ref access based on func_dep_tables) + return TRUE; + + Apply the same do-while loop-based approach to available equalities + T.column1=func(other columns) + to spread the set of functionally-dependent columns. The goal is to get + all columns of a certain unique key to be bound. + } + + +3.4.4 Functional dependency propagation - option 2 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Analyze the ON expression(s) and build a list of + + tbl.field = expr(...) + +equalities. tbl here is a table that belongs to a join nest that could +potentially be eliminated. + +besides those, add to the list + - An element for each unique key in the table that needs to be eliminated + - An element for each table that needs to be eliminated + - An element for each join nest that can be eliminated (i.e. has no + references from outside). + +Then, setup "reverse dependencies": each element should have pointers to +elements that are functionally dependent on it: + +- "tbl.field=expr(...)" equality is functionally dependent on all fields that + are used in "expr(...)" (here we take into account only fields that belong + to tables that can potentially be eliminated). +- a unique key is dependent on all of its components +- a table is dependent on all of its unique keys +- a join nest is dependent on all tables that it contains + +These pointers are stored in form of one bitmap, such that: + + "X depends on Y" == test( bitmap[(X's number)*n_objects + (Y's number)] ) + +Each object also stores a number of dependencies it needs to be satisfied +before it itself is satisfied: + +- "tbl.field=expr(...)" needs all its underlying fields (if a field is + referenced many times it is counted only once) + +- a unique key needs all of its key parts + +- a table needs only one of its unique keys + +- a join nest needs all of its tables + +(TODO: so what do we do when we've marked a table as constant? We'll need to +update the "field=expr(....)" elements that use fields of that table. And the +problem is that we won't know how much to decrement from the counters of those +elements. + +Solution#1: switch to table_map() based approach. +Solution#2: introduce separate elements for each involved field. + field will depend on its table, + "field=expr" will depend on fields. +) + +Besides the above, let each element have a pointer to another element, so that +we can have a linked list of elements. + +After the above structures have been created, we start the main algorithm. + +The first step is to create a list of functionally-dependent elements. We walk +across array of dependencies and mark those elements that are already bound +(i.e. their dependencies are satisfied). At the moment those immediately-bound +are only "field=expr" dependencies that don't refer to any columns that are +not bound. + +The second step is the loop + + while (bound_list is not empty) + { + Take the first bound element F off the list. + Use the bitmap to find out what other elements depended on it + for each such element E + { + if (E becomes bound after F is bound) + add E to the list; + } + } + +The last step is to walk through elements that represent the join nests. Those +that are bound can be eliminated. + +4. Removal operation properties +=============================== +* There is always one way to remove (no choice to remove either this or that) +* It is always better to remove as much tables as possible (at least within + our cost model). +Thus, no need for any cost calculations/etc. It's an unconditional rewrite. + + +5. Removal operation +==================== +(This depends a lot on whether we make table elimination a one-off rewrite or +conditional) + +At the moment table elimination is re-done for each join re-execution, hence +the removal operation is designed not to modify any statement's permanent +members. + +* Remove the outer join nest's nested join structure (i.e. get the + outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, + $OJ->embedding->nested_join. Update table_map's of all ancestor nested + joins). [MARK2] + +* Move the tables and their JOIN_TABs to the front of join order, like it is + done with const tables, with exception that if eliminated outer join nest + was within another outer join nest, that shouldn't prevent us from moving + away the eliminated tables. + +* Update join->table_count and all-join-tables bitmap. + ^ TODO: not true anymore ^ + +* That's it. Nothing else? + +6. User interface +================= + +6.1 @@optimizer_switch flag +--------------------------- +Argument againist adding the flag: +* It is always better to perform table elimination than not to do it. + +Arguments for the flag: +* It is always theoretically possible that the new code will cause unintended + slowdowns. +* Having the flag is useful for QA and comparative benchmarking. + +Decision so far: add the flag under #ifdef. Make the flag be present in debug +builds. + +6.2 EXPLAIN [EXTENDED] +---------------------- +There are two possible options: +1. Show eliminated tables, like we do with const tables. +2. Do not show eliminated tables. + +We chose option 2, because: +- the table is not accessed at all (besides locking it) +- it is more natural for anchor model user - when he's querying an anchor- + and attributes view, he doesn't care about the unused attributes. + +EXPLAIN EXTENDED+SHOW WARNINGS won't show the removed table either. + +NOTE: Before this WL, the warning text was generated after all JOIN objects +have been destroyed. This didn't allow to use information about join plan +when printing the warning. We've fixed this by keeping the JOIN objects until +the warning text has been generated. + +Table elimination removes inner sides of outer join, and logically the ON +clause is also removed. If this clause has any subqueries, they will be +also removed from EXPLAIN output. + +An exception to the above is that if we eliminate a derived table, it will +still be shown in EXPLAIN output. This comes from the fact that the FROM +subqueries are evaluated before table elimination is invoked. +TODO: Is the above ok or still remove parts of FROM subqueries? + +7. Miscellaneous adjustments +============================ + +7.1 Fix used_tables() of aggregate functions +-------------------------------------------- +Aggregate functions used to report that they depend on all tables, that is, + + item_agg_func->used_tables() == (1ULL << join->tables) - 1 + +always. Fixed it, now aggregate function reports that it depends on the +tables that its arguments depend on. In particular, COUNT(*) reports that it +depends on no tables (item_count_star->used_tables()==0). One consequence of +that is that "item->used_tables()==0" is not equivalent to +"item->const_item()==true" anymore (not sure if it's "anymore" or this has +been already so for some items). + +7.2 Make subquery predicates collect their outer references +----------------------------------------------------------- +Per-column functional dependency analysis requires us to take a + + tbl.field = func(...) + +equality and tell which columns of which tables are referred from func(...) +expression. For scalar expressions, this is accomplished by Item::walk()-based +traversal. It should be reasonably cheap (the only practical Item that can be +expensive to traverse seems to be a special case of "col IN (const1,const2, +...)". check if we traverse the long list for such items). + +For correlated subqueries, traversal can be expensive, it is cheaper to make +each subquery item have a list of its outer references. The list can be +collected at fix_fields() stage with very little extra cost, and then it could +be used for other optimizations. + + +8. Other concerns +================= + +8.1 Relationship with outer->inner joins converter +-------------------------------------------------- +One could suspect that outer->inner join conversion could get in the way +of table elimination by changing outer joins (which could be eliminated) +to inner (which we will not try to eliminate). +This concern is not valid: we make outer->inner conversions based on +predicates in WHERE. If the WHERE referred to an inner table (this is a +requirement for the conversion) then table elimination would not be +applicable anyway. + +8.2 Relationship with prepared statements +----------------------------------------- +On one hand, it's natural to desire to make table elimination a +once-per-statement operation, like outer->inner join conversion. We'll have +to limit the applicability by removing [MARK1] as that can change during +lifetime of the statement. + +The other option is to do table elimination every time. This will require to +rework operation [MARK2] to be undoable. + + +8.3 Relationship with constant table detection +---------------------------------------------- +Table elimination is performed after constant table detection (but before +the range analysis). Constant tables are technically different from +eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). +Considering we've already done the join_read_const_table() call, is there any +real difference between constant table and eliminated one? If there is, should +we mark const tables also as eliminated? +from user/EXPLAIN point of view: no. constant table is the one that we read +one record from. eliminated table is the one that we don't acccess at all. +TODO + +9. Tests and benchmarks +======================= +Create a benchmark in sql-bench which checks if the DBMS has table +elimination. +[According to Monty] Run + - query Q1 that would use elimination + - query Q2 that is very similar to Q1 (so that they would have same + QEP, execution cost, etc) but cannot use table elimination. +then compare run times and make a conclusion about whether the used dbms +supports table elimination. -=-=(Guest - Thu, 23 Jul 2009, 20:07)=-=- Dependency created: 29 now depends on 17 -=-=(Monty - Thu, 23 Jul 2009, 09:19)=-=- Version updated. --- /tmp/wklog.17.old.24090 2009-07-23 09:19:32.000000000 +0300 +++ /tmp/wklog.17.new.24090 2009-07-23 09:19:32.000000000 +0300 @@ -1 +1 @@ -Server-9.x +Server-5.1 -=-=(Guest - Mon, 20 Jul 2009, 14:28)=-=- deukje weg Worked 1 hour and estimate 3 hours remain (original estimate increased by 4 hours). ------------------------------------------------------------ -=-=(View All Progress Notes, 32 total)=-=- http://askmonty.org/worklog/index.pl?tid=17&nolimit=1 DESCRIPTION: Eliminate not needed tables from SELECT queries.. This will speed up some views and automatically generated queries. Example: CREATE TABLE B (id int primary key); select A.colA from tableA A left outer join tableB B on B.id = A.id; In this case we can remove table B and the join from the query. HIGH-LEVEL SPECIFICATION: Here is an extended explanation of table elimination. Table elimination is a feature found in some modern query optimizers, of which Microsoft SQL Server 2005/2008 seems to have the most advanced implementation. Oracle 11g has also been confirmed to use table elimination but not to the same extent. Basically, what table elimination does, is to remove tables from the execution plan when it is unnecessary to include them. This can, of course, only happen if the right circumstances arise. Let us for example look at the following query: select A.colA from tableA A left outer join tableB B on B.id = A.id; When using A as the left table we ensure that the query will return at least as many rows as there are in that table. For rows where the join condition (B.id = A.id) is not met the selected column (A.colA) will still contain it's original value. The not seen B.* row would contain all NULL:s. However, the result set could actually contain more rows than what is found in tableA if there are duplicates of the column B.id in tableB. If A contains a row [1, "val1"] and B the rows [1, "other1a"],[1, "other1b"] then two rows will match in the join condition. The only way to know what the result will look like is to actually touch both tables during execution. Instead, let's say that tableB contains rows that make it possible to place a unique constraint on the column B.id, for example and often the case a primary key. In this situation we know that we will get exactly as many rows as there are in tableA, since joining with tableB cannot introduce any duplicates. If further, as in the example query, we do not select any columns from tableB, touching that table during execution is unnecessary. We can remove the whole join operation from the execution plan. Both SQL Server 2005/2008 and Oracle 11g will deploy table elimination in the case described above. Let us look at a more advanced query, where Oracle fails. select A.colA from tableA A left outer join tableB B on B.id = A.id and B.fromDate = ( select max(sub.fromDate) from tableB sub where sub.id = A.id ); In this example we have added another join condition, which ensures that we only pick the matching row from tableB having the latest fromDate. In this case tableB will contain duplicates of the column B.id, so in order to ensure uniqueness the primary key has to contain the fromDate column as well. In other words the primary key of tableB is (B.id, B.fromDate). Furthermore, since the subselect ensures that we only pick the latest B.fromDate for a given B.id we know that at most one row will match the join condition. We will again have the situation where joining with tableB cannot affect the number of rows in the result set. Since we do not select any columns from tableB, the whole join operation can be eliminated from the execution plan. SQL Server 2005/2008 will deploy table elimination in this situation as well. We have not found a way to make Oracle 11g use it for this type of query. Queries like these arise in two situations. Either when you have denormalized model consisting of a fact table with several related dimension tables, or when you have a highly normalized model where each attribute is stored in its own table. The example with the subselect is common whenever you store historized/versioned data. LOW-LEVEL DESIGN: The code (currently in development) is at lp: ~maria-captains/maria/maria-5.1-table-elimination tree. <contents> 1. Elimination criteria 2. No outside references check 2.1 Quick check if there are tables with no outside references 3. One-match check 3.1 Functional dependency source #1: Potential eq_ref access 3.2 Functional dependency source #2: col2=func(col1) 3.3 Functional dependency source #3: One or zero records in the table 3.4 Functional dependency check implementation 3.4.1 Equality collection: Option1 3.4.2 Equality collection: Option2 3.4.3 Functional dependency propagation - option 1 3.4.4 Functional dependency propagation - option 2 4. Removal operation properties 5. Removal operation 6. User interface 6.1 @@optimizer_switch flag 6.2 EXPLAIN [EXTENDED] 7. Miscellaneous adjustments 7.1 Fix used_tables() of aggregate functions 7.2 Make subquery predicates collect their outer references 8. Other concerns 8.1 Relationship with outer->inner joins converter 8.2 Relationship with prepared statements 8.3 Relationship with constant table detection 9. Tests and benchmarks </contents> It's not really about elimination of tables, it's about elimination of inner sides of outer joins. 1. Elimination criteria ======================= We can eliminate inner side of an outer join nest if: 1. There are no references to columns of the inner tables anywhere else in the query. 2. For each record combination of outer tables, it will always produce exactly one matching record combination. Most of effort in this WL entry is checking these two conditions. 2. No outside references check ============================== Criterion #1 means that the WHERE clause, ON clauses of embedding/subsequent outer joins, ORDER BY, GROUP BY and HAVING must have no references to inner tables of the outer join nest we're trying to remove. For multi-table UPDATE/DELETE we also must not remove tables that we're updating/deleting from or tables that are used in UPDATE's SET clause. 2.1 Quick check if there are tables with no outside references ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Before we start searching for outer join nests that could be eliminated, we'll do a quick and cheap check if there possibly could be something that could be eliminated: if (there are outer joins && (tables used in select_list | tables used in group/order by UNION | tables used in where) != bitmap_of_all_join_tables) { attempt table elimination; } 3. One-match check ================== We can eliminate inner side of outer join if it will always generate exactly one matching record combination. By definition of OUTER JOIN, a NULL-complemented record combination will be generated when the inner side of outer join has not produced any matches. What remains to be checked is that there is no possiblity that inner side of the outer join could produce more than one matching record combination. We'll refer to one-match property as "functional dependency": - A outer join nest is functionally dependent [wrt outer tables] if it will produce one matching record combination per each record combination of outer tables - A table is functionally dependent wrt certain set of dependency tables, if record combination of dependency tables uniquely identifies zero or one matching record in the table - Definitions of functional dependency of keys (=column tuples) and columns are apparent. Our goal is to prove that the entire join nest is functionally-dependent. Join nest is functionally dependent (on the otside tables) if each of its elements (those can be either base tables or join nests) is functionally dependent. Functional dependency is transitive: if table A is f-dependent on the outer tables and table B is f.dependent on {A, outer_tables} then B is functionally dependent on the outer tables. Subsequent sections list cases when we can declare a table to be functionally-dependent. 3.1 Functional dependency source #1: Potential eq_ref access ------------------------------------------------------------ This is the most practically-important case. Taking the example from the HLD of this WL entry: select A.colA from tableA A left outer join tableB B on B.id = A.id; and generalizing it: a table TBL is functionally-dependent if the ON expression allows to construct a potential eq_ref access to table TBL that uses only outer or functionally-dependent tables. In other words: table TBL will have one match if the ON expression can be converted into this form TBL.unique_key=func(one_match_tables) AND .. remainder ... (with appropriate extension for multi-part keys), where one_match_tables= { tables that are not on the inner side of the outer join in question, and functionally dependent tables } Note that this will cover constant tables, except those that are constant because they have 0/1 record or are partitioned and have no used partitions. 3.2 Functional dependency source #2: col2=func(col1) ---------------------------------------------------- This comes from the second example in the HLS: create unique index idx on tableB (id, fromDate); ... left outer join tableB B on B.id = A.id and B.fromDate = (select max(sub.fromDate) from tableB sub where sub.id = A.id); Here it is apparent that tableB can be eliminated. It is not possible to construct eq_ref access to tableB, though, because for the second part of the primary key (fromDate column) we only got a condition in this form: B.fromDate= func(tableB) (we write "func(tableB)" because ref optimizer can only determine which tables the right part of the equality depends on). In general case, equality like this doesn't guarantee functional dependency. For example, if func() == { return fromDate;}, i.e the ON expression is ... ON B.id = A.id and B.fromDate = B.fromDate then that would allow table B to have multiple matches per record of table A. In order to be able to distinguish between these two cases, we'll need to go down to column level: - A table is functionally dependent if it has a unique key that's functionally dependent - A unique key is functionally dependent when all of its columns are functionally dependent - A table column is functionally dependent if the ON clause allows to extract an AND-part in this form: tbl.column = f(functionally-dependent columns or columns of outer tables) 3.3 Functional dependency source #3: One or zero records in the table --------------------------------------------------------------------- A table with one or zero records cannot generate more than one matching record. This source is of lesser importance as one/zero-record tables are only MyISAM tables. 3.4 Functional dependency check implementation ---------------------------------------------- As shown above, we need something similar to KEYUSE structures, but not exactly that (we need things that current ref optimizer considers unusable and don't need things that it considers usable). 3.4.1 Equality collection: Option1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ We could - extend KEYUSE structures to store all kinds of equalities we need - change update_ref_and_keys() and co. to collect equalities both for ref access and for table elimination = [possibly] Improve [eq_]ref access to be able to use equalities in form keypart2=func(keypart1) - process the KEYUSE array both by table elimination and by ref access optimizer. + This requires less effort. - Code will have to be changed all over sql_select.cc - update_ref_and_keys() and co. already do several unrelated things. Hooking up table elimination will make it even worse. 3.4.2 Equality collection: Option2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Alternatively, we could process the WHERE clause totally on our own. + Table elimination is standalone and easy to detach module. - Some code duplication with update_ref_and_keys() and co. Having got the equalities, we'll to propagate functional dependency property to unique keys, tables and, ultimately, join nests. 3.4.3 Functional dependency propagation - option 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Borrow the approach used in constant table detection code: do { converted= FALSE; for each table T in join nest { if (check_if_functionally_dependent(T)) converted= TRUE; } } while (converted == TRUE); check_if_functionally_dependent(T) { if (T has eq_ref access based on func_dep_tables) return TRUE; Apply the same do-while loop-based approach to available equalities T.column1=func(other columns) to spread the set of functionally-dependent columns. The goal is to get all columns of a certain unique key to be bound. } 3.4.4 Functional dependency propagation - option 2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Analyze the ON expression(s) and build a list of tbl.field = expr(...) equalities. tbl here is a table that belongs to a join nest that could potentially be eliminated. besides those, add to the list - An element for each unique key in the table that needs to be eliminated - An element for each table that needs to be eliminated - An element for each join nest that can be eliminated (i.e. has no references from outside). Then, setup "reverse dependencies": each element should have pointers to elements that are functionally dependent on it: - "tbl.field=expr(...)" equality is functionally dependent on all fields that are used in "expr(...)" (here we take into account only fields that belong to tables that can potentially be eliminated). - a unique key is dependent on all of its components - a table is dependent on all of its unique keys - a join nest is dependent on all tables that it contains These pointers are stored in form of one bitmap, such that: "X depends on Y" == test( bitmap[(X's number)*n_objects + (Y's number)] ) Each object also stores a number of dependencies it needs to be satisfied before it itself is satisfied: - "tbl.field=expr(...)" needs all its underlying fields (if a field is referenced many times it is counted only once) - a unique key needs all of its key parts - a table needs only one of its unique keys - a join nest needs all of its tables (TODO: so what do we do when we've marked a table as constant? We'll need to update the "field=expr(....)" elements that use fields of that table. And the problem is that we won't know how much to decrement from the counters of those elements. Solution#1: switch to table_map() based approach. Solution#2: introduce separate elements for each involved field. field will depend on its table, "field=expr" will depend on fields. ) Besides the above, let each element have a pointer to another element, so that we can have a linked list of elements. After the above structures have been created, we start the main algorithm. The first step is to create a list of functionally-dependent elements. We walk across array of dependencies and mark those elements that are already bound (i.e. their dependencies are satisfied). At the moment those immediately-bound are only "field=expr" dependencies that don't refer to any columns that are not bound. The second step is the loop while (bound_list is not empty) { Take the first bound element F off the list. Use the bitmap to find out what other elements depended on it for each such element E { if (E becomes bound after F is bound) add E to the list; } } The last step is to walk through elements that represent the join nests. Those that are bound can be eliminated. 4. Removal operation properties =============================== * There is always one way to remove (no choice to remove either this or that) * It is always better to remove as much tables as possible (at least within our cost model). Thus, no need for any cost calculations/etc. It's an unconditional rewrite. 5. Removal operation ==================== (This depends a lot on whether we make table elimination a one-off rewrite or conditional) At the moment table elimination is re-done for each join re-execution, hence the removal operation is designed not to modify any statement's permanent members. * Remove the outer join nest's nested join structure (i.e. get the outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, $OJ->embedding->nested_join. Update table_map's of all ancestor nested joins). [MARK2] * Move the tables and their JOIN_TABs to the front of join order, like it is done with const tables, with exception that if eliminated outer join nest was within another outer join nest, that shouldn't prevent us from moving away the eliminated tables. * Update join->table_count and all-join-tables bitmap. ^ TODO: not true anymore ^ * That's it. Nothing else? 6. User interface ================= 6.1 @@optimizer_switch flag --------------------------- Argument againist adding the flag: * It is always better to perform table elimination than not to do it. Arguments for the flag: * It is always theoretically possible that the new code will cause unintended slowdowns. * Having the flag is useful for QA and comparative benchmarking. Decision so far: add the flag under #ifdef. Make the flag be present in debug builds. 6.2 EXPLAIN [EXTENDED] ---------------------- There are two possible options: 1. Show eliminated tables, like we do with const tables. 2. Do not show eliminated tables. We chose option 2, because: - the table is not accessed at all (besides locking it) - it is more natural for anchor model user - when he's querying an anchor- and attributes view, he doesn't care about the unused attributes. EXPLAIN EXTENDED+SHOW WARNINGS won't show the removed table either. NOTE: Before this WL, the warning text was generated after all JOIN objects have been destroyed. This didn't allow to use information about join plan when printing the warning. We've fixed this by keeping the JOIN objects until the warning text has been generated. Table elimination removes inner sides of outer join, and logically the ON clause is also removed. If this clause has any subqueries, they will be also removed from EXPLAIN output. An exception to the above is that if we eliminate a derived table, it will still be shown in EXPLAIN output. This comes from the fact that the FROM subqueries are evaluated before table elimination is invoked. TODO: Is the above ok or still remove parts of FROM subqueries? 7. Miscellaneous adjustments ============================ 7.1 Fix used_tables() of aggregate functions -------------------------------------------- Aggregate functions used to report that they depend on all tables, that is, item_agg_func->used_tables() == (1ULL << join->tables) - 1 always. Fixed it, now aggregate function reports that it depends on the tables that its arguments depend on. In particular, COUNT(*) reports that it depends on no tables (item_count_star->used_tables()==0). One consequence of that is that "item->used_tables()==0" is not equivalent to "item->const_item()==true" anymore (not sure if it's "anymore" or this has been already so for some items). 7.2 Make subquery predicates collect their outer references ----------------------------------------------------------- Per-column functional dependency analysis requires us to take a tbl.field = func(...) equality and tell which columns of which tables are referred from func(...) expression. For scalar expressions, this is accomplished by Item::walk()-based traversal. It should be reasonably cheap (the only practical Item that can be expensive to traverse seems to be a special case of "col IN (const1,const2, ...)". check if we traverse the long list for such items). For correlated subqueries, traversal can be expensive, it is cheaper to make each subquery item have a list of its outer references. The list can be collected at fix_fields() stage with very little extra cost, and then it could be used for other optimizations. 8. Other concerns ================= 8.1 Relationship with outer->inner joins converter -------------------------------------------------- One could suspect that outer->inner join conversion could get in the way of table elimination by changing outer joins (which could be eliminated) to inner (which we will not try to eliminate). This concern is not valid: we make outer->inner conversions based on predicates in WHERE. If the WHERE referred to an inner table (this is a requirement for the conversion) then table elimination would not be applicable anyway. 8.2 Relationship with prepared statements ----------------------------------------- On one hand, it's natural to desire to make table elimination a once-per-statement operation, like outer->inner join conversion. We'll have to limit the applicability by removing [MARK1] as that can change during lifetime of the statement. The other option is to do table elimination every time. This will require to rework operation [MARK2] to be undoable. 8.3 Relationship with constant table detection ---------------------------------------------- Table elimination is performed after constant table detection (but before the range analysis). Constant tables are technically different from eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). Considering we've already done the join_read_const_table() call, is there any real difference between constant table and eliminated one? If there is, should we mark const tables also as eliminated? from user/EXPLAIN point of view: no. constant table is the one that we read one record from. eliminated table is the one that we don't acccess at all. TODO 9. Tests and benchmarks ======================= Create a benchmark in sql-bench which checks if the DBMS has table elimination. [According to Monty] Run - query Q1 that would use elimination - query Q2 that is very similar to Q1 (so that they would have same QEP, execution cost, etc) but cannot use table elimination. then compare run times and make a conclusion about whether the used dbms supports table elimination. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Guest): Table elimination (17)
by worklog-noreply＠askmonty.org 09 Apr '10

09 Apr '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Table elimination CREATION DATE..: Sun, 10 May 2009, 19:57 SUPERVISOR.....: Monty IMPLEMENTOR....: Igor COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 17 (http://askmonty.org/worklog/?tid=17) VERSION........: Server-9.x STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 1 ESTIMATE.......: 3 (hours remain) ORIG. ESTIMATE.: 3 PROGRESS NOTES: -=-=(Guest - Thu, 01 Apr 2010, 01:56)=-=- New attachment: '16_bitMICRO-PROCESSOR.htm' -=-=(Guest - Wed, 24 Mar 2010, 05:58)=-=- Status updated. --- /tmp/wklog.17.old.13223 2010-03-24 05:58:26.000000000 +0000 +++ /tmp/wklog.17.new.13223 2010-03-24 05:58:26.000000000 +0000 @@ -1 +1 @@ -In-Progress +Assigned -=-=(Guest - Wed, 24 Mar 2010, 05:58)=-=- Privacy level updated. --- /tmp/wklog.17.old.13223 2010-03-24 05:58:26.000000000 +0000 +++ /tmp/wklog.17.new.13223 2010-03-24 05:58:26.000000000 +0000 @@ -1 +1 @@ -n +y -=-=(Guest - Wed, 28 Oct 2009, 02:01)=-=- Version updated. --- /tmp/wklog.17.old.24041 2009-10-28 02:01:58.000000000 +0200 +++ /tmp/wklog.17.new.24041 2009-10-28 02:01:58.000000000 +0200 @@ -1 +1 @@ -9.x +Server-9.x -=-=(Guest - Sun, 16 Aug 2009, 16:16)=-=- Category updated. --- /tmp/wklog.17.old.24882 2009-08-16 16:16:49.000000000 +0300 +++ /tmp/wklog.17.new.24882 2009-08-16 16:16:49.000000000 +0300 @@ -1 +1 @@ -Client-BackLog +Server-Sprint -=-=(Guest - Sun, 16 Aug 2009, 16:16)=-=- Version updated. --- /tmp/wklog.17.old.24882 2009-08-16 16:16:49.000000000 +0300 +++ /tmp/wklog.17.new.24882 2009-08-16 16:16:49.000000000 +0300 @@ -1 +1 @@ -Server-5.1 +9.x -=-=(Guest - Wed, 29 Jul 2009, 21:41)=-=- Low Level Design modified. --- /tmp/wklog.17.old.26011 2009-07-29 21:41:04.000000000 +0300 +++ /tmp/wklog.17.new.26011 2009-07-29 21:41:04.000000000 +0300 @@ -2,163 +2,146 @@ ~maria-captains/maria/maria-5.1-table-elimination tree. <contents> -1. Conditions for removal -1.1 Quick check if there are candidates -2. Removal operation properties -3. Removal operation -4. User interface -5. Tests and benchmarks -6. Todo, issues to resolve -6.1 To resolve -6.2 Resolved -7. Additional issues +1. Elimination criteria +2. No outside references check +2.1 Quick check if there are tables with no outside references +3. One-match check +3.1 Functional dependency source #1: Potential eq_ref access +3.2 Functional dependency source #2: col2=func(col1) +3.3 Functional dependency source #3: One or zero records in the table +3.4 Functional dependency check implementation +3.4.1 Equality collection: Option1 +3.4.2 Equality collection: Option2 +3.4.3 Functional dependency propagation - option 1 +3.4.4 Functional dependency propagation - option 2 +4. Removal operation properties +5. Removal operation +6. User interface +6.1 @@optimizer_switch flag +6.2 EXPLAIN [EXTENDED] +7. Miscellaneous adjustments +7.1 Fix used_tables() of aggregate functions +7.2 Make subquery predicates collect their outer references +8. Other concerns +8.1 Relationship with outer->inner joins converter +8.2 Relationship with prepared statements +8.3 Relationship with constant table detection +9. Tests and benchmarks </contents> It's not really about elimination of tables, it's about elimination of inner sides of outer joins. -1. Conditions for removal -------------------------- -We can eliminate an inner side of outer join if: -1. For each record combination of outer tables, it will always produce - exactly one record. -2. There are no references to columns of the inner tables anywhere else in +1. Elimination criteria +======================= +We can eliminate inner side of an outer join nest if: + +1. There are no references to columns of the inner tables anywhere else in the query. +2. For each record combination of outer tables, it will always produce + exactly one matching record combination. + +Most of effort in this WL entry is checking these two conditions. -#1 means that every table inside the outer join nest is: - - is a constant table: - = because it can be accessed via eq_ref(const) access, or - = it is a zero-rows or one-row MyISAM-like table [MARK1] - - has an eq_ref access method candidate. - -#2 means that WHERE clause, ON clauses of embedding outer joins, ORDER BY, - GROUP BY and HAVING do not refer to the inner tables of the outer join - nest. - -1.1 Quick check if there are candidates -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -Before we start to enumerate join nests, here is a quick way to check if -there *can be* something to be removed: +2. No outside references check +============================== +Criterion #1 means that the WHERE clause, ON clauses of embedding/subsequent +outer joins, ORDER BY, GROUP BY and HAVING must have no references to inner +tables of the outer join nest we're trying to remove. + +For multi-table UPDATE/DELETE we also must not remove tables that we're +updating/deleting from or tables that are used in UPDATE's SET clause. + +2.1 Quick check if there are tables with no outside references +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Before we start searching for outer join nests that could be eliminated, +we'll do a quick and cheap check if there possibly could be something that +could be eliminated: - if ((tables used in select_list | + if (there are outer joins && + (tables used in select_list | tables used in group/order by UNION | - tables used in where) != bitmap_of_all_tables) + tables used in where) != bitmap_of_all_join_tables) { attempt table elimination; } -2. Removal operation properties -------------------------------- -* There is always one way to remove (no choice to remove either this or that) -* It is always better to remove as much tables as possible (at least within - our cost model). -Thus, no need for any cost calculations/etc. It's an unconditional rewrite. -3. Removal operation --------------------- -* Remove the outer join nest's nested join structure (i.e. get the - outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, - $OJ->embedding->nested_join. Update table_map's of all ancestor nested - joins). [MARK2] +3. One-match check +================== +We can eliminate inner side of outer join if it will always generate exactly +one matching record combination. -* Move the tables and their JOIN_TABs to front like it is done with const - tables, with exception that if eliminated outer join nest was within - another outer join nest, that shouldn't prevent us from moving away the - eliminated tables. +By definition of OUTER JOIN, a NULL-complemented record combination will be +generated when the inner side of outer join has not produced any matches. -* Update join->table_count and all-join-tables bitmap. +What remains to be checked is that there is no possiblity that inner side of +the outer join could produce more than one matching record combination. -* That's it. Nothing else? +We'll refer to one-match property as "functional dependency": -4. User interface ------------------ -* We'll add an @@optimizer switch flag for table elimination. Tentative - name: 'table_elimination'. - (Note ^^ utility of the above questioned ^, as table elimination can never - be worse than no elimination. We're leaning towards not adding the flag) - -* EXPLAIN will not show the removed tables at all. This will allow to check - if tables were removed, and also will behave nicely with anchor model and - VIEWs: stuff that user doesn't care about just won't be there. +- A outer join nest is functionally dependent [wrt outer tables] if it will + produce one matching record combination per each record combination of + outer tables -5. Tests and benchmarks ------------------------ -Create a benchmark in sql-bench which checks if the DBMS has table -elimination. -[According to Monty] Run - - queries that would use elimination - - queries that are very similar to one above (so that they would have same - QEP, execution cost, etc) but cannot use table elimination. -then compare run times and make a conclusion about whether dbms supports table -elimination. +- A table is functionally dependent wrt certain set of dependency tables, if + record combination of dependency tables uniquely identifies zero or one + matching record in the table -6. Todo, issues to resolve --------------------------- +- Definitions of functional dependency of keys (=column tuples) and columns are + apparent. -6.1 To resolve -~~~~~~~~~~~~~~ -- Relationship with prepared statements. - On one hand, it's natural to desire to make table elimination a - once-per-statement operation, like outer->inner join conversion. We'll have - to limit the applicability by removing [MARK1] as that can change during - lifetime of the statement. - - The other option is to do table elimination every time. This will require to - rework operation [MARK2] to be undoable. - - I'm leaning towards doing the former. With anchor modeling, it is unlikely - that we'll meet outer joins which have N inner tables of which some are 1-row - MyISAM tables that do not have primary key. - -6.2 Resolved -~~~~~~~~~~~~ -* outer->inner join conversion is not a problem for table elimination. - We make outer->inner conversions based on predicates in WHERE. If the WHERE - referred to an inner table (requirement for OJ->IJ conversion) then table - elimination would not be applicable anyway. - -* For Multi-table UPDATEs/DELETEs, need to also analyze the SET clause: - - affected tables must not be eliminated - - tables that are used on the right side of the SET x=y assignments must - not be eliminated either. +Our goal is to prove that the entire join nest is functionally-dependent. -* Aggregate functions used to report that they depend on all tables, that is, +Join nest is functionally dependent (on the otside tables) if each of its +elements (those can be either base tables or join nests) is functionally +dependent. - item_agg_func->used_tables() == (1ULL << join->tables) - 1 +Functional dependency is transitive: if table A is f-dependent on the outer +tables and table B is f.dependent on {A, outer_tables} then B is functionally +dependent on the outer tables. + +Subsequent sections list cases when we can declare a table to be +functionally-dependent. + +3.1 Functional dependency source #1: Potential eq_ref access +------------------------------------------------------------ +This is the most practically-important case. Taking the example from the HLD +of this WL entry: + + select + A.colA + from + tableA A + left outer join + tableB B + on + B.id = A.id; - always. Fixed it, now aggregate function reports it depends on - tables that its arguments depend on. In particular, COUNT(*) reports - that it depends on no tables (item_count_star->used_tables()==0). - One consequence of that is that "item->used_tables()==0" is not - equivalent to "item->const_item()==true" anymore (not sure if it's - "anymore" or this has been already happening). - -* EXPLAIN EXTENDED warning text was generated after the JOIN object has - been discarded. This didn't allow to use information about join plan - when printing the warning. Fixed this by keeping the JOIN objects until - we've printed the warning (have also an intent to remove the const - tables from the join output). - -7. Additional issues --------------------- -* We remove ON clauses within outer join nests. If these clauses contain - subqueries, they probably should be gone from EXPLAIN output also? - Yes. Current approach: when removing an outer join nest, walk the ON clause - and mark subselects as eliminated. Then let EXPLAIN code check if the - SELECT was eliminated before the printing (EXPLAIN is generated by doing - a recursive descent, so the check will also cause children of eliminated - selects not to be printed) - -* Table elimination is performed after constant table detection (but before - the range analysis). Constant tables are technically different from - eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). - Considering we've already done the join_read_const_table() call, is there any - real difference between constant table and eliminated one? If there is, should - we mark const tables also as eliminated? - from user/EXPLAIN point of view: no. constant table is the one that we read - one record from. eliminated table is the one that we don't acccess at all. +and generalizing it: a table TBL is functionally-dependent if the ON +expression allows to construct a potential eq_ref access to table TBL that +uses only outer or functionally-dependent tables. + +In other words: table TBL will have one match if the ON expression can be +converted into this form + + TBL.unique_key=func(one_match_tables) AND .. remainder ... + +(with appropriate extension for multi-part keys), where + + one_match_tables= { + tables that are not on the inner side of the outer join in question, and + functionally dependent tables + } + +Note that this will cover constant tables, except those that are constant because +they have 0/1 record or are partitioned and have no used partitions. + + +3.2 Functional dependency source #2: col2=func(col1) +---------------------------------------------------- +This comes from the second example in the HLS: -* What is described above will not be able to eliminate this outer join create unique index idx on tableB (id, fromDate); ... left outer join @@ -169,32 +152,331 @@ B.fromDate = (select max(sub.fromDate) from tableB sub where sub.id = A.id); - This is because condition "B.fromDate= func(tableB)" cannot be used. - Reason#1: update_ref_and_keys() does not consider such conditions to - be of any use (and indeed they are not usable for ref access) - so they are not put into KEYUSE array. - Reason#2: even if they were put there, we would need to be able to tell - between predicates like - B.fromDate= func(B.id) // guarantees only one matching row as - // B.id is already bound by B.id=A.id - // hence B.fromDate becomes bound too. - and - "B.fromDate= func(B.*)" // Can potentially have many matching - // records. - We need to - - Have update_ref_and_keys() create KEYUSE elements for such equalities - - Have eliminate_tables() and friends make a more accurate check. - The right check is to check whether all parts of a unique key are bound. - If we have keypartX to be bound, then t.keypartY=func(keypartX) makes - keypartY to be bound. - The difficulty here is that correlated subquery predicate cannot tell what - columns it depends on (it only remembers tables). - Traversing the predicate is expensive and complicated. - We're leaning towards making each subquery predicate have a List<Item> with - items that - - are in the current select - - and it depends on. - This list will be useful in certain other subquery optimizations as well, - it is cheap to collect it in fix_fields() phase, so it will be collected - for every subquery predicate. +Here it is apparent that tableB can be eliminated. It is not possible to +construct eq_ref access to tableB, though, because for the second part of the +primary key (fromDate column) we only got a condition in this form: + + B.fromDate= func(tableB) + +(we write "func(tableB)" because ref optimizer can only determine which tables +the right part of the equality depends on). + +In general case, equality like this doesn't guarantee functional dependency. +For example, if func() == { return fromDate;}, i.e the ON expression is + + ... ON B.id = A.id and B.fromDate = B.fromDate + +then that would allow table B to have multiple matches per record of table A. + +In order to be able to distinguish between these two cases, we'll need to go +down to column level: + +- A table is functionally dependent if it has a unique key that's functionally + dependent + +- A unique key is functionally dependent when all of its columns are + functionally dependent + +- A table column is functionally dependent if the ON clause allows to extract + an AND-part in this form: + + tbl.column = f(functionally-dependent columns or columns of outer tables) + +3.3 Functional dependency source #3: One or zero records in the table +--------------------------------------------------------------------- +A table with one or zero records cannot generate more than one matching +record. This source is of lesser importance as one/zero-record tables are only +MyISAM tables. + +3.4 Functional dependency check implementation +---------------------------------------------- +As shown above, we need something similar to KEYUSE structures, but not +exactly that (we need things that current ref optimizer considers unusable and +don't need things that it considers usable). + +3.4.1 Equality collection: Option1 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +We could +- extend KEYUSE structures to store all kinds of equalities we need +- change update_ref_and_keys() and co. to collect equalities both for ref + access and for table elimination + = [possibly] Improve [eq_]ref access to be able to use equalities in + form keypart2=func(keypart1) +- process the KEYUSE array both by table elimination and by ref access + optimizer. + ++ This requires less effort. +- Code will have to be changed all over sql_select.cc +- update_ref_and_keys() and co. already do several unrelated things. Hooking + up table elimination will make it even worse. + +3.4.2 Equality collection: Option2 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Alternatively, we could process the WHERE clause totally on our own. ++ Table elimination is standalone and easy to detach module. +- Some code duplication with update_ref_and_keys() and co. + +Having got the equalities, we'll to propagate functional dependency property +to unique keys, tables and, ultimately, join nests. + +3.4.3 Functional dependency propagation - option 1 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Borrow the approach used in constant table detection code: + + do + { + converted= FALSE; + for each table T in join nest + { + if (check_if_functionally_dependent(T)) + converted= TRUE; + } + } while (converted == TRUE); + + check_if_functionally_dependent(T) + { + if (T has eq_ref access based on func_dep_tables) + return TRUE; + + Apply the same do-while loop-based approach to available equalities + T.column1=func(other columns) + to spread the set of functionally-dependent columns. The goal is to get + all columns of a certain unique key to be bound. + } + + +3.4.4 Functional dependency propagation - option 2 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Analyze the ON expression(s) and build a list of + + tbl.field = expr(...) + +equalities. tbl here is a table that belongs to a join nest that could +potentially be eliminated. + +besides those, add to the list + - An element for each unique key in the table that needs to be eliminated + - An element for each table that needs to be eliminated + - An element for each join nest that can be eliminated (i.e. has no + references from outside). + +Then, setup "reverse dependencies": each element should have pointers to +elements that are functionally dependent on it: + +- "tbl.field=expr(...)" equality is functionally dependent on all fields that + are used in "expr(...)" (here we take into account only fields that belong + to tables that can potentially be eliminated). +- a unique key is dependent on all of its components +- a table is dependent on all of its unique keys +- a join nest is dependent on all tables that it contains + +These pointers are stored in form of one bitmap, such that: + + "X depends on Y" == test( bitmap[(X's number)*n_objects + (Y's number)] ) + +Each object also stores a number of dependencies it needs to be satisfied +before it itself is satisfied: + +- "tbl.field=expr(...)" needs all its underlying fields (if a field is + referenced many times it is counted only once) + +- a unique key needs all of its key parts + +- a table needs only one of its unique keys + +- a join nest needs all of its tables + +(TODO: so what do we do when we've marked a table as constant? We'll need to +update the "field=expr(....)" elements that use fields of that table. And the +problem is that we won't know how much to decrement from the counters of those +elements. + +Solution#1: switch to table_map() based approach. +Solution#2: introduce separate elements for each involved field. + field will depend on its table, + "field=expr" will depend on fields. +) + +Besides the above, let each element have a pointer to another element, so that +we can have a linked list of elements. + +After the above structures have been created, we start the main algorithm. + +The first step is to create a list of functionally-dependent elements. We walk +across array of dependencies and mark those elements that are already bound +(i.e. their dependencies are satisfied). At the moment those immediately-bound +are only "field=expr" dependencies that don't refer to any columns that are +not bound. + +The second step is the loop + + while (bound_list is not empty) + { + Take the first bound element F off the list. + Use the bitmap to find out what other elements depended on it + for each such element E + { + if (E becomes bound after F is bound) + add E to the list; + } + } + +The last step is to walk through elements that represent the join nests. Those +that are bound can be eliminated. + +4. Removal operation properties +=============================== +* There is always one way to remove (no choice to remove either this or that) +* It is always better to remove as much tables as possible (at least within + our cost model). +Thus, no need for any cost calculations/etc. It's an unconditional rewrite. + + +5. Removal operation +==================== +(This depends a lot on whether we make table elimination a one-off rewrite or +conditional) + +At the moment table elimination is re-done for each join re-execution, hence +the removal operation is designed not to modify any statement's permanent +members. + +* Remove the outer join nest's nested join structure (i.e. get the + outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, + $OJ->embedding->nested_join. Update table_map's of all ancestor nested + joins). [MARK2] + +* Move the tables and their JOIN_TABs to the front of join order, like it is + done with const tables, with exception that if eliminated outer join nest + was within another outer join nest, that shouldn't prevent us from moving + away the eliminated tables. + +* Update join->table_count and all-join-tables bitmap. + ^ TODO: not true anymore ^ + +* That's it. Nothing else? + +6. User interface +================= + +6.1 @@optimizer_switch flag +--------------------------- +Argument againist adding the flag: +* It is always better to perform table elimination than not to do it. + +Arguments for the flag: +* It is always theoretically possible that the new code will cause unintended + slowdowns. +* Having the flag is useful for QA and comparative benchmarking. + +Decision so far: add the flag under #ifdef. Make the flag be present in debug +builds. + +6.2 EXPLAIN [EXTENDED] +---------------------- +There are two possible options: +1. Show eliminated tables, like we do with const tables. +2. Do not show eliminated tables. + +We chose option 2, because: +- the table is not accessed at all (besides locking it) +- it is more natural for anchor model user - when he's querying an anchor- + and attributes view, he doesn't care about the unused attributes. + +EXPLAIN EXTENDED+SHOW WARNINGS won't show the removed table either. + +NOTE: Before this WL, the warning text was generated after all JOIN objects +have been destroyed. This didn't allow to use information about join plan +when printing the warning. We've fixed this by keeping the JOIN objects until +the warning text has been generated. + +Table elimination removes inner sides of outer join, and logically the ON +clause is also removed. If this clause has any subqueries, they will be +also removed from EXPLAIN output. + +An exception to the above is that if we eliminate a derived table, it will +still be shown in EXPLAIN output. This comes from the fact that the FROM +subqueries are evaluated before table elimination is invoked. +TODO: Is the above ok or still remove parts of FROM subqueries? + +7. Miscellaneous adjustments +============================ + +7.1 Fix used_tables() of aggregate functions +-------------------------------------------- +Aggregate functions used to report that they depend on all tables, that is, + + item_agg_func->used_tables() == (1ULL << join->tables) - 1 + +always. Fixed it, now aggregate function reports that it depends on the +tables that its arguments depend on. In particular, COUNT(*) reports that it +depends on no tables (item_count_star->used_tables()==0). One consequence of +that is that "item->used_tables()==0" is not equivalent to +"item->const_item()==true" anymore (not sure if it's "anymore" or this has +been already so for some items). + +7.2 Make subquery predicates collect their outer references +----------------------------------------------------------- +Per-column functional dependency analysis requires us to take a + + tbl.field = func(...) + +equality and tell which columns of which tables are referred from func(...) +expression. For scalar expressions, this is accomplished by Item::walk()-based +traversal. It should be reasonably cheap (the only practical Item that can be +expensive to traverse seems to be a special case of "col IN (const1,const2, +...)". check if we traverse the long list for such items). + +For correlated subqueries, traversal can be expensive, it is cheaper to make +each subquery item have a list of its outer references. The list can be +collected at fix_fields() stage with very little extra cost, and then it could +be used for other optimizations. + + +8. Other concerns +================= + +8.1 Relationship with outer->inner joins converter +-------------------------------------------------- +One could suspect that outer->inner join conversion could get in the way +of table elimination by changing outer joins (which could be eliminated) +to inner (which we will not try to eliminate). +This concern is not valid: we make outer->inner conversions based on +predicates in WHERE. If the WHERE referred to an inner table (this is a +requirement for the conversion) then table elimination would not be +applicable anyway. + +8.2 Relationship with prepared statements +----------------------------------------- +On one hand, it's natural to desire to make table elimination a +once-per-statement operation, like outer->inner join conversion. We'll have +to limit the applicability by removing [MARK1] as that can change during +lifetime of the statement. + +The other option is to do table elimination every time. This will require to +rework operation [MARK2] to be undoable. + + +8.3 Relationship with constant table detection +---------------------------------------------- +Table elimination is performed after constant table detection (but before +the range analysis). Constant tables are technically different from +eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). +Considering we've already done the join_read_const_table() call, is there any +real difference between constant table and eliminated one? If there is, should +we mark const tables also as eliminated? +from user/EXPLAIN point of view: no. constant table is the one that we read +one record from. eliminated table is the one that we don't acccess at all. +TODO + +9. Tests and benchmarks +======================= +Create a benchmark in sql-bench which checks if the DBMS has table +elimination. +[According to Monty] Run + - query Q1 that would use elimination + - query Q2 that is very similar to Q1 (so that they would have same + QEP, execution cost, etc) but cannot use table elimination. +then compare run times and make a conclusion about whether the used dbms +supports table elimination. -=-=(Guest - Thu, 23 Jul 2009, 20:07)=-=- Dependency created: 29 now depends on 17 -=-=(Monty - Thu, 23 Jul 2009, 09:19)=-=- Version updated. --- /tmp/wklog.17.old.24090 2009-07-23 09:19:32.000000000 +0300 +++ /tmp/wklog.17.new.24090 2009-07-23 09:19:32.000000000 +0300 @@ -1 +1 @@ -Server-9.x +Server-5.1 -=-=(Guest - Mon, 20 Jul 2009, 14:28)=-=- deukje weg Worked 1 hour and estimate 3 hours remain (original estimate increased by 4 hours). ------------------------------------------------------------ -=-=(View All Progress Notes, 32 total)=-=- http://askmonty.org/worklog/index.pl?tid=17&nolimit=1 DESCRIPTION: Eliminate not needed tables from SELECT queries.. This will speed up some views and automatically generated queries. Example: CREATE TABLE B (id int primary key); select A.colA from tableA A left outer join tableB B on B.id = A.id; In this case we can remove table B and the join from the query. HIGH-LEVEL SPECIFICATION: Here is an extended explanation of table elimination. Table elimination is a feature found in some modern query optimizers, of which Microsoft SQL Server 2005/2008 seems to have the most advanced implementation. Oracle 11g has also been confirmed to use table elimination but not to the same extent. Basically, what table elimination does, is to remove tables from the execution plan when it is unnecessary to include them. This can, of course, only happen if the right circumstances arise. Let us for example look at the following query: select A.colA from tableA A left outer join tableB B on B.id = A.id; When using A as the left table we ensure that the query will return at least as many rows as there are in that table. For rows where the join condition (B.id = A.id) is not met the selected column (A.colA) will still contain it's original value. The not seen B.* row would contain all NULL:s. However, the result set could actually contain more rows than what is found in tableA if there are duplicates of the column B.id in tableB. If A contains a row [1, "val1"] and B the rows [1, "other1a"],[1, "other1b"] then two rows will match in the join condition. The only way to know what the result will look like is to actually touch both tables during execution. Instead, let's say that tableB contains rows that make it possible to place a unique constraint on the column B.id, for example and often the case a primary key. In this situation we know that we will get exactly as many rows as there are in tableA, since joining with tableB cannot introduce any duplicates. If further, as in the example query, we do not select any columns from tableB, touching that table during execution is unnecessary. We can remove the whole join operation from the execution plan. Both SQL Server 2005/2008 and Oracle 11g will deploy table elimination in the case described above. Let us look at a more advanced query, where Oracle fails. select A.colA from tableA A left outer join tableB B on B.id = A.id and B.fromDate = ( select max(sub.fromDate) from tableB sub where sub.id = A.id ); In this example we have added another join condition, which ensures that we only pick the matching row from tableB having the latest fromDate. In this case tableB will contain duplicates of the column B.id, so in order to ensure uniqueness the primary key has to contain the fromDate column as well. In other words the primary key of tableB is (B.id, B.fromDate). Furthermore, since the subselect ensures that we only pick the latest B.fromDate for a given B.id we know that at most one row will match the join condition. We will again have the situation where joining with tableB cannot affect the number of rows in the result set. Since we do not select any columns from tableB, the whole join operation can be eliminated from the execution plan. SQL Server 2005/2008 will deploy table elimination in this situation as well. We have not found a way to make Oracle 11g use it for this type of query. Queries like these arise in two situations. Either when you have denormalized model consisting of a fact table with several related dimension tables, or when you have a highly normalized model where each attribute is stored in its own table. The example with the subselect is common whenever you store historized/versioned data. LOW-LEVEL DESIGN: The code (currently in development) is at lp: ~maria-captains/maria/maria-5.1-table-elimination tree. <contents> 1. Elimination criteria 2. No outside references check 2.1 Quick check if there are tables with no outside references 3. One-match check 3.1 Functional dependency source #1: Potential eq_ref access 3.2 Functional dependency source #2: col2=func(col1) 3.3 Functional dependency source #3: One or zero records in the table 3.4 Functional dependency check implementation 3.4.1 Equality collection: Option1 3.4.2 Equality collection: Option2 3.4.3 Functional dependency propagation - option 1 3.4.4 Functional dependency propagation - option 2 4. Removal operation properties 5. Removal operation 6. User interface 6.1 @@optimizer_switch flag 6.2 EXPLAIN [EXTENDED] 7. Miscellaneous adjustments 7.1 Fix used_tables() of aggregate functions 7.2 Make subquery predicates collect their outer references 8. Other concerns 8.1 Relationship with outer->inner joins converter 8.2 Relationship with prepared statements 8.3 Relationship with constant table detection 9. Tests and benchmarks </contents> It's not really about elimination of tables, it's about elimination of inner sides of outer joins. 1. Elimination criteria ======================= We can eliminate inner side of an outer join nest if: 1. There are no references to columns of the inner tables anywhere else in the query. 2. For each record combination of outer tables, it will always produce exactly one matching record combination. Most of effort in this WL entry is checking these two conditions. 2. No outside references check ============================== Criterion #1 means that the WHERE clause, ON clauses of embedding/subsequent outer joins, ORDER BY, GROUP BY and HAVING must have no references to inner tables of the outer join nest we're trying to remove. For multi-table UPDATE/DELETE we also must not remove tables that we're updating/deleting from or tables that are used in UPDATE's SET clause. 2.1 Quick check if there are tables with no outside references ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Before we start searching for outer join nests that could be eliminated, we'll do a quick and cheap check if there possibly could be something that could be eliminated: if (there are outer joins && (tables used in select_list | tables used in group/order by UNION | tables used in where) != bitmap_of_all_join_tables) { attempt table elimination; } 3. One-match check ================== We can eliminate inner side of outer join if it will always generate exactly one matching record combination. By definition of OUTER JOIN, a NULL-complemented record combination will be generated when the inner side of outer join has not produced any matches. What remains to be checked is that there is no possiblity that inner side of the outer join could produce more than one matching record combination. We'll refer to one-match property as "functional dependency": - A outer join nest is functionally dependent [wrt outer tables] if it will produce one matching record combination per each record combination of outer tables - A table is functionally dependent wrt certain set of dependency tables, if record combination of dependency tables uniquely identifies zero or one matching record in the table - Definitions of functional dependency of keys (=column tuples) and columns are apparent. Our goal is to prove that the entire join nest is functionally-dependent. Join nest is functionally dependent (on the otside tables) if each of its elements (those can be either base tables or join nests) is functionally dependent. Functional dependency is transitive: if table A is f-dependent on the outer tables and table B is f.dependent on {A, outer_tables} then B is functionally dependent on the outer tables. Subsequent sections list cases when we can declare a table to be functionally-dependent. 3.1 Functional dependency source #1: Potential eq_ref access ------------------------------------------------------------ This is the most practically-important case. Taking the example from the HLD of this WL entry: select A.colA from tableA A left outer join tableB B on B.id = A.id; and generalizing it: a table TBL is functionally-dependent if the ON expression allows to construct a potential eq_ref access to table TBL that uses only outer or functionally-dependent tables. In other words: table TBL will have one match if the ON expression can be converted into this form TBL.unique_key=func(one_match_tables) AND .. remainder ... (with appropriate extension for multi-part keys), where one_match_tables= { tables that are not on the inner side of the outer join in question, and functionally dependent tables } Note that this will cover constant tables, except those that are constant because they have 0/1 record or are partitioned and have no used partitions. 3.2 Functional dependency source #2: col2=func(col1) ---------------------------------------------------- This comes from the second example in the HLS: create unique index idx on tableB (id, fromDate); ... left outer join tableB B on B.id = A.id and B.fromDate = (select max(sub.fromDate) from tableB sub where sub.id = A.id); Here it is apparent that tableB can be eliminated. It is not possible to construct eq_ref access to tableB, though, because for the second part of the primary key (fromDate column) we only got a condition in this form: B.fromDate= func(tableB) (we write "func(tableB)" because ref optimizer can only determine which tables the right part of the equality depends on). In general case, equality like this doesn't guarantee functional dependency. For example, if func() == { return fromDate;}, i.e the ON expression is ... ON B.id = A.id and B.fromDate = B.fromDate then that would allow table B to have multiple matches per record of table A. In order to be able to distinguish between these two cases, we'll need to go down to column level: - A table is functionally dependent if it has a unique key that's functionally dependent - A unique key is functionally dependent when all of its columns are functionally dependent - A table column is functionally dependent if the ON clause allows to extract an AND-part in this form: tbl.column = f(functionally-dependent columns or columns of outer tables) 3.3 Functional dependency source #3: One or zero records in the table --------------------------------------------------------------------- A table with one or zero records cannot generate more than one matching record. This source is of lesser importance as one/zero-record tables are only MyISAM tables. 3.4 Functional dependency check implementation ---------------------------------------------- As shown above, we need something similar to KEYUSE structures, but not exactly that (we need things that current ref optimizer considers unusable and don't need things that it considers usable). 3.4.1 Equality collection: Option1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ We could - extend KEYUSE structures to store all kinds of equalities we need - change update_ref_and_keys() and co. to collect equalities both for ref access and for table elimination = [possibly] Improve [eq_]ref access to be able to use equalities in form keypart2=func(keypart1) - process the KEYUSE array both by table elimination and by ref access optimizer. + This requires less effort. - Code will have to be changed all over sql_select.cc - update_ref_and_keys() and co. already do several unrelated things. Hooking up table elimination will make it even worse. 3.4.2 Equality collection: Option2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Alternatively, we could process the WHERE clause totally on our own. + Table elimination is standalone and easy to detach module. - Some code duplication with update_ref_and_keys() and co. Having got the equalities, we'll to propagate functional dependency property to unique keys, tables and, ultimately, join nests. 3.4.3 Functional dependency propagation - option 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Borrow the approach used in constant table detection code: do { converted= FALSE; for each table T in join nest { if (check_if_functionally_dependent(T)) converted= TRUE; } } while (converted == TRUE); check_if_functionally_dependent(T) { if (T has eq_ref access based on func_dep_tables) return TRUE; Apply the same do-while loop-based approach to available equalities T.column1=func(other columns) to spread the set of functionally-dependent columns. The goal is to get all columns of a certain unique key to be bound. } 3.4.4 Functional dependency propagation - option 2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Analyze the ON expression(s) and build a list of tbl.field = expr(...) equalities. tbl here is a table that belongs to a join nest that could potentially be eliminated. besides those, add to the list - An element for each unique key in the table that needs to be eliminated - An element for each table that needs to be eliminated - An element for each join nest that can be eliminated (i.e. has no references from outside). Then, setup "reverse dependencies": each element should have pointers to elements that are functionally dependent on it: - "tbl.field=expr(...)" equality is functionally dependent on all fields that are used in "expr(...)" (here we take into account only fields that belong to tables that can potentially be eliminated). - a unique key is dependent on all of its components - a table is dependent on all of its unique keys - a join nest is dependent on all tables that it contains These pointers are stored in form of one bitmap, such that: "X depends on Y" == test( bitmap[(X's number)*n_objects + (Y's number)] ) Each object also stores a number of dependencies it needs to be satisfied before it itself is satisfied: - "tbl.field=expr(...)" needs all its underlying fields (if a field is referenced many times it is counted only once) - a unique key needs all of its key parts - a table needs only one of its unique keys - a join nest needs all of its tables (TODO: so what do we do when we've marked a table as constant? We'll need to update the "field=expr(....)" elements that use fields of that table. And the problem is that we won't know how much to decrement from the counters of those elements. Solution#1: switch to table_map() based approach. Solution#2: introduce separate elements for each involved field. field will depend on its table, "field=expr" will depend on fields. ) Besides the above, let each element have a pointer to another element, so that we can have a linked list of elements. After the above structures have been created, we start the main algorithm. The first step is to create a list of functionally-dependent elements. We walk across array of dependencies and mark those elements that are already bound (i.e. their dependencies are satisfied). At the moment those immediately-bound are only "field=expr" dependencies that don't refer to any columns that are not bound. The second step is the loop while (bound_list is not empty) { Take the first bound element F off the list. Use the bitmap to find out what other elements depended on it for each such element E { if (E becomes bound after F is bound) add E to the list; } } The last step is to walk through elements that represent the join nests. Those that are bound can be eliminated. 4. Removal operation properties =============================== * There is always one way to remove (no choice to remove either this or that) * It is always better to remove as much tables as possible (at least within our cost model). Thus, no need for any cost calculations/etc. It's an unconditional rewrite. 5. Removal operation ==================== (This depends a lot on whether we make table elimination a one-off rewrite or conditional) At the moment table elimination is re-done for each join re-execution, hence the removal operation is designed not to modify any statement's permanent members. * Remove the outer join nest's nested join structure (i.e. get the outer join's TABLE_LIST object $OJ and remove it from $OJ->embedding, $OJ->embedding->nested_join. Update table_map's of all ancestor nested joins). [MARK2] * Move the tables and their JOIN_TABs to the front of join order, like it is done with const tables, with exception that if eliminated outer join nest was within another outer join nest, that shouldn't prevent us from moving away the eliminated tables. * Update join->table_count and all-join-tables bitmap. ^ TODO: not true anymore ^ * That's it. Nothing else? 6. User interface ================= 6.1 @@optimizer_switch flag --------------------------- Argument againist adding the flag: * It is always better to perform table elimination than not to do it. Arguments for the flag: * It is always theoretically possible that the new code will cause unintended slowdowns. * Having the flag is useful for QA and comparative benchmarking. Decision so far: add the flag under #ifdef. Make the flag be present in debug builds. 6.2 EXPLAIN [EXTENDED] ---------------------- There are two possible options: 1. Show eliminated tables, like we do with const tables. 2. Do not show eliminated tables. We chose option 2, because: - the table is not accessed at all (besides locking it) - it is more natural for anchor model user - when he's querying an anchor- and attributes view, he doesn't care about the unused attributes. EXPLAIN EXTENDED+SHOW WARNINGS won't show the removed table either. NOTE: Before this WL, the warning text was generated after all JOIN objects have been destroyed. This didn't allow to use information about join plan when printing the warning. We've fixed this by keeping the JOIN objects until the warning text has been generated. Table elimination removes inner sides of outer join, and logically the ON clause is also removed. If this clause has any subqueries, they will be also removed from EXPLAIN output. An exception to the above is that if we eliminate a derived table, it will still be shown in EXPLAIN output. This comes from the fact that the FROM subqueries are evaluated before table elimination is invoked. TODO: Is the above ok or still remove parts of FROM subqueries? 7. Miscellaneous adjustments ============================ 7.1 Fix used_tables() of aggregate functions -------------------------------------------- Aggregate functions used to report that they depend on all tables, that is, item_agg_func->used_tables() == (1ULL << join->tables) - 1 always. Fixed it, now aggregate function reports that it depends on the tables that its arguments depend on. In particular, COUNT(*) reports that it depends on no tables (item_count_star->used_tables()==0). One consequence of that is that "item->used_tables()==0" is not equivalent to "item->const_item()==true" anymore (not sure if it's "anymore" or this has been already so for some items). 7.2 Make subquery predicates collect their outer references ----------------------------------------------------------- Per-column functional dependency analysis requires us to take a tbl.field = func(...) equality and tell which columns of which tables are referred from func(...) expression. For scalar expressions, this is accomplished by Item::walk()-based traversal. It should be reasonably cheap (the only practical Item that can be expensive to traverse seems to be a special case of "col IN (const1,const2, ...)". check if we traverse the long list for such items). For correlated subqueries, traversal can be expensive, it is cheaper to make each subquery item have a list of its outer references. The list can be collected at fix_fields() stage with very little extra cost, and then it could be used for other optimizations. 8. Other concerns ================= 8.1 Relationship with outer->inner joins converter -------------------------------------------------- One could suspect that outer->inner join conversion could get in the way of table elimination by changing outer joins (which could be eliminated) to inner (which we will not try to eliminate). This concern is not valid: we make outer->inner conversions based on predicates in WHERE. If the WHERE referred to an inner table (this is a requirement for the conversion) then table elimination would not be applicable anyway. 8.2 Relationship with prepared statements ----------------------------------------- On one hand, it's natural to desire to make table elimination a once-per-statement operation, like outer->inner join conversion. We'll have to limit the applicability by removing [MARK1] as that can change during lifetime of the statement. The other option is to do table elimination every time. This will require to rework operation [MARK2] to be undoable. 8.3 Relationship with constant table detection ---------------------------------------------- Table elimination is performed after constant table detection (but before the range analysis). Constant tables are technically different from eliminated ones (e.g. the former are shown in EXPLAIN and the latter aren't). Considering we've already done the join_read_const_table() call, is there any real difference between constant table and eliminated one? If there is, should we mark const tables also as eliminated? from user/EXPLAIN point of view: no. constant table is the one that we read one record from. eliminated table is the one that we don't acccess at all. TODO 9. Tests and benchmarks ======================= Create a benchmark in sql-bench which checks if the DBMS has table elimination. [According to Monty] Run - query Q1 that would use elimination - query Q2 that is very similar to Q1 (so that they would have same QEP, execution cost, etc) but cannot use table elimination. then compare run times and make a conclusion about whether the used dbms supports table elimination. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Rev 2783: merged in http://bazaar.launchpad.net/~maria-captains/maria/5.2/
by serg＠askmonty.org 08 Apr '10

08 Apr '10

At http://bazaar.launchpad.net/~maria-captains/maria/5.2/ ------------------------------------------------------------ revno: 2783 [merge] revision-id: sergii(a)pisem.net-20100408210307-j82aap23b14h7j3t parent: sergii(a)pisem.net-20100408171901-npw9ltfhwwma6kvi parent: igor(a)askmonty.org-20100408205517-43ppipg5lrwlnk5u committer: Sergei Golubchik <sergii(a)pisem.net> branch nick: 5.2 timestamp: Thu 2010-04-08 23:03:07 +0200 message: merged modified: sql/sql_select.cc sp1f-sql_select.cc-19700101030959-egb7whpkh76zzvikycs5nsnuviu4fdlb === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2010-03-15 11:51:23 +0000 +++ b/sql/sql_select.cc 2010-04-08 21:03:07 +0000 @@ -2335,7 +2335,13 @@ JOIN::destroy() tab->cleanup(); } tmp_join->tmp_join= 0; + /* + We need to clean up tmp_table_param for reusable JOINs (having non-zero + and different from self tmp_join) because it's not being cleaned up + anywhere else (as we need to keep the join is reusable). + */ tmp_table_param.cleanup(); + tmp_join->tmp_table_param.copy_field= 0; DBUG_RETURN(tmp_join->destroy()); } cond_equal= 0;

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2845: Fixed a problem of merge from mysql-5.1 baseline.
by noreply＠launchpad.net 08 Apr '10

08 Apr '10

------------------------------------------------------------ revno: 2845 committer: Igor Babaev <igor(a)askmonty.org> branch nick: maria-5.1 timestamp: Thu 2010-04-08 13:55:17 -0700 message: Fixed a problem of merge from mysql-5.1 baseline. modified: sql/sql_select.cc -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription

1 0