mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-20 05:52:27 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/mysql-test/r/profiling.result

416 lines
16 KiB
Text
Raw Normal View History

Unify profiling SHOW code and INFORMATION_SCHEMA code for profiling. Also, Bug#26938: profiling client hang if used before enabled In the SHOW command, not sending header data because we had no rows to send was a protocol violation. Porting the SHOW PROFILE command to use the Information Schema table avoids that problem.
2007-07-02 13:27:39 +02:00
show profiles;
Query_ID Duration Query
show profile all;
Status Duration CPU_user CPU_system Context_voluntary Context_involuntary Block_ops_in Block_ops_out Messages_sent Messages_received Page_faults_major Page_faults_minor Swaps Source_function Source_file Source_line
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.
2007-02-22 16:03:08 +01:00
show session variables like 'profil%';
Variable_name Value
profiling OFF
profiling_history_size 15
select @@profiling;
@@profiling
0
set global profiling = ON;
ERROR HY000: Variable 'profiling' is a SESSION variable and can't be used with SET GLOBAL
set global profiling_history_size=100;
show global variables like 'profil%';
Variable_name Value
profiling OFF
profiling_history_size 100
set session profiling = ON;
set session profiling_history_size=30;
show session variables like 'profil%';
Variable_name Value
profiling ON
profiling_history_size 30
select @@profiling;
@@profiling
1
create table t1 (
a int,
b int
);
insert into t1 values (1,1), (2,null), (3, 4);
insert into t1 values (5,1), (6,null), (7, 4);
insert into t1 values (1,1), (2,null), (3, 4);
insert into t1 values (5,1), (6,null), (7, 4);
select max(x) from (select sum(a) as x from t1 group by b) as teeone;
max(x)
20
insert into t1 select * from t1;
select count(*) from t1;
count(*)
24
insert into t1 select * from t1;
insert into t1 select * from t1;
insert into t1 select * from t1;
select count(*) from t1;
count(*)
192
insert into t1 select * from t1;
insert into t1 select * from t1;
insert into t1 select * from t1;
select count(*) from t1;
count(*)
1536
select sum(a) from t1;
sum(a)
6144
select sum(a) from t1 group by b;
sum(a)
2048
1536
2560
select sum(a) + sum(b) from t1 group by b;
sum(a) + sum(b)
NULL
2048
4608
select max(x) from (select sum(a) as x from t1 group by b) as teeone;
max(x)
2560
select '012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890' as big_string;
big_string
012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890
show profiles;
Query_ID Duration Query
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
1 # set session profiling_history_size=30
2 # show session variables like 'profil%'
3 # select @@profiling
4 # create table t1 (
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.
2007-02-22 16:03:08 +01:00
a int,
b int
)
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
5 # insert into t1 values (1,1), (2,null), (3, 4)
6 # insert into t1 values (5,1), (6,null), (7, 4)
7 # insert into t1 values (1,1), (2,null), (3, 4)
8 # insert into t1 values (5,1), (6,null), (7, 4)
9 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
10 # insert into t1 select * from t1
11 # select count(*) from t1
12 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
13 # insert into t1 select * from t1
14 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
15 # select count(*) from t1
16 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
17 # insert into t1 select * from t1
18 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
19 # select count(*) from t1
20 # select sum(a) from t1
21 # select sum(a) from t1 group by b
22 # select sum(a) + sum(b) from t1 group by b
23 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
24 # select '0123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345
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.
2007-02-22 16:03:08 +01:00
show profile for query 15;
show profile cpu for query 15;
show profile cpu, block io for query 15;
show profile cpu for query 9 limit 2 offset 2;
show profile cpu for query 10 limit 0;
show profile cpu for query 65534;
show profile memory;
show profile block io;
show profile context switches;
show profile page faults;
show profile ipc;
show profile swaps limit 1 offset 2;
show profile source;
show profile all for query 0 limit 0;
show profile all for query 15;
select * from information_schema.profiling;
select query_id, state, duration from information_schema.profiling;
select query_id, sum(duration) from information_schema.profiling group by query_id;
select query_id, count(*) from information_schema.profiling group by query_id;
select sum(duration) from information_schema.profiling;
select query_id, count(*), sum(duration) from information_schema.profiling group by query_id;
select CPU_user, CPU_system, Context_voluntary, Context_involuntary, Block_ops_in, Block_ops_out, Messages_sent, Messages_received, Page_faults_major, Page_faults_minor, Swaps, Source_function, Source_file, Source_line from information_schema.profiling;
drop table if exists t1, t2, t3;
Warnings:
Note 1051 Unknown table 't2'
Note 1051 Unknown table 't3'
create table t1 (id int );
create table t2 (id int not null);
create table t3 (id int not null primary key);
insert into t1 values (1), (2), (3);
insert into t2 values (1), (2), (3);
insert into t3 values (1), (2), (3);
show profiles;
Query_ID Duration Query
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
10 # insert into t1 select * from t1
11 # select count(*) from t1
12 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
13 # insert into t1 select * from t1
14 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
15 # select count(*) from t1
16 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
17 # insert into t1 select * from t1
18 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
19 # select count(*) from t1
20 # select sum(a) from t1
21 # select sum(a) from t1 group by b
22 # select sum(a) + sum(b) from t1 group by b
23 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
24 # select '0123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345
25 # select * from information_schema.profiling
26 # select query_id, state, duration from information_schema.profiling
27 # select query_id, sum(duration) from information_schema.profiling group by query_id
28 # select query_id, count(*) from information_schema.profiling group by query_id
29 # select sum(duration) from information_schema.profiling
30 # select query_id, count(*), sum(duration) from information_schema.profiling group by query_id
31 # select CPU_user, CPU_system, Context_voluntary, Context_involuntary, Block_ops_in, Block_ops_out, Messages_sent, Messages_received, Page_faults_major, Page_faults_minor, Swaps, Source_function, Source_file, Source_line from information_schema.profiling
32 # drop table if exists t1, t2, t3
33 # SHOW WARNINGS
34 # create table t1 (id int )
35 # create table t2 (id int not null)
36 # create table t3 (id int not null primary key)
37 # insert into t1 values (1), (2), (3)
38 # insert into t2 values (1), (2), (3)
39 # insert into t3 values (1), (2), (3)
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.
2007-02-22 16:03:08 +01:00
select * from t1;
id
1
2
3
show profiles;
Query_ID Duration Query
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
11 # select count(*) from t1
12 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
13 # insert into t1 select * from t1
14 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
15 # select count(*) from t1
16 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
17 # insert into t1 select * from t1
18 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
19 # select count(*) from t1
20 # select sum(a) from t1
21 # select sum(a) from t1 group by b
22 # select sum(a) + sum(b) from t1 group by b
23 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
24 # select '0123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345
25 # select * from information_schema.profiling
26 # select query_id, state, duration from information_schema.profiling
27 # select query_id, sum(duration) from information_schema.profiling group by query_id
28 # select query_id, count(*) from information_schema.profiling group by query_id
29 # select sum(duration) from information_schema.profiling
30 # select query_id, count(*), sum(duration) from information_schema.profiling group by query_id
31 # select CPU_user, CPU_system, Context_voluntary, Context_involuntary, Block_ops_in, Block_ops_out, Messages_sent, Messages_received, Page_faults_major, Page_faults_minor, Swaps, Source_function, Source_file, Source_line from information_schema.profiling
32 # drop table if exists t1, t2, t3
33 # SHOW WARNINGS
34 # create table t1 (id int )
35 # create table t2 (id int not null)
36 # create table t3 (id int not null primary key)
37 # insert into t1 values (1), (2), (3)
38 # insert into t2 values (1), (2), (3)
39 # insert into t3 values (1), (2), (3)
40 # select * from t1
The test suite erroneously removes backslashes.
2007-04-03 20:05:00 +02:00
This ^^ should end in "select * from t1;"
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.
2007-02-22 16:03:08 +01:00
delete from t1;
insert into t1 values (1), (2), (3);
insert into t1 values (1), (2), (3);
select * from t1;
id
1
2
3
1
2
3
show profiles;
Query_ID Duration Query
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
15 # select count(*) from t1
16 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
17 # insert into t1 select * from t1
18 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
19 # select count(*) from t1
20 # select sum(a) from t1
21 # select sum(a) from t1 group by b
22 # select sum(a) + sum(b) from t1 group by b
23 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
24 # select '0123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345
25 # select * from information_schema.profiling
26 # select query_id, state, duration from information_schema.profiling
27 # select query_id, sum(duration) from information_schema.profiling group by query_id
28 # select query_id, count(*) from information_schema.profiling group by query_id
29 # select sum(duration) from information_schema.profiling
30 # select query_id, count(*), sum(duration) from information_schema.profiling group by query_id
31 # select CPU_user, CPU_system, Context_voluntary, Context_involuntary, Block_ops_in, Block_ops_out, Messages_sent, Messages_received, Page_faults_major, Page_faults_minor, Swaps, Source_function, Source_file, Source_line from information_schema.profiling
32 # drop table if exists t1, t2, t3
33 # SHOW WARNINGS
34 # create table t1 (id int )
35 # create table t2 (id int not null)
36 # create table t3 (id int not null primary key)
37 # insert into t1 values (1), (2), (3)
38 # insert into t2 values (1), (2), (3)
39 # insert into t3 values (1), (2), (3)
40 # select * from t1
41 # delete from t1
42 # insert into t1 values (1), (2), (3)
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.
2007-02-22 16:03:08 +01:00
43 # insert into t1 values (1), (2), (3)
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
44 # select * from t1
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.
2007-02-22 16:03:08 +01:00
set session profiling = OFF;
select sum(id) from t1;
sum(id)
12
show profiles;
Query_ID Duration Query
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
15 # select count(*) from t1
16 # insert into t1 select * from t1
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.
2007-02-22 16:03:08 +01:00
17 # insert into t1 select * from t1
18 # insert into t1 select * from t1
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
19 # select count(*) from t1
20 # select sum(a) from t1
21 # select sum(a) from t1 group by b
22 # select sum(a) + sum(b) from t1 group by b
23 # select max(x) from (select sum(a) as x from t1 group by b) as teeone
24 # select '0123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345678900123456789001234567890012345
25 # select * from information_schema.profiling
26 # select query_id, state, duration from information_schema.profiling
27 # select query_id, sum(duration) from information_schema.profiling group by query_id
28 # select query_id, count(*) from information_schema.profiling group by query_id
29 # select sum(duration) from information_schema.profiling
30 # select query_id, count(*), sum(duration) from information_schema.profiling group by query_id
31 # select CPU_user, CPU_system, Context_voluntary, Context_involuntary, Block_ops_in, Block_ops_out, Messages_sent, Messages_received, Page_faults_major, Page_faults_minor, Swaps, Source_function, Source_file, Source_line from information_schema.profiling
32 # drop table if exists t1, t2, t3
33 # SHOW WARNINGS
34 # create table t1 (id int )
35 # create table t2 (id int not null)
36 # create table t3 (id int not null primary key)
37 # insert into t1 values (1), (2), (3)
38 # insert into t2 values (1), (2), (3)
39 # insert into t3 values (1), (2), (3)
40 # select * from t1
41 # delete from t1
42 # insert into t1 values (1), (2), (3)
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.
2007-02-22 16:03:08 +01:00
43 # insert into t1 values (1), (2), (3)
Changing the state of whether we're recording profiling information halfway through a query (as happens in "SET SESSION PROFILING = ...") has a few side-effects, the worst of which is a memory leak for prepared statements, which poke directly from the parser into the profiling code (we don't have the query text when we need it) and that overwrites a pointer to heap-allocated memory when the previous statement turns on profiling. Instead, now set a flag when we begin a new statement that tracks whether profiling is on _at the start_ of the query. Use that to track whether we gather info. Additionally, use that AND use the state of the profiling variable after the end of a query to know whether to store information about the query that just finished.
2007-04-06 15:15:18 +02:00
44 # select * from t1
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.
2007-02-22 16:03:08 +01:00
set session profiling = ON;
select @@profiling;
@@profiling
1
create function f1() returns varchar(50) return 'hello';
select @@profiling;
@@profiling
1
select * from t1 where id <> f1();
id
1
2
3
1
2
3
select @@profiling;
@@profiling
1
Insert profiling instructions into s-p code to make each statement be profiled separately. Expand the time formats in i_s.profiling to wide enough for larger numbers.
2007-11-13 15:46:17 +01:00
set session profiling = OFF;
drop table if exists profile_log;
Warnings:
Note 1051 Unknown table 'profile_log'
create table profile_log (how_many int);
drop procedure if exists p1;
drop procedure if exists p2;
drop procedure if exists p3;
Correct complaints of reviewers and clean up test.
2007-12-14 14:57:37 +01:00
create procedure p1 ()
modifies sql data
begin
set profiling = ON;
select 'This p1 should show up in profiling';
insert into profile_log select count(*) from information_schema.profiling;
end//
create procedure p2()
deterministic
begin
set profiling = ON;
call p1();
select 'This p2 should show up in profiling';
end//
create procedure p3 ()
reads sql data
begin
set profiling = ON;
select 'This p3 should show up in profiling';
show profile;
end//
Insert profiling instructions into s-p code to make each statement be profiled separately. Expand the time formats in i_s.profiling to wide enough for larger numbers.
2007-11-13 15:46:17 +01:00
first call to p1
call p1;
select * from profile_log;
second call to p1
call p1;
select * from profile_log;
third call to p1
call p1;
select * from profile_log;
set session profiling = OFF;
call p2;
set session profiling = OFF;
call p3;
show profiles;
drop procedure if exists p1;
drop procedure if exists p2;
drop procedure if exists p3;
drop table if exists profile_log;
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.
2007-02-22 16:03:08 +01:00
set session profiling = ON;
drop table if exists t2;
create table t2 (id int not null);
create trigger t2_bi before insert on t2 for each row set @x=0;
select @@profiling;
@@profiling
1
insert into t2 values (1), (2), (3);
select @@profiling;
@@profiling
1
set session profiling = ON;
drop table if exists t1, t2;
create table t1 (id int not null primary key);
create table t2 (id int not null primary key, id1 int not null);
select @@profiling;
@@profiling
1
alter table t2 add foreign key (id1) references t1 (id) on delete cascade;
select @@profiling;
@@profiling
1
lock table t1 write;
select @@profiling;
@@profiling
1
unlock table;
select @@profiling;
@@profiling
1
set autocommit=0;
select @@profiling, @@autocommit;
@@profiling @@autocommit
1 0
begin;
select @@profiling;
@@profiling
1
insert into t1 values (1);
insert into t2 values (1,1);
testing referential integrity cascade
delete from t1 where id = 1;
select @@profiling;
@@profiling
1
testing rollback
rollback;
select @@profiling;
@@profiling
1
testing commit
begin;
select @@profiling;
@@profiling
1
commit;
select @@profiling;
@@profiling
1
drop table if exists t1, t2, t3;
drop view if exists v1;
Warnings:
Note 1051 Unknown table 'test.v1'
drop function if exists f1;
set session profiling = OFF;
End of 5.0 tests
Reference in a new issue Copy permalink