You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

672 lines
22 KiB

Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Backport of revno 2630.28.10, 2630.28.31, 2630.28.26, 2630.33.1, 2630.39.1, 2630.28.29, 2630.34.3, 2630.34.2, 2630.34.1, 2630.29.29, 2630.29.28, 2630.31.1, 2630.28.13, 2630.28.10, 2617.23.14 and some other minor revisions. This patch implements: WL#4264 "Backup: Stabilize Service Interface" -- all the server prerequisites except si_objects.{h,cc} themselves (they can be just copied over, when needed). WL#4435: Support OUT-parameters in prepared statements. (and all issues in the initial patches for these two tasks, that were discovered in pushbuild and during testing). Bug#39519: mysql_stmt_close() should flush all data associated with the statement. After execution of a prepared statement, send OUT parameters of the invoked stored procedure, if any, to the client. When using the binary protocol, send the parameters in an additional result set over the wire. When using the text protocol, assign out parameters to the user variables from the CALL(@var1, @var2, ...) specification. The following refactoring has been made: - Protocol::send_fields() was renamed to Protocol::send_result_set_metadata(); - A new Protocol::send_result_set_row() was introduced to incapsulate common functionality for sending row data. - Signature of Protocol::prepare_for_send() was changed: this operation does not need a list of items, the number of items is fully sufficient. The following backward incompatible changes have been made: - CLIENT_MULTI_RESULTS is now enabled by default in the client; - CLIENT_PS_MULTI_RESUTLS is now enabled by default in the client. include/mysql.h: Add a new flag to MYSQL_METHODS::flush_use_result function pointer. This flag determines if all results should be flushed or only the first one: - if flush_all_results is TRUE, then cli_flush_use_result() will read/flush all pending results. I.e. it will read all packets while server status attribute indicates that there are more results. This is a new semantic, required to fix the bug. - if flush_all_results is FALSE, the old sematic is preserved -- i.e. cli_flush_use_result() reads data until first EOF-packet. include/mysql.h.pp: Update the ABI with new calls (compatible changes). include/mysql_com.h: Add CLIENT_PS_OUT_PARAMS -- a client capability indicating that the client supportsю libmysql/libmysql.c: Add mysql_stmt_next_result() -- analogue of mysql_next_result() for binary protocol. Fix a minor bug in alloc_fields() -- not all members were copied over, and some only shallow-copied (catalog). Flush all results in mysql_stmt_close() (Bug#39519). libmysqld/lib_sql.cc: Rename send_fields() -> send_result_set_metadata(). Refactoring: change prepare_for_send() so that it accepts only what it really needs -- a number of elements in the list. mysql-test/r/ps.result: Update results: WL#4435. mysql-test/t/ps.test: WL#4435: A test case for an SQL-part of the problem. sql-common/client.c: Bug#39519. Implement new functionality in cli_flush_use_result(): if flush_all_delete is TRUE, then it should read/flush all pending results. sql/Makefile.am: Add a new header sql_prepare.h to the list of build headers. sql/events.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/handler.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/mysql_priv.h: Move sql_prepare.cc-specific declarations to a new header - sql_prepare.h. sql/procedure.h: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/protocol.cc: Move the logic responsible for sending of one result set row to the Protocol class. Define a template for end-of-statement action. Refactoring: change prepare_for_send() so that it accepts only what it really needs -- a number of elements in the list. Rename send_fields() to send_result_set_metadata(). sql/protocol.h: Update with new declarations (WL#4435). Rename send_fields() -> send_result_set_metadata(). prepare_for_send() only needs the number of columns to send, and doesn't use the item list - update signature to require only what's needed. Add a new protocol type -- Protocol_local. sql/repl_failsafe.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/slave.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_acl.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_base.cc: Include sql_prepare.h (for Reprepare_observer). sql/sql_cache.cc: Extend the query cache flags block to be able to store a numeric id for the result format, not just a flag binary/non-binary. sql/sql_class.cc: Update to use the rename of Protocol::send_fields() to Protocol::send_result_set_metadata(). Use Protocol::send_one_result_set_row(). sql/sql_class.h: Move the declaration of Reprepare_observer to the new header - sql_prepare.h. Update to the new signature of class Protocol::send_fields(). sql/sql_connect.cc: Use a protocol template method instead of raw NET layer API at the end of a statement. sql/sql_cursor.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_error.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_handler.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). Use new method Protocol::send_one_result_set_row(). sql/sql_help.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_lex.cc: Initialize multi_statements variable. Add a handy constant for empty lex string. sql/sql_lex.h: Add a separate member for a standalone parsing option - multi-statements support. sql/sql_list.cc: sql_list.h is a standalone header now, no need to include mysql_priv.h. sql/sql_list.h: Make sql_list.h a stand-alone header. sql/sql_parse.cc: Include sql_prepare.h for prepared statements- related declarations. Use a new Protocol template method to end each statement (send OK, EOF or ERROR to the client). sql/sql_prepare.cc: Implement Execute Direct API (WL#4264), currently unused. It will be used by the service interface (Backup). Use a new header - sql_prepare.h. Add support for OUT parameters in the binary and text protocol (prepared statements only). sql/sql_prepare.h: Add a new header to contain (for now) all prepared statement- external related declarations. sql/sql_profile.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_repl.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_select.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_show.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_string.h: Add a way to convert a String to LEX_STRING. sql/sql_table.cc: Rename: Protocol::send_fields() -> Protocol::send_result_set_metadata(). sql/sql_update.cc: Remove an extraneous my_error(). The error is already reported in update_non_unique_table_error(). sql/sql_yacc.yy: Support for multi-statements is an independent property of parsing, not derived from the protocol type. tests/mysql_client_test.c: Add tests for WL#4435 (binary protocol).
16 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Bug#24795: Add SHOW PROFILE Patch contributed by Jeremy Cole. CLA received Oct 2006 by Kaj Arnö Add rudimentary query profiling support. libmysqld/Makefile.am: Add profile file to source list. sql/Makefile.am: Add profiling files to source and header lists. sql/ha_archive.cc: Macro-ized other discovered instances of setting proc_info. sql/ha_myisam.cc: Macroize setting thread-state info sql/item_func.cc: Macro-ized other discovered instances of setting proc_info. sql/lex.h: Add lexer info for profiling. sql/lock.cc: Macroize setting thread-state info sql/log_event.cc: Macro-ized other discovered instances of setting proc_info. sql/mysql_priv.h: Set constants for profiling. sql/repl_failsafe.cc: Macro-ized other discovered instances of setting proc_info. sql/slave.cc: Macro-ized other discovered instances of setting proc_info. sql/sp_head.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_base.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_cache.cc: Macroize setting thread-state info sql/sql_class.cc: Integrate profiling. sql/sql_class.h: Instantiate profiling object. sql/sql_delete.cc: Macroize setting thread-state info sql/sql_insert.cc: Macroize setting thread-state info --- Macro-ized other discovered instances of setting proc_info. sql/sql_lex.cc: Initialize profiling. sql/sql_lex.h: Define lex tokens and allocate space for profiling options. sql/sql_parse.cc: Integrate profiling. --- Macro-ized other discovered instances of setting proc_info. sql/sql_repl.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_select.cc: Macroize setting thread-state info. Clean up some lines. sql/sql_show.cc: Macro-ized other discovered instances of setting proc_info. --- Revert bad use of macro. sql/sql_table.cc: Macroize setting thread-state info sql/sql_update.cc: Macroize setting thread-state info sql/sql_view.cc: Macro-ized other discovered instances of setting proc_info. sql/sql_yacc.yy: Add parser info for profiling. --- Fix new YACC shift/reduce conflict. (Now at 249.) mysql-test/r/profile.result: Test profiling code. --- A not-very-useful result. mysql-test/t/profile.test: Test profiling code. --- Test syntax, but not values of profiles code. sql/sql_profile.cc: Add profiling code. --- Add wishlist comment. sql/sql_profile.h: Add profiling code. --- Changed the value of the macro so that it's syntactically equivalent to a single statement.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Rewrite profiler code to be easier to maintain and less buggy. Between 5.0 and 5.1, the step of incrementing the global query id changed, which broke how the profiler noticed when a new query had started. That reset the state list and caused all but the last five (or so) states to be thrown away. Now, don't watch for query_id changes in the lower level. Add a bogus state change at the end of profiling so that the last real state change is timed. Emit source reference for the start of the span of time instead of the end of it. mysql-test/r/profiling.result: Add a test that shows continuation of execution with multi-statement packets. mysql-test/t/profiling.test: Add a test that shows continuation of execution with multi-statement packets. sql/sql_parse.cc: Insert profiling calls at beginnings and ends of each query. Remove the old way of keeping or discarding profiles, and flipping to new query profiles. sql/sql_profile.cc: No longer use the thread's query_id to determine when we flip to a new statement. Some status statements are set to be NULL in the server. We don't log those, as it doesn't fit this style of profiling yet. Rewrite the parser code to be more active and legible. Relying on passive/lazy discovery of new queries was buggy. Add a bogus status change before ending a profile, so that the previous real status has a endpoint. Emit source reference of the start of the span-of-time instead of the end of it. sql/sql_profile.h: Store the server_query_id at instantiation time for a new query. Rewrite the parser code to be more active. Relying on passive/lazy discovery of new queries was buggy. Name first state to more honestly describe the state. We don't really know of initialization that will follow. sql/sql_show.cc: Update comment to note the decidedly weird field_length behavior on Decimal types in information_schema.
18 years ago
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. BitKeeper/deleted/.del-profile.result: Rename: mysql-test/r/profile.result -> BitKeeper/deleted/.del-profile.result BitKeeper/deleted/.del-profile.test: Rename: mysql-test/t/profile.test -> BitKeeper/deleted/.del-profile.test BitKeeper/deleted/.del-sql_profile.cc: Rename: sql/sql_profile.cc -> BitKeeper/deleted/.del-sql_profile.cc BitKeeper/deleted/.del-sql_profile.h: Rename: sql/sql_profile.h -> BitKeeper/deleted/.del-sql_profile.h configure.in: Add a configure-time option to enable/disable query profiling. The default is enabled. include/my_dbug.h: DBUG_* statements should be syntactically equivalent to a single statement. myisam/mi_open.c: DBUG_* statements should be syntactically equivalent to a single statement. mysql-test/r/information_schema.result: Updated (re-recorded) tests that I missed somehow. I verified these for correctness. mysql-test/r/information_schema_db.result: Updated test results I missed. mysql-test/r/mysqlshow.result: Fix merge problems. ndb/src/ndbapi/DictCache.cpp: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_archive.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_berkeley.cc: Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/ha_myisam.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/ha_myisammrg.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/ha_ndbcluster.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_cmpfunc.cc: DBUG_* statements should be syntactically equivalent to a single statement. sql/item_func.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/lock.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/log_event.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/mysql_priv.h: Use 64-bit constants for the 64-bit bit field. Add a new option bit for whether profiling is active or not. sql/mysqld.cc: Add semicolon to DBUG statement. Add a new system variable and set it. --- Set the minimum, which is zero and not 50. sql/repl_failsafe.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/set_var.cc: Make a new system global variable and session variable, to determine behavior of profiling. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/set_var.h: The THD::options bit field is ulonglong, not ulong. sql/slave.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sp_head.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_base.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_cache.cc: DBUG_* statements should be syntactically equivalent to a single statement. --- Fix merge problems. sql/sql_class.cc: Insert a pointer to the current thread in the profiling code. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_class.h: Create a new system variable, profiling_history_size, and add a member to THD to hold profiling information about this thread's execution. --- Manual merge, undoing first patch. sql/sql_delete.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_insert.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_lex.cc: Initialize profiling options to empty. --- Manual merge, undoing first patch. sql/sql_lex.h: Add info to the lexer object so that we can hold data that comes from parsing statements. Reuse memory addresses of uints that can't occur in the same state- ment. This is dangerous because it involves knowledge of what symbols are never used together, which is information stored obliquely in another file. --- Manual merge, undoing first patch. sql/sql_parse.cc: Add hooks to the parser to jump to profiling code. If profiling is not present, then return an error message upon being used. --- Manual merge, undoing first patch. --- Fix merge problems. --- Include patch backported to 5.0-global. THD::options is a ulonglong, not ulong. sql/sql_prepare.cc: From prepared statement execution, set the query source in the profiler, as we can't get it from thd . --- Make it less expensive to limit the size of the queries. sql/sql_repl.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_select.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. --- Fix merge problems. sql/sql_show.cc: Abstract setting thread-info into a function or macro. Also, remove "static" qualification on schema_table_store_record() so that external functions may use it. --- Manual merge, undoing first patch. sql/sql_table.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_update.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_view.cc: Abstract setting thread-info into a function or macro. --- Manual merge, undoing first patch. sql/sql_yacc.yy: Add new lexer symbols and insert new grammatical rules necessary to retreive profiling information. --- Manual merge, undoing first patch. --- Fix merge problems. sql/table.h: Add enum item for query profiling. BitKeeper/deleted/.del-profiling-master.opt: New BitKeeper file ``mysql-test/t/profiling-master.opt'' mysql-test/r/profiling.result: New BitKeeper file ``mysql-test/r/profiling.result'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. mysql-test/t/profiling.test: New BitKeeper file ``mysql-test/t/profiling.test'' --- Include more verbose that describes the queries so far. Include Giuseppe's tests. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. sql/sql_profile.cc: New BitKeeper file ``sql/sql_profile.cc'' --- If query_source is NULL, as can sometimes happen, then don't try to copy that memory. --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length. sql/sql_profile.h: New BitKeeper file ``sql/sql_profile.h'' --- Make each new session use its own numbering of query_ids, and not show the global-pool numbers to the user. Provide a way for prepared statements to set the query_source. --- Make it less expensive to limit the size of the queries.
19 years ago
  1. /* Copyright (C) 2007 MySQL AB
  2. This program is free software; you can redistribute it and/or modify
  3. it under the terms of the GNU General Public License as published by
  4. the Free Software Foundation; version 2 of the License.
  5. This program is distributed in the hope that it will be useful,
  6. but WITHOUT ANY WARRANTY; without even the implied warranty of
  7. MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  8. GNU General Public License for more details.
  9. You should have received a copy of the GNU General Public License
  10. along with this program; if not, write to the Free Software
  11. Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
  12. /**
  13. @file
  14. Implement query profiling as as list of metaphorical fences, with one fence
  15. per query, and each fencepost a change of thd->proc_info state (with a
  16. snapshot of system statistics). When asked, we can then iterate over the
  17. fenceposts and calculate the distance between them, to inform the user what
  18. happened during a particular query or thd->proc_info state.
  19. User variables that inform profiling behavior:
  20. - "profiling", boolean, session only, "Are queries profiled?"
  21. - "profiling_history_size", integer, session + global, "Num queries stored?"
  22. */
  23. #include "mysql_priv.h"
  24. #include "my_sys.h"
  25. #define TIME_FLOAT_DIGITS 9
  26. /** two vals encoded: (dec*100)+len */
  27. #define TIME_I_S_DECIMAL_SIZE (TIME_FLOAT_DIGITS*100)+(TIME_FLOAT_DIGITS-3)
  28. #define MAX_QUERY_LENGTH 300
  29. /* Reserved for systems that can't record the function name in source. */
  30. const char * const _unknown_func_ = "<unknown>";
  31. /**
  32. Connects Information_Schema and Profiling.
  33. */
  34. int fill_query_profile_statistics_info(THD *thd, TABLE_LIST *tables,
  35. Item *cond)
  36. {
  37. #if defined(ENABLED_PROFILING)
  38. return(thd->profiling.fill_statistics_info(thd, tables, cond));
  39. #else
  40. my_error(ER_FEATURE_DISABLED, MYF(0), "SHOW PROFILE", "enable-profiling");
  41. return(1);
  42. #endif
  43. }
  44. ST_FIELD_INFO query_profile_statistics_info[]=
  45. {
  46. /* name, length, type, value, maybe_null, old_name, open_method */
  47. {"QUERY_ID", 20, MYSQL_TYPE_LONG, 0, false, "Query_id", SKIP_OPEN_TABLE},
  48. {"SEQ", 20, MYSQL_TYPE_LONG, 0, false, "Seq", SKIP_OPEN_TABLE},
  49. {"STATE", 30, MYSQL_TYPE_STRING, 0, false, "Status", SKIP_OPEN_TABLE},
  50. {"DURATION", TIME_I_S_DECIMAL_SIZE, MYSQL_TYPE_DECIMAL, 0, false, "Duration", SKIP_OPEN_TABLE},
  51. {"CPU_USER", TIME_I_S_DECIMAL_SIZE, MYSQL_TYPE_DECIMAL, 0, true, "CPU_user", SKIP_OPEN_TABLE},
  52. {"CPU_SYSTEM", TIME_I_S_DECIMAL_SIZE, MYSQL_TYPE_DECIMAL, 0, true, "CPU_system", SKIP_OPEN_TABLE},
  53. {"CONTEXT_VOLUNTARY", 20, MYSQL_TYPE_LONG, 0, true, "Context_voluntary", SKIP_OPEN_TABLE},
  54. {"CONTEXT_INVOLUNTARY", 20, MYSQL_TYPE_LONG, 0, true, "Context_involuntary", SKIP_OPEN_TABLE},
  55. {"BLOCK_OPS_IN", 20, MYSQL_TYPE_LONG, 0, true, "Block_ops_in", SKIP_OPEN_TABLE},
  56. {"BLOCK_OPS_OUT", 20, MYSQL_TYPE_LONG, 0, true, "Block_ops_out", SKIP_OPEN_TABLE},
  57. {"MESSAGES_SENT", 20, MYSQL_TYPE_LONG, 0, true, "Messages_sent", SKIP_OPEN_TABLE},
  58. {"MESSAGES_RECEIVED", 20, MYSQL_TYPE_LONG, 0, true, "Messages_received", SKIP_OPEN_TABLE},
  59. {"PAGE_FAULTS_MAJOR", 20, MYSQL_TYPE_LONG, 0, true, "Page_faults_major", SKIP_OPEN_TABLE},
  60. {"PAGE_FAULTS_MINOR", 20, MYSQL_TYPE_LONG, 0, true, "Page_faults_minor", SKIP_OPEN_TABLE},
  61. {"SWAPS", 20, MYSQL_TYPE_LONG, 0, true, "Swaps", SKIP_OPEN_TABLE},
  62. {"SOURCE_FUNCTION", 30, MYSQL_TYPE_STRING, 0, true, "Source_function", SKIP_OPEN_TABLE},
  63. {"SOURCE_FILE", 20, MYSQL_TYPE_STRING, 0, true, "Source_file", SKIP_OPEN_TABLE},
  64. {"SOURCE_LINE", 20, MYSQL_TYPE_LONG, 0, true, "Source_line", SKIP_OPEN_TABLE},
  65. {NULL, 0, MYSQL_TYPE_STRING, 0, true, NULL, 0}
  66. };
  67. int make_profile_table_for_show(THD *thd, ST_SCHEMA_TABLE *schema_table)
  68. {
  69. int profile_options = thd->lex->profile_options;
  70. int fields_include_condition_truth_values[]= {
  71. FALSE, /* Query_id */
  72. FALSE, /* Seq */
  73. TRUE, /* Status */
  74. TRUE, /* Duration */
  75. profile_options & PROFILE_CPU, /* CPU_user */
  76. profile_options & PROFILE_CPU, /* CPU_system */
  77. profile_options & PROFILE_CONTEXT, /* Context_voluntary */
  78. profile_options & PROFILE_CONTEXT, /* Context_involuntary */
  79. profile_options & PROFILE_BLOCK_IO, /* Block_ops_in */
  80. profile_options & PROFILE_BLOCK_IO, /* Block_ops_out */
  81. profile_options & PROFILE_IPC, /* Messages_sent */
  82. profile_options & PROFILE_IPC, /* Messages_received */
  83. profile_options & PROFILE_PAGE_FAULTS, /* Page_faults_major */
  84. profile_options & PROFILE_PAGE_FAULTS, /* Page_faults_minor */
  85. profile_options & PROFILE_SWAPS, /* Swaps */
  86. profile_options & PROFILE_SOURCE, /* Source_function */
  87. profile_options & PROFILE_SOURCE, /* Source_file */
  88. profile_options & PROFILE_SOURCE, /* Source_line */
  89. };
  90. ST_FIELD_INFO *field_info;
  91. Name_resolution_context *context= &thd->lex->select_lex.context;
  92. int i;
  93. for (i= 0; schema_table->fields_info[i].field_name != NULL; i++)
  94. {
  95. if (! fields_include_condition_truth_values[i])
  96. continue;
  97. field_info= &schema_table->fields_info[i];
  98. Item_field *field= new Item_field(context,
  99. NullS, NullS, field_info->field_name);
  100. if (field)
  101. {
  102. field->set_name(field_info->old_name,
  103. (uint) strlen(field_info->old_name),
  104. system_charset_info);
  105. if (add_item_to_list(thd, field))
  106. return 1;
  107. }
  108. }
  109. return 0;
  110. }
  111. #if defined(ENABLED_PROFILING)
  112. #define RUSAGE_USEC(tv) ((tv).tv_sec*1000*1000 + (tv).tv_usec)
  113. #define RUSAGE_DIFF_USEC(tv1, tv2) (RUSAGE_USEC((tv1))-RUSAGE_USEC((tv2)))
  114. PROF_MEASUREMENT::PROF_MEASUREMENT(QUERY_PROFILE *profile_arg, const char
  115. *status_arg)
  116. :profile(profile_arg)
  117. {
  118. collect();
  119. set_label(status_arg, NULL, NULL, 0);
  120. }
  121. PROF_MEASUREMENT::PROF_MEASUREMENT(QUERY_PROFILE *profile_arg,
  122. const char *status_arg,
  123. const char *function_arg,
  124. const char *file_arg,
  125. unsigned int line_arg)
  126. :profile(profile_arg)
  127. {
  128. collect();
  129. set_label(status_arg, function_arg, file_arg, line_arg);
  130. }
  131. PROF_MEASUREMENT::~PROF_MEASUREMENT()
  132. {
  133. if (allocated_status_memory != NULL)
  134. my_free(allocated_status_memory, MYF(0));
  135. status= function= file= NULL;
  136. }
  137. void PROF_MEASUREMENT::set_label(const char *status_arg,
  138. const char *function_arg,
  139. const char *file_arg, unsigned int line_arg)
  140. {
  141. size_t sizes[3]; /* 3 == status+function+file */
  142. char *cursor;
  143. /*
  144. Compute all the space we'll need to allocate one block for everything
  145. we'll need, instead of N mallocs.
  146. */
  147. sizes[0]= (status_arg == NULL) ? 0 : strlen(status_arg) + 1;
  148. sizes[1]= (function_arg == NULL) ? 0 : strlen(function_arg) + 1;
  149. sizes[2]= (file_arg == NULL) ? 0 : strlen(file_arg) + 1;
  150. allocated_status_memory= (char *) my_malloc(sizes[0] + sizes[1] + sizes[2], MYF(0));
  151. DBUG_ASSERT(allocated_status_memory != NULL);
  152. cursor= allocated_status_memory;
  153. if (status_arg != NULL)
  154. {
  155. strcpy(cursor, status_arg);
  156. status= cursor;
  157. cursor+= sizes[0];
  158. }
  159. else
  160. status= NULL;
  161. if (function_arg != NULL)
  162. {
  163. strcpy(cursor, function_arg);
  164. function= cursor;
  165. cursor+= sizes[1];
  166. }
  167. else
  168. function= NULL;
  169. if (file_arg != NULL)
  170. {
  171. strcpy(cursor, file_arg);
  172. file= cursor;
  173. cursor+= sizes[2];
  174. }
  175. else
  176. file= NULL;
  177. line= line_arg;
  178. }
  179. /**
  180. This updates the statistics for this moment of time. It captures the state
  181. of the running system, so later we can compare points in time and infer what
  182. happened in the mean time. It should only be called immediately upon
  183. instantiation of this PROF_MEASUREMENT.
  184. @todo Implement resource capture for OSes not like BSD.
  185. */
  186. void PROF_MEASUREMENT::collect()
  187. {
  188. time_usecs= (double) my_getsystime() / 10.0; /* 1 sec was 1e7, now is 1e6 */
  189. #ifdef HAVE_GETRUSAGE
  190. getrusage(RUSAGE_SELF, &rusage);
  191. #endif
  192. }
  193. QUERY_PROFILE::QUERY_PROFILE(PROFILING *profiling_arg, const char *status_arg)
  194. :profiling(profiling_arg), profiling_query_id(0), query_source(NULL)
  195. {
  196. profile_start= new PROF_MEASUREMENT(this, status_arg);
  197. entries.push_back(profile_start);
  198. profile_end= profile_start;
  199. }
  200. QUERY_PROFILE::~QUERY_PROFILE()
  201. {
  202. while (! entries.is_empty())
  203. delete entries.pop();
  204. if (query_source != NULL)
  205. my_free(query_source, MYF(0));
  206. }
  207. /**
  208. @todo Provide a way to include the full text, as in SHOW PROCESSLIST.
  209. */
  210. void QUERY_PROFILE::set_query_source(char *query_source_arg,
  211. uint query_length_arg)
  212. {
  213. /* Truncate to avoid DoS attacks. */
  214. uint length= min(MAX_QUERY_LENGTH, query_length_arg);
  215. DBUG_ASSERT(query_source == NULL); /* we don't leak memory */
  216. if (query_source_arg != NULL)
  217. query_source= my_strndup(query_source_arg, length, MYF(0));
  218. }
  219. void QUERY_PROFILE::new_status(const char *status_arg,
  220. const char *function_arg, const char *file_arg,
  221. unsigned int line_arg)
  222. {
  223. PROF_MEASUREMENT *prof;
  224. DBUG_ENTER("QUERY_PROFILE::status");
  225. DBUG_ASSERT(status_arg != NULL);
  226. if ((function_arg != NULL) && (file_arg != NULL))
  227. prof= new PROF_MEASUREMENT(this, status_arg, function_arg, file_arg, line_arg);
  228. else
  229. prof= new PROF_MEASUREMENT(this, status_arg);
  230. profile_end= prof;
  231. entries.push_back(prof);
  232. DBUG_VOID_RETURN;
  233. }
  234. PROFILING::PROFILING()
  235. :profile_id_counter(1), current(NULL), last(NULL)
  236. {
  237. }
  238. PROFILING::~PROFILING()
  239. {
  240. while (! history.is_empty())
  241. delete history.pop();
  242. if (current != NULL)
  243. delete current;
  244. }
  245. /**
  246. A new state is given, and that signals the profiler to start a new
  247. timed step for the current query's profile.
  248. @param status_arg name of this step
  249. @param function_arg calling function (usually supplied from compiler)
  250. @param function_arg calling file (usually supplied from compiler)
  251. @param function_arg calling line number (usually supplied from compiler)
  252. */
  253. void PROFILING::status_change(const char *status_arg,
  254. const char *function_arg,
  255. const char *file_arg, unsigned int line_arg)
  256. {
  257. DBUG_ENTER("PROFILING::status_change");
  258. if (status_arg == NULL) /* We don't know how to handle that */
  259. DBUG_VOID_RETURN;
  260. if (current == NULL) /* This profile was already discarded. */
  261. DBUG_VOID_RETURN;
  262. if (unlikely(enabled))
  263. current->new_status(status_arg, function_arg, file_arg, line_arg);
  264. DBUG_VOID_RETURN;
  265. }
  266. /**
  267. Prepare to start processing a new query. It is an error to do this
  268. if there's a query already in process; nesting is not supported.
  269. @param initial_state (optional) name of period before first state change
  270. */
  271. void PROFILING::start_new_query(const char *initial_state)
  272. {
  273. DBUG_ENTER("PROFILING::start_new_query");
  274. /* This should never happen unless the server is radically altered. */
  275. if (unlikely(current != NULL))
  276. {
  277. DBUG_PRINT("warning", ("profiling code was asked to start a new query "
  278. "before the old query was finished. This is "
  279. "probably a bug."));
  280. finish_current_query();
  281. }
  282. enabled= (((thd)->options & OPTION_PROFILING) != 0);
  283. if (! enabled) DBUG_VOID_RETURN;
  284. DBUG_ASSERT(current == NULL);
  285. current= new QUERY_PROFILE(this, initial_state);
  286. DBUG_VOID_RETURN;
  287. }
  288. /**
  289. Throw away the current profile, because it's useless or unwanted
  290. or corrupted.
  291. */
  292. void PROFILING::discard_current_query()
  293. {
  294. DBUG_ENTER("PROFILING::discard_current_profile");
  295. delete current;
  296. current= NULL;
  297. DBUG_VOID_RETURN;
  298. }
  299. /**
  300. Try to save the current profile entry, clean up the data if it shouldn't be
  301. saved, and maintain the profile history size. Naturally, this may not
  302. succeed if the profile was previously discarded, and that's expected.
  303. */
  304. void PROFILING::finish_current_query()
  305. {
  306. DBUG_ENTER("PROFILING::finish_current_profile");
  307. if (current != NULL)
  308. {
  309. /* The last fence-post, so we can support the span before this. */
  310. status_change("ending", NULL, NULL, 0);
  311. if ((enabled) && /* ON at start? */
  312. ((thd->options & OPTION_PROFILING) != 0) && /* and ON at end? */
  313. (current->query_source != NULL) &&
  314. (! current->entries.is_empty()))
  315. {
  316. current->profiling_query_id= next_profile_id(); /* assign an id */
  317. history.push_back(current);
  318. last= current; /* never contains something that is not in the history. */
  319. current= NULL;
  320. }
  321. else
  322. {
  323. delete current;
  324. current= NULL;
  325. }
  326. }
  327. /* Maintain the history size. */
  328. while (history.elements > thd->variables.profiling_history_size)
  329. delete history.pop();
  330. DBUG_VOID_RETURN;
  331. }
  332. bool PROFILING::show_profiles()
  333. {
  334. DBUG_ENTER("PROFILING::show_profiles");
  335. QUERY_PROFILE *prof;
  336. List<Item> field_list;
  337. field_list.push_back(new Item_return_int("Query_ID", 10,
  338. MYSQL_TYPE_LONG));
  339. field_list.push_back(new Item_return_int("Duration", TIME_FLOAT_DIGITS-1,
  340. MYSQL_TYPE_DOUBLE));
  341. field_list.push_back(new Item_empty_string("Query", 40));
  342. if (thd->protocol->send_result_set_metadata(&field_list,
  343. Protocol::SEND_NUM_ROWS | Protocol::SEND_EOF))
  344. DBUG_RETURN(TRUE);
  345. SELECT_LEX *sel= &thd->lex->select_lex;
  346. SELECT_LEX_UNIT *unit= &thd->lex->unit;
  347. ha_rows idx= 0;
  348. Protocol *protocol= thd->protocol;
  349. unit->set_limit(sel);
  350. void *iterator;
  351. for (iterator= history.new_iterator();
  352. iterator != NULL;
  353. iterator= history.iterator_next(iterator))
  354. {
  355. prof= history.iterator_value(iterator);
  356. String elapsed;
  357. PROF_MEASUREMENT *ps= prof->profile_start;
  358. PROF_MEASUREMENT *pe= prof->profile_end;
  359. if (++idx <= unit->offset_limit_cnt)
  360. continue;
  361. if (idx > unit->select_limit_cnt)
  362. break;
  363. protocol->prepare_for_resend();
  364. protocol->store((uint32)(prof->profiling_query_id));
  365. protocol->store((double)(pe->time_usecs - ps->time_usecs)/(1000.0*1000),
  366. (uint32) TIME_FLOAT_DIGITS-1, &elapsed);
  367. if (prof->query_source != NULL)
  368. protocol->store(prof->query_source, strlen(prof->query_source),
  369. system_charset_info);
  370. else
  371. protocol->store_null();
  372. if (protocol->write())
  373. DBUG_RETURN(TRUE);
  374. }
  375. my_eof(thd);
  376. DBUG_RETURN(FALSE);
  377. }
  378. /**
  379. At a point in execution where we know the query source, save the text
  380. of it in the query profile.
  381. This must be called exactly once per descrete statement.
  382. */
  383. void PROFILING::set_query_source(char *query_source_arg, uint query_length_arg)
  384. {
  385. DBUG_ENTER("PROFILING::set_query_source");
  386. if (! enabled)
  387. DBUG_VOID_RETURN;
  388. if (current != NULL)
  389. current->set_query_source(query_source_arg, query_length_arg);
  390. else
  391. DBUG_PRINT("info", ("no current profile to send query source to"));
  392. DBUG_VOID_RETURN;
  393. }
  394. /**
  395. Fill the information schema table, "query_profile", as defined in show.cc .
  396. There are two ways to get to this function: Selecting from the information
  397. schema, and a SHOW command.
  398. */
  399. int PROFILING::fill_statistics_info(THD *thd, TABLE_LIST *tables, Item *cond)
  400. {
  401. DBUG_ENTER("PROFILING::fill_statistics_info");
  402. TABLE *table= tables->table;
  403. ulonglong row_number= 0;
  404. QUERY_PROFILE *query;
  405. /* Go through each query in this thread's stored history... */
  406. void *history_iterator;
  407. for (history_iterator= history.new_iterator();
  408. history_iterator != NULL;
  409. history_iterator= history.iterator_next(history_iterator))
  410. {
  411. query= history.iterator_value(history_iterator);
  412. /*
  413. Because we put all profiling info into a table that may be reordered, let
  414. us also include a numbering of each state per query. The query_id and
  415. the "seq" together are unique.
  416. */
  417. ulonglong seq;
  418. void *entry_iterator;
  419. PROF_MEASUREMENT *entry, *previous= NULL;
  420. /* ...and for each query, go through all its state-change steps. */
  421. for (seq= 0, entry_iterator= query->entries.new_iterator();
  422. entry_iterator != NULL;
  423. entry_iterator= query->entries.iterator_next(entry_iterator),
  424. seq++, previous=entry, row_number++)
  425. {
  426. entry= query->entries.iterator_value(entry_iterator);
  427. /* Skip the first. We count spans of fence, not fence-posts. */
  428. if (previous == NULL) continue;
  429. if (thd->lex->sql_command == SQLCOM_SHOW_PROFILE)
  430. {
  431. /*
  432. We got here via a SHOW command. That means that we stored
  433. information about the query we wish to show and that isn't
  434. in a WHERE clause at a higher level to filter out rows we
  435. wish to exclude.
  436. Because that functionality isn't available in the server yet,
  437. we must filter here, at the wrong level. Once one can con-
  438. struct where and having conditions at the SQL layer, then this
  439. condition should be ripped out.
  440. */
  441. if (thd->lex->profile_query_id == 0) /* 0 == show final query */
  442. {
  443. if (query != last)
  444. continue;
  445. }
  446. else
  447. {
  448. if (thd->lex->profile_query_id != query->profiling_query_id)
  449. continue;
  450. }
  451. }
  452. /* Set default values for this row. */
  453. restore_record(table, s->default_values);
  454. /*
  455. The order of these fields is set by the query_profile_statistics_info
  456. array.
  457. */
  458. table->field[0]->store((ulonglong) query->profiling_query_id, TRUE);
  459. table->field[1]->store((ulonglong) seq, TRUE); /* the step in the sequence */
  460. /*
  461. This entry, n, has a point in time, T(n), and a status phrase, S(n).
  462. The status phrase S(n) describes the period of time that begins at
  463. T(n). The previous status phrase S(n-1) describes the period of time
  464. that starts at T(n-1) and ends at T(n). Since we want to describe the
  465. time that a status phrase took T(n)-T(n-1), this line must describe the
  466. previous status.
  467. */
  468. table->field[2]->store(previous->status, strlen(previous->status),
  469. system_charset_info);
  470. my_decimal duration_decimal;
  471. double2my_decimal(E_DEC_FATAL_ERROR,
  472. (entry->time_usecs-previous->time_usecs)/(1000.0*1000),
  473. &duration_decimal);
  474. table->field[3]->store_decimal(&duration_decimal);
  475. #ifdef HAVE_GETRUSAGE
  476. my_decimal cpu_utime_decimal, cpu_stime_decimal;
  477. double2my_decimal(E_DEC_FATAL_ERROR,
  478. RUSAGE_DIFF_USEC(entry->rusage.ru_utime,
  479. previous->rusage.ru_utime) /
  480. (1000.0*1000),
  481. &cpu_utime_decimal);
  482. double2my_decimal(E_DEC_FATAL_ERROR,
  483. RUSAGE_DIFF_USEC(entry->rusage.ru_stime,
  484. previous->rusage.ru_stime) /
  485. (1000.0*1000),
  486. &cpu_stime_decimal);
  487. table->field[4]->store_decimal(&cpu_utime_decimal);
  488. table->field[5]->store_decimal(&cpu_stime_decimal);
  489. table->field[4]->set_notnull();
  490. table->field[5]->set_notnull();
  491. #else
  492. /* TODO: Add CPU-usage info for non-BSD systems */
  493. #endif
  494. #ifdef HAVE_GETRUSAGE
  495. table->field[6]->store((uint32)(entry->rusage.ru_nvcsw -
  496. previous->rusage.ru_nvcsw));
  497. table->field[6]->set_notnull();
  498. table->field[7]->store((uint32)(entry->rusage.ru_nivcsw -
  499. previous->rusage.ru_nivcsw));
  500. table->field[7]->set_notnull();
  501. #else
  502. /* TODO: Add context switch info for non-BSD systems */
  503. #endif
  504. #ifdef HAVE_GETRUSAGE
  505. table->field[8]->store((uint32)(entry->rusage.ru_inblock -
  506. previous->rusage.ru_inblock));
  507. table->field[8]->set_notnull();
  508. table->field[9]->store((uint32)(entry->rusage.ru_oublock -
  509. previous->rusage.ru_oublock));
  510. table->field[9]->set_notnull();
  511. #else
  512. /* TODO: Add block IO info for non-BSD systems */
  513. #endif
  514. #ifdef HAVE_GETRUSAGE
  515. table->field[10]->store((uint32)(entry->rusage.ru_msgsnd -
  516. previous->rusage.ru_msgsnd), true);
  517. table->field[10]->set_notnull();
  518. table->field[11]->store((uint32)(entry->rusage.ru_msgrcv -
  519. previous->rusage.ru_msgrcv), true);
  520. table->field[11]->set_notnull();
  521. #else
  522. /* TODO: Add message info for non-BSD systems */
  523. #endif
  524. #ifdef HAVE_GETRUSAGE
  525. table->field[12]->store((uint32)(entry->rusage.ru_majflt -
  526. previous->rusage.ru_majflt), true);
  527. table->field[12]->set_notnull();
  528. table->field[13]->store((uint32)(entry->rusage.ru_minflt -
  529. previous->rusage.ru_minflt), true);
  530. table->field[13]->set_notnull();
  531. #else
  532. /* TODO: Add page fault info for non-BSD systems */
  533. #endif
  534. #ifdef HAVE_GETRUSAGE
  535. table->field[14]->store((uint32)(entry->rusage.ru_nswap -
  536. previous->rusage.ru_nswap), true);
  537. table->field[14]->set_notnull();
  538. #else
  539. /* TODO: Add swap info for non-BSD systems */
  540. #endif
  541. /* Emit the location that started this step, not that ended it. */
  542. if ((previous->function != NULL) && (previous->file != NULL))
  543. {
  544. table->field[15]->store(previous->function, strlen(previous->function),
  545. system_charset_info);
  546. table->field[15]->set_notnull();
  547. table->field[16]->store(previous->file, strlen(previous->file), system_charset_info);
  548. table->field[16]->set_notnull();
  549. table->field[17]->store(previous->line, true);
  550. table->field[17]->set_notnull();
  551. }
  552. if (schema_table_store_record(thd, table))
  553. DBUG_RETURN(1);
  554. }
  555. }
  556. DBUG_RETURN(0);
  557. }
  558. #endif /* ENABLED_PROFILING */