mariadb/sql/sql_class.cc

4645 lines
134 KiB
C++
Raw Normal View History

/* Copyright (C) 2000-2008 MySQL AB, 2008-2009 Sun Microsystems, Inc.
2000-07-31 21:29:14 +02:00
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
2000-07-31 21:29:14 +02:00
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
2000-07-31 21:29:14 +02:00
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/*****************************************************************************
**
** This file implements classes defined in sql_class.h
** Especially the classes to handle a result from a select
**
*****************************************************************************/
#ifdef USE_PRAGMA_IMPLEMENTATION
2000-07-31 21:29:14 +02:00
#pragma implementation // gcc: Class implementation
#endif
#include "my_global.h" /* NO_EMBEDDED_ACCESS_CHECKS */
#include "sql_priv.h"
#include "unireg.h" // REQUIRED: for other includes
#include "sql_class.h"
#include "lock.h" // unlock_global_read_lock, mysql_unlock_tables
#include "sql_cache.h" // query_cache_abort
#include "sql_base.h" // close_thread_tables
#include "sql_time.h" // date_time_format_copy
#include "sql_acl.h" // NO_ACCESS,
// acl_getroot_no_password
#include "sql_base.h" // close_temporary_tables
#include "sql_handler.h" // mysql_ha_cleanup
#include "rpl_rli.h"
#include "rpl_filter.h"
#include "rpl_record.h"
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
#include "slave.h"
#include <my_bitmap.h>
#include "log_event.h"
#include "sql_audit.h"
2000-07-31 21:29:14 +02:00
#include <m_ctype.h>
#include <sys/stat.h>
#include <thr_alarm.h>
2000-07-31 21:29:14 +02:00
#ifdef __WIN__
#include <io.h>
#endif
#include <mysys_err.h>
#include <limits.h>
2000-07-31 21:29:14 +02:00
#include "sp_rcontext.h"
#include "sp_cache.h"
#include "transaction.h"
#include "debug_sync.h"
#include "sql_parse.h" // is_update_query
/*
The following is used to initialise Table_ident with a internal
table name
*/
char internal_table_name[2]= "*";
char empty_c_string[1]= {0}; /* used for not defined db */
const char * const THD::DEFAULT_WHERE= "field list";
2000-07-31 21:29:14 +02:00
/*****************************************************************************
** Instansiate templates
*****************************************************************************/
#ifdef HAVE_EXPLICIT_TEMPLATE_INSTANTIATION
2000-07-31 21:29:14 +02:00
/* Used templates */
template class List<Key>;
template class List_iterator<Key>;
template class List<Key_part_spec>;
template class List_iterator<Key_part_spec>;
2000-07-31 21:29:14 +02:00
template class List<Alter_drop>;
template class List_iterator<Alter_drop>;
template class List<Alter_column>;
template class List_iterator<Alter_column>;
#endif
/****************************************************************************
** User variables
****************************************************************************/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
extern "C" uchar *get_var_key(user_var_entry *entry, size_t *length,
my_bool not_used __attribute__((unused)))
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
*length= entry->name.length;
return (uchar*) entry->name.str;
2000-07-31 21:29:14 +02:00
}
extern "C" void free_user_var(user_var_entry *entry)
2000-07-31 21:29:14 +02:00
{
char *pos= (char*) entry+ALIGN_SIZE(sizeof(*entry));
if (entry->value && entry->value != pos)
my_free(entry->value, MYF(0));
my_free((char*) entry,MYF(0));
}
bool Key_part_spec::operator==(const Key_part_spec& other) const
{
return length == other.length &&
!my_strcasecmp(system_charset_info, field_name.str,
other.field_name.str);
}
5.1 version of a fix and test cases for bugs: Bug#4968 ""Stored procedure crash if cursor opened on altered table" Bug#6895 "Prepared Statements: ALTER TABLE DROP COLUMN does nothing" Bug#19182 "CREATE TABLE bar (m INT) SELECT n FROM foo; doesn't work from stored procedure." Bug#19733 "Repeated alter, or repeated create/drop, fails" Bug#22060 "ALTER TABLE x AUTO_INCREMENT=y in SP crashes server" Bug#24879 "Prepared Statements: CREATE TABLE (UTF8 KEY) produces a growing key length" (this bug is not fixed in 5.0) Re-execution of CREATE DATABASE, CREATE TABLE and ALTER TABLE statements in stored routines or as prepared statements caused incorrect results (and crashes in versions prior to 5.0.25). In 5.1 the problem occured only for CREATE DATABASE, CREATE TABLE SELECT and CREATE TABLE with INDEX/DATA DIRECTOY options). The problem of bugs 4968, 19733, 19282 and 6895 was that functions mysql_prepare_table, mysql_create_table and mysql_alter_table are not re-execution friendly: during their operation they modify contents of LEX (members create_info, alter_info, key_list, create_list), thus making the LEX unusable for the next execution. In particular, these functions removed processed columns and keys from create_list, key_list and drop_list. Search the code in sql_table.cc for drop_it.remove() and similar patterns to find evidence. The fix is to supply to these functions a usable copy of each of the above structures at every re-execution of an SQL statement. To simplify memory management, LEX::key_list and LEX::create_list were added to LEX::alter_info, a fresh copy of which is created for every execution. The problem of crashing bug 22060 stemmed from the fact that the above metnioned functions were not only modifying HA_CREATE_INFO structure in LEX, but also were changing it to point to areas in volatile memory of the execution memory root. The patch solves this problem by creating and using an on-stack copy of HA_CREATE_INFO in mysql_execute_command. Additionally, this patch splits the part of mysql_alter_table that analizes and rewrites information from the parser into a separate function - mysql_prepare_alter_table, in analogy with mysql_prepare_table, which is renamed to mysql_prepare_create_table.
2007-05-28 13:30:01 +02:00
/**
Construct an (almost) deep copy of this key. Only those
elements that are known to never change are not copied.
If out of memory, a partial copy is returned and an error is set
in THD.
*/
Key::Key(const Key &rhs, MEM_ROOT *mem_root)
:type(rhs.type),
key_create_info(rhs.key_create_info),
columns(rhs.columns, mem_root),
name(rhs.name),
generated(rhs.generated)
{
list_copy_and_replace_each_value(columns, mem_root);
}
/**
Construct an (almost) deep copy of this foreign key. Only those
elements that are known to never change are not copied.
If out of memory, a partial copy is returned and an error is set
in THD.
*/
Foreign_key::Foreign_key(const Foreign_key &rhs, MEM_ROOT *mem_root)
5.1 version of a fix and test cases for bugs: Bug#4968 ""Stored procedure crash if cursor opened on altered table" Bug#6895 "Prepared Statements: ALTER TABLE DROP COLUMN does nothing" Bug#19182 "CREATE TABLE bar (m INT) SELECT n FROM foo; doesn't work from stored procedure." Bug#19733 "Repeated alter, or repeated create/drop, fails" Bug#22060 "ALTER TABLE x AUTO_INCREMENT=y in SP crashes server" Bug#24879 "Prepared Statements: CREATE TABLE (UTF8 KEY) produces a growing key length" (this bug is not fixed in 5.0) Re-execution of CREATE DATABASE, CREATE TABLE and ALTER TABLE statements in stored routines or as prepared statements caused incorrect results (and crashes in versions prior to 5.0.25). In 5.1 the problem occured only for CREATE DATABASE, CREATE TABLE SELECT and CREATE TABLE with INDEX/DATA DIRECTOY options). The problem of bugs 4968, 19733, 19282 and 6895 was that functions mysql_prepare_table, mysql_create_table and mysql_alter_table are not re-execution friendly: during their operation they modify contents of LEX (members create_info, alter_info, key_list, create_list), thus making the LEX unusable for the next execution. In particular, these functions removed processed columns and keys from create_list, key_list and drop_list. Search the code in sql_table.cc for drop_it.remove() and similar patterns to find evidence. The fix is to supply to these functions a usable copy of each of the above structures at every re-execution of an SQL statement. To simplify memory management, LEX::key_list and LEX::create_list were added to LEX::alter_info, a fresh copy of which is created for every execution. The problem of crashing bug 22060 stemmed from the fact that the above metnioned functions were not only modifying HA_CREATE_INFO structure in LEX, but also were changing it to point to areas in volatile memory of the execution memory root. The patch solves this problem by creating and using an on-stack copy of HA_CREATE_INFO in mysql_execute_command. Additionally, this patch splits the part of mysql_alter_table that analizes and rewrites information from the parser into a separate function - mysql_prepare_alter_table, in analogy with mysql_prepare_table, which is renamed to mysql_prepare_create_table.
2007-05-28 13:30:01 +02:00
:Key(rhs),
ref_table(rhs.ref_table),
ref_columns(rhs.ref_columns),
delete_opt(rhs.delete_opt),
update_opt(rhs.update_opt),
match_opt(rhs.match_opt)
{
list_copy_and_replace_each_value(ref_columns, mem_root);
}
/*
Test if a foreign key (= generated key) is a prefix of the given key
(ignoring key name, key type and order of columns)
NOTES:
This is only used to test if an index for a FOREIGN KEY exists
IMPLEMENTATION
We only compare field names
RETURN
0 Generated key is a prefix of other key
1 Not equal
*/
bool foreign_key_prefix(Key *a, Key *b)
{
/* Ensure that 'a' is the generated key */
if (a->generated)
{
if (b->generated && a->columns.elements > b->columns.elements)
swap_variables(Key*, a, b); // Put shorter key in 'a'
}
else
{
if (!b->generated)
return TRUE; // No foreign key
swap_variables(Key*, a, b); // Put generated key in 'a'
}
/* Test if 'a' is a prefix of 'b' */
if (a->columns.elements > b->columns.elements)
return TRUE; // Can't be prefix
List_iterator<Key_part_spec> col_it1(a->columns);
List_iterator<Key_part_spec> col_it2(b->columns);
const Key_part_spec *col1, *col2;
#ifdef ENABLE_WHEN_INNODB_CAN_HANDLE_SWAPED_FOREIGN_KEY_COLUMNS
while ((col1= col_it1++))
{
bool found= 0;
col_it2.rewind();
while ((col2= col_it2++))
{
if (*col1 == *col2)
{
found= TRUE;
break;
}
}
if (!found)
return TRUE; // Error
}
return FALSE; // Is prefix
#else
while ((col1= col_it1++))
{
col2= col_it2++;
if (!(*col1 == *col2))
return TRUE;
}
return FALSE; // Is prefix
#endif
}
2000-07-31 21:29:14 +02:00
/****************************************************************************
** Thread specific functions
****************************************************************************/
/*
The following functions form part of the C plugin API
*/
extern "C" int mysql_tmpfile(const char *prefix)
{
char filename[FN_REFLEN];
File fd = create_temp_file(filename, mysql_tmpdir, prefix,
#ifdef __WIN__
O_BINARY | O_TRUNC | O_SEQUENTIAL |
O_SHORT_LIVED |
#endif /* __WIN__ */
O_CREAT | O_EXCL | O_RDWR | O_TEMPORARY,
MYF(MY_WME));
if (fd >= 0) {
#ifndef __WIN__
/*
This can be removed once the following bug is fixed:
Bug #28903 create_temp_file() doesn't honor O_TEMPORARY option
(file not removed) (Unix)
*/
unlink(filename);
#endif /* !__WIN__ */
}
return fd;
}
extern "C"
int thd_in_lock_tables(const THD *thd)
{
return test(thd->in_lock_tables);
}
extern "C"
int thd_tablespace_op(const THD *thd)
{
return test(thd->tablespace_op);
}
extern "C"
const char *set_thd_proc_info(THD *thd, const char *info,
const char *calling_function,
const char *calling_file,
const unsigned int calling_line)
{
if (!thd)
thd= current_thd;
const char *old_info= thd->proc_info;
DBUG_PRINT("proc_info", ("%s:%d %s", calling_file, calling_line,
(info != NULL) ? info : "(null)"));
#if defined(ENABLED_PROFILING)
thd->profiling.status_change(info, calling_function, calling_file, calling_line);
#endif
thd->proc_info= info;
return old_info;
}
extern "C"
const char* thd_enter_cond(MYSQL_THD thd, mysql_cond_t *cond,
mysql_mutex_t *mutex, const char *msg)
{
if (!thd)
thd= current_thd;
return thd->enter_cond(cond, mutex, msg);
}
extern "C"
void thd_exit_cond(MYSQL_THD thd, const char *old_msg)
{
if (!thd)
thd= current_thd;
thd->exit_cond(old_msg);
return;
}
extern "C"
void **thd_ha_data(const THD *thd, const struct handlerton *hton)
{
return (void **) &thd->ha_data[hton->slot].ha_ptr;
}
extern "C"
long long thd_test_options(const THD *thd, long long test_options)
{
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
return thd->variables.option_bits & test_options;
}
extern "C"
int thd_sql_command(const THD *thd)
{
return (int) thd->lex->sql_command;
}
extern "C"
int thd_tx_isolation(const THD *thd)
{
return (int) thd->variables.tx_isolation;
}
extern "C"
void thd_inc_row_count(THD *thd)
{
thd->warning_info->inc_current_row_for_warning();
}
/**
Dumps a text description of a thread, its security context
(user, host) and the current query.
@param thd thread context
@param buffer pointer to preferred result buffer
@param length length of buffer
@param max_query_len how many chars of query to copy (0 for all)
@req LOCK_thread_count
@note LOCK_thread_count mutex is not necessary when the function is invoked on
the currently running thread (current_thd) or if the caller in some other
way guarantees that access to thd->query is serialized.
@return Pointer to string
*/
extern "C"
2007-04-13 19:23:02 +02:00
char *thd_security_context(THD *thd, char *buffer, unsigned int length,
unsigned int max_query_len)
{
String str(buffer, length, &my_charset_latin1);
const Security_context *sctx= &thd->main_security_ctx;
char header[64];
int len;
/*
The pointers thd->query and thd->proc_info might change since they are
being modified concurrently. This is acceptable for proc_info since its
values doesn't have to very accurate and the memory it points to is static,
but we need to attempt a snapshot on the pointer values to avoid using NULL
values. The pointer to thd->query however, doesn't point to static memory
and has to be protected by LOCK_thread_count or risk pointing to
uninitialized memory.
*/
const char *proc_info= thd->proc_info;
len= my_snprintf(header, sizeof(header),
"MySQL thread id %lu, query id %lu",
thd->thread_id, (ulong) thd->query_id);
str.length(0);
str.append(header, len);
if (sctx->host)
{
str.append(' ');
str.append(sctx->host);
}
if (sctx->ip)
{
str.append(' ');
str.append(sctx->ip);
}
if (sctx->user)
{
str.append(' ');
str.append(sctx->user);
}
if (proc_info)
{
str.append(' ');
str.append(proc_info);
}
mysql_mutex_lock(&thd->LOCK_thd_data);
if (thd->query())
{
if (max_query_len < 1)
len= thd->query_length();
else
len= min(thd->query_length(), max_query_len);
str.append('\n');
str.append(thd->query(), len);
}
mysql_mutex_unlock(&thd->LOCK_thd_data);
if (str.c_ptr_safe() == buffer)
return buffer;
/*
We have to copy the new string to the destination buffer because the string
was reallocated to a larger buffer to be able to fit.
*/
DBUG_ASSERT(buffer != NULL);
length= min(str.length(), length-1);
memcpy(buffer, str.c_ptr_quick(), length);
/* Make sure that the new string is null terminated */
buffer[length]= '\0';
return buffer;
}
2000-07-31 21:29:14 +02:00
/**
Backporting a patch for Bug#43138. That patch had been already backported to 5.1 partially. This patch brings what was left to mysql-next-mr. Original revisions in 6.0: ------------------------------------------------------------ revno: 2617.31.26 committer: Alexander Nozdrin <alik@sun.com> branch nick: 6.0-rt-bug43138.3 timestamp: Thu 2009-04-30 19:31:30 +0400 message: Fix for Bug#43138: DROP DATABASE failure does not clean up message list. The problem was that the high-level function mysql_rm_db() invoked low-level mysql_rm_table_part2(), which reported low-level error (Unknown table) if SE refused to delete a table. Also when mysql_rm_table_part2() reported an error, it didn't add corresponding warning into the list (because it is used from other places where such behaviour is required). The fix is to 1. Remove no_warnings_for_error usage from sql_table.cc 2. Improve internal error handler support in THD, so that a stack of error handlers is allowed. 3. Create an internal error handler (Drop_table_error_handler) to silence useless warnings. 4. Use the handler in DROP DATABASE and DROP TABLE statements. ------------------------------------------------------------ revno: 2617.69.38 committer: Alexander Nozdrin <alik@sun.com> branch nick: mysql-next-bugfixing-bug37431 timestamp: Mon 2009-08-24 21:52:09 +0400 message: A test case for Bug#37431 (DROP TABLE does not report errors correctly). ------------------------------------------------------------ revno: 2617.31.29 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-runtime timestamp: Fri 2009-05-01 17:37:34 +0400 message: Follow-up for fix for bug "Bug#43138: DROP DATABASE failure does not clean up message list". Fixed drop.test failure under non-debug server by moving part of test dependent on debug-only feature to separate .test file, which won't be run for non-debug versions of server. ------------------------------------------------------------ revno: 2617.45.17 committer: Sergei Golubchik <serg@mysql.com> branch nick: 6.0-maria timestamp: Wed 2009-05-13 20:08:58 +0200 message: followup for bug#43138 if delete fails with a permission denied error, we want to show it ------------------------------------------------------------ The patch was backported to 5.1 in scope of Bug#42364 by the following revision: ------------------------------------------------------------ revno: 2497.975.3 committer: Sergey Glukhov <Sergey.Glukhov@sun.com> branch nick: mysql-5.1-bugteam timestamp: Fri 2009-07-03 13:22:06 +0500 message: Bug#42364 SHOW ERRORS returns empty resultset after dropping non existent table enabled message storing into error message list for 'drop table' command ------------------------------------------------------------
2009-10-22 20:22:53 +02:00
Implementation of Drop_table_error_handler::handle_condition().
The reason in having this implementation is to silence technical low-level
warnings during DROP TABLE operation. Currently we don't want to expose
the following warnings during DROP TABLE:
- Some of table files are missed or invalid (the table is going to be
deleted anyway, so why bother that something was missed);
- A trigger associated with the table does not have DEFINER (One of the
MySQL specifics now is that triggers are loaded for the table being
dropped. So, we may have a warning that trigger does not have DEFINER
attribute during DROP TABLE operation).
@return TRUE if the condition is handled.
*/
bool Drop_table_error_handler::handle_condition(THD *thd,
uint sql_errno,
const char* sqlstate,
MYSQL_ERROR::enum_warning_level level,
const char* msg,
MYSQL_ERROR ** cond_hdl)
{
*cond_hdl= NULL;
return ((sql_errno == EE_DELETE && my_errno == ENOENT) ||
sql_errno == ER_TRG_NO_DEFINER);
}
THD::THD()
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
:Statement(&main_lex, &main_mem_root, CONVENTIONAL_EXECUTION,
/* statement id */ 0),
rli_fake(0),
lock_id(&main_lock_id),
user_time(0), in_sub_stmt(0),
sql_log_bin_toplevel(false),
binlog_unsafe_warning_flags(0), binlog_table_maps(0),
2008-08-07 03:33:01 +02:00
table_map_for_update(0),
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
arg_of_last_insert_id_function(FALSE),
first_successful_insert_id_in_prev_stmt(0),
first_successful_insert_id_in_prev_stmt_for_binlog(0),
first_successful_insert_id_in_cur_stmt(0),
stmt_depends_on_first_successful_insert_id_in_prev_stmt(FALSE),
examined_row_count(0),
warning_info(&main_warning_info),
stmt_da(&main_da),
2007-04-16 18:16:17 +02:00
is_fatal_error(0),
2007-08-02 02:59:41 +02:00
transaction_rollback_request(0),
is_fatal_sub_stmt_error(0),
2007-04-13 19:23:02 +02:00
rand_used(0),
time_zone_used(0),
in_lock_tables(0),
bootstrap(0),
derived_tables_processing(FALSE),
spcont(NULL),
m_parser_state(NULL),
#if defined(ENABLED_DEBUG_SYNC)
2009-10-14 10:25:39 +02:00
debug_sync_control(0),
#endif /* defined(ENABLED_DEBUG_SYNC) */
main_warning_info(0)
2000-07-31 21:29:14 +02:00
{
ulong tmp;
mdl_context.init(this);
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
/*
Pass nominal parameters to init_alloc_root only to ensure that
the destructor works OK in case of an error. The main_mem_root
will be re-initialized in init_for_queries().
*/
init_sql_alloc(&main_mem_root, ALLOC_ROOT_MIN_BLOCK_SIZE, 0);
stmt_arena= this;
thread_stack= 0;
This will be pushed only after I fix the testsuite. This is the main commit for Worklog tasks: * A more dynamic binlog format which allows small changes (1064) * Log session variables in Query_log_event (1063) Below 5.0 means 5.0.0. MySQL 5.0 is able to replicate FOREIGN_KEY_CHECKS, UNIQUE_KEY_CHECKS (for speed), SQL_AUTO_IS_NULL, SQL_MODE. Not charsets (WL#1062), not some vars (I can only think of SQL_SELECT_LIMIT, which deserves a special treatment). Note that this works for queries, except LOAD DATA INFILE (for this it would have to wait for Dmitri's push of WL#874, which in turns waits for the present push, so... the deadlock must be broken!). Note that when Dmitri pushes WL#874 in 5.0.1, 5.0.0 won't be able to replicate a LOAD DATA INFILE from 5.0.1. Apart from that, the new binlog format is designed so that it can tolerate a little variation in the events (so that a 5.0.0 slave could replicate a 5.0.1 master, except for LOAD DATA INFILE unfortunately); that is, when I later add replication of charsets it should break nothing. And when I later add a UID to every event, it should break nothing. The main change brought by this patch is a new type of event, Format_description_log_event, which describes some lengthes in other event types. This event is needed for the master/slave/mysqlbinlog to understand a 5.0 log. Thanks to this event, we can later add more bytes to the header of every event without breaking compatibility. Inside Query_log_event, we have some additional dynamic format, as every Query_log_event can have a different number of status variables, stored as pairs (code, value); that's how SQL_MODE and session variables and catalog are stored. Like this, we can later add count of affected rows, charsets... and we can have options --don't-log-count-affected-rows if we want. MySQL 5.0 is able to run on 4.x relay logs, 4.x binlogs. Upgrading a 4.x master to 5.0 is ok (no need to delete binlogs), upgrading a 4.x slave to 5.0 is ok (no need to delete relay logs); so both can be "hot" upgrades. Upgrading a 3.23 master to 5.0 requires as much as upgrading it to 4.0. 3.23 and 4.x can't be slaves of 5.0. So downgrading from 5.0 to 4.x may be complicated. Log_event::log_pos is now the position of the end of the event, which is more useful than the position of the beginning. We take care about compatibility with <5.0 (in which log_pos is the beginning). I added a short test for replication of SQL_MODE and some other variables. TODO: - after committing this, merge the latest 5.0 into it - fix all tests - update the manual with upgrade notes.
2003-12-18 01:09:05 +01:00
catalog= (char*)"std"; // the only catalog we have for now
main_security_ctx.init();
security_ctx= &main_security_ctx;
A prerequisite patch for the fix for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". Introduce a notion of a sentinel to MDL_context. A sentinel is a ticket that separates all tickets in the context into two groups: before and after it. Currently we can have (and need) only one designated sentinel -- it separates all locks taken by LOCK TABLE or HANDLER statement, which must survive COMMIT and ROLLBACK and all other locks, which must be released at COMMIT or ROLLBACK. The tricky part is maintaining the sentinel up to date when someone release its corresponding ticket. This can happen, e.g. if someone issues DROP TABLE under LOCK TABLES (generally, see all calls to release_all_locks_for_name()). MDL_context::release_ticket() is modified to take care of it. ****** A fix and a test case for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". An attempt to mix HANDLER SQL statements, which are transaction- agnostic, an open multi-statement transaction, and DDL against the involved tables (in a concurrent connection) could lead to a deadlock. The deadlock would occur when HANDLER OPEN or HANDLER READ would have to wait on a conflicting metadata lock. If the connection that issued HANDLER statement also had other metadata locks (say, acquired in scope of a transaction), a classical deadlock situation of mutual wait could occur. Incompatible change: entering LOCK TABLES mode automatically closes all open HANDLERs in the current connection. Incompatible change: previously an attempt to wait on a lock in a connection that has an open HANDLER statement could wait indefinitely/deadlock. After this patch, an error ER_LOCK_DEADLOCK is produced. The idea of the fix is to merge thd->handler_mdl_context with the main mdl_context of the connection, used for transactional locks. This makes deadlock detection possible, since all waits with locks are "visible" and available to analysis in a single MDL context of the connection. Since HANDLER locks and transactional locks have a different life cycle -- HANDLERs are explicitly open and closed, and so are HANDLER locks, explicitly acquired and released, whereas transactional locks "accumulate" till the end of a transaction and are released only with COMMIT, ROLLBACK and ROLLBACK TO SAVEPOINT, a concept of "sentinel" was introduced to MDL_context. All locks, HANDLER and others, reside in the same linked list. However, a selected element of the list separates locks with different life cycle. HANDLER locks always reside at the end of the list, after the sentinel. Transactional locks are prepended to the beginning of the list, before the sentinel. Thus, ROLLBACK, COMMIT or ROLLBACK TO SAVEPOINT, only release those locks that reside before the sentinel. HANDLER locks must be released explicitly as part of HANDLER CLOSE statement, or an implicit close. The same approach with sentinel is also employed for LOCK TABLES locks. Since HANDLER and LOCK TABLES statement has never worked together, the implementation is made simple and only maintains one sentinel, which is used either for HANDLER locks, or for LOCK TABLES locks.
2009-12-22 17:09:15 +01:00
no_errors=password= 0;
query_start_used= 0;
count_cuted_fields= CHECK_FIELD_IGNORE;
killed= NOT_KILLED;
col_access=0;
is_slave_error= thread_specific_used= thread_temporary_used= FALSE;
my_hash_clear(&handler_tables_hash);
2000-07-31 21:29:14 +02:00
tmp_table=0;
used_tables=0;
cuted_fields= 0L;
sent_row_count= 0L;
limit_found_rows= 0;
m_row_count_func= -1;
statement_id_counter= 0UL;
// Must be reset to handle error with THD's created for init of mysqld
2003-12-04 20:08:26 +01:00
lex->current_select= 0;
2000-07-31 21:29:14 +02:00
start_time=(time_t) 0;
start_utime= prior_thr_create_utime= 0L;
utime_after_lock= 0L;
current_linfo = 0;
slave_thread = 0;
bzero(&variables, sizeof(variables));
thread_id= 0;
one_shot_set= 0;
file_id = 0;
query_id= 0;
query_name_consts= 0;
db_charset= global_system_variables.collation_database;
2005-01-16 13:16:23 +01:00
bzero(ha_data, sizeof(ha_data));
mysys_var=0;
binlog_evt_union.do_union= FALSE;
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
enable_slow_log= 0;
#ifndef DBUG_OFF
dbug_sentry=THD_SENTRY_MAGIC;
#endif
#ifndef EMBEDDED_LIBRARY
mysql_audit_init_thd(this);
net.vio=0;
#endif
client_capabilities= 0; // minimalistic client
ull=0;
system_thread= NON_SYSTEM_THREAD;
cleanup_done= abort_on_warning= no_warnings_for_error= 0;
peer_port= 0; // For SHOW PROCESSLIST
transaction.m_pending_rows_event= 0;
transaction.on= 1;
#ifdef SIGNAL_WITH_VIO_CLOSE
active_vio = 0;
#endif
mysql_mutex_init(key_LOCK_thd_data, &LOCK_thd_data, MY_MUTEX_INIT_FAST);
/* Variables with default values */
proc_info="login";
where= THD::DEFAULT_WHERE;
server_id = ::server_id;
slave_net = 0;
2000-07-31 21:29:14 +02:00
command=COM_CONNECT;
2003-07-18 16:57:21 +02:00
*scramble= '\0';
/* Call to init() below requires fully initialized Open_tables_state. */
init_open_tables_state(this, refresh_version);
init();
#if defined(ENABLED_PROFILING)
profiling.set_thd(this);
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
#endif
user_connect=(USER_CONN *)0;
my_hash_init(&user_vars, system_charset_info, USER_VARS_HASH_SIZE, 0, 0,
(my_hash_get_key) get_var_key,
(my_hash_free_key) free_user_var, 0);
sp_proc_cache= NULL;
sp_func_cache= NULL;
2003-01-30 18:39:54 +01:00
/* For user vars replication*/
if (opt_bin_log)
my_init_dynamic_array(&user_var_events,
2005-01-16 13:16:23 +01:00
sizeof(BINLOG_USER_VAR_EVENT *), 16, 16);
2003-01-30 18:39:54 +01:00
else
bzero((char*) &user_var_events, sizeof(user_var_events));
/* Protocol */
protocol= &protocol_text; // Default protocol
protocol_text.init(this);
protocol_binary.init(this);
tablespace_op=FALSE;
tmp= sql_rnd_with_mutex();
randominit(&rand, tmp + (ulong) &rand, tmp + (ulong) ::global_query_id);
substitute_null_with_insert_id = FALSE;
thr_lock_info_init(&lock_info); /* safety: will be reset after start */
thr_lock_owner_init(&main_lock_id, &lock_info);
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
m_internal_handler= NULL;
}
void THD::push_internal_handler(Internal_error_handler *handler)
{
if (m_internal_handler)
{
handler->m_prev_internal_handler= m_internal_handler;
m_internal_handler= handler;
}
else
{
m_internal_handler= handler;
}
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
}
bool THD::handle_condition(uint sql_errno,
const char* sqlstate,
MYSQL_ERROR::enum_warning_level level,
const char* msg,
MYSQL_ERROR ** cond_hdl)
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
{
if (!m_internal_handler)
{
*cond_hdl= NULL;
return FALSE;
}
for (Internal_error_handler *error_handler= m_internal_handler;
error_handler;
error_handler= error_handler->m_prev_internal_handler)
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
{
if (error_handler->handle_condition(this, sql_errno, sqlstate, level, msg,
cond_hdl))
{
return TRUE;
}
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
}
return FALSE;
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
}
Internal_error_handler *THD::pop_internal_handler()
Bug#8407 (Stored functions/triggers ignore exception handler) Bug 18914 (Calling certain SPs from triggers fail) Bug 20713 (Functions will not not continue for SQLSTATE VALUE '42S02') Bug 21825 (Incorrect message error deleting records in a table with a trigger for inserting) Bug 22580 (DROP TABLE in nested stored procedure causes strange dependency error) Bug 25345 (Cursors from Functions) This fix resolves a long standing issue originally reported with bug 8407, which affect the behavior of Stored Procedures, Stored Functions and Trigger in many different ways, causing symptoms reported by all the bugs listed. In all cases, the root cause of the problem traces back to 8407 and how the server locks tables involved with sub statements. Prior to this fix, the implementation of stored routines would: - compute the transitive closure of all the tables referenced by a top level statement - open and lock all the tables involved - execute the top level statement "transitive closure of tables" means collecting: - all the tables, - all the stored functions, - all the views, - all the table triggers - all the stored procedures involved, and recursively inspect these objects definition to find more references to more objects, until the list of every object referenced does not grow any more. This mechanism is known as "pre-locking" tables before execution. The motivation for locking all the tables (possibly) used at once is to prevent dead locks. One problem with this approach is that, if the execution path the code really takes during runtime does not use a given table, and if the table is missing, the server would not execute the statement. This in particular has a major impact on triggers, since a missing table referenced by an update/delete trigger would prevent an insert trigger to run. Another problem is that stored routines might define SQL exception handlers to deal with missing tables, but the server implementation would never give user code a chance to execute this logic, since the routine is never executed when a missing table cause the pre-locking code to fail. With this fix, the internal implementation of the pre-locking code has been relaxed of some constraints, so that failure to open a table does not necessarily prevent execution of a stored routine. In particular, the pre-locking mechanism is now behaving as follows: 1) the first step, to compute the transitive closure of all the tables possibly referenced by a statement, is unchanged. 2) the next step, which is to open all the tables involved, only attempts to open the tables added by the pre-locking code, but silently fails without reporting any error or invoking any exception handler is the table is not present. This is achieved by trapping internal errors with Prelock_error_handler 3) the locking step only locks tables that were successfully opened. 4) when executing sub statements, the list of tables used by each statements is evaluated as before. The tables needed by the sub statement are expected to be already opened and locked. Statement referencing tables that were not opened in step 2) will fail to find the table in the open list, and only at this point will execution of the user code fail. 5) when a runtime exception is raised at 4), the instruction continuation destination (the next instruction to execute in case of SQL continue handlers) is evaluated. This is achieved with sp_instr::exec_open_and_lock_tables() 6) if a user exception handler is present in the stored routine, that handler is invoked as usual, so that ER_NO_SUCH_TABLE exceptions can be trapped by stored routines. If no handler exists, then the runtime execution will fail as expected. With all these changes, a side effect is that view security is impacted, in two different ways. First, a view defined as "select stored_function()", where the stored function references a table that may not exist, is considered valid. The rationale is that, because the stored function might trap exceptions during execution and still return a valid result, there is no way to decide when the view is created if a missing table really cause the view to be invalid. Secondly, testing for existence of tables is now done later during execution. View security, which consist of trapping errors and return a generic ER_VIEW_INVALID (to prevent disclosing information) was only implemented at very specific phases covering *opening* tables, but not covering the runtime execution. Because of this existing limitation, errors that were previously trapped and converted into ER_VIEW_INVALID are not trapped, causing table names to be reported to the user. This change is exposing an existing problem, which is independent and will be resolved separately.
2007-03-06 03:42:07 +01:00
{
DBUG_ASSERT(m_internal_handler != NULL);
Internal_error_handler *popped_handler= m_internal_handler;
m_internal_handler= m_internal_handler->m_prev_internal_handler;
return popped_handler;
2000-07-31 21:29:14 +02:00
}
void THD::raise_error(uint sql_errno)
{
const char* msg= ER(sql_errno);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_ERROR,
msg);
}
void THD::raise_error_printf(uint sql_errno, ...)
{
va_list args;
char ebuff[MYSQL_ERRMSG_SIZE];
DBUG_ENTER("THD::raise_error_printf");
DBUG_PRINT("my", ("nr: %d errno: %d", sql_errno, errno));
const char* format= ER(sql_errno);
va_start(args, sql_errno);
my_vsnprintf(ebuff, sizeof(ebuff), format, args);
va_end(args);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_ERROR,
ebuff);
DBUG_VOID_RETURN;
}
void THD::raise_warning(uint sql_errno)
{
const char* msg= ER(sql_errno);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_WARN,
msg);
}
void THD::raise_warning_printf(uint sql_errno, ...)
{
va_list args;
char ebuff[MYSQL_ERRMSG_SIZE];
DBUG_ENTER("THD::raise_warning_printf");
DBUG_PRINT("enter", ("warning: %u", sql_errno));
const char* format= ER(sql_errno);
va_start(args, sql_errno);
my_vsnprintf(ebuff, sizeof(ebuff), format, args);
va_end(args);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_WARN,
ebuff);
DBUG_VOID_RETURN;
}
void THD::raise_note(uint sql_errno)
{
DBUG_ENTER("THD::raise_note");
DBUG_PRINT("enter", ("code: %d", sql_errno));
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
if (!(variables.option_bits & OPTION_SQL_NOTES))
DBUG_VOID_RETURN;
const char* msg= ER(sql_errno);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_NOTE,
msg);
DBUG_VOID_RETURN;
}
void THD::raise_note_printf(uint sql_errno, ...)
{
va_list args;
char ebuff[MYSQL_ERRMSG_SIZE];
DBUG_ENTER("THD::raise_note_printf");
DBUG_PRINT("enter",("code: %u", sql_errno));
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
if (!(variables.option_bits & OPTION_SQL_NOTES))
DBUG_VOID_RETURN;
const char* format= ER(sql_errno);
va_start(args, sql_errno);
my_vsnprintf(ebuff, sizeof(ebuff), format, args);
va_end(args);
(void) raise_condition(sql_errno,
NULL,
MYSQL_ERROR::WARN_LEVEL_NOTE,
ebuff);
DBUG_VOID_RETURN;
}
MYSQL_ERROR* THD::raise_condition(uint sql_errno,
const char* sqlstate,
MYSQL_ERROR::enum_warning_level level,
const char* msg)
{
MYSQL_ERROR *cond= NULL;
DBUG_ENTER("THD::raise_condition");
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
if (!(variables.option_bits & OPTION_SQL_NOTES) &&
(level == MYSQL_ERROR::WARN_LEVEL_NOTE))
DBUG_RETURN(NULL);
warning_info->opt_clear_warning_info(query_id);
/*
TODO: replace by DBUG_ASSERT(sql_errno != 0) once all bugs similar to
Bug#36768 are fixed: a SQL condition must have a real (!=0) error number
so that it can be caught by handlers.
*/
if (sql_errno == 0)
sql_errno= ER_UNKNOWN_ERROR;
if (msg == NULL)
msg= ER(sql_errno);
if (sqlstate == NULL)
sqlstate= mysql_errno_to_sqlstate(sql_errno);
if ((level == MYSQL_ERROR::WARN_LEVEL_WARN) &&
really_abort_on_warning())
{
/*
FIXME:
push_warning and strict SQL_MODE case.
*/
level= MYSQL_ERROR::WARN_LEVEL_ERROR;
killed= THD::KILL_BAD_DATA;
}
switch (level)
{
case MYSQL_ERROR::WARN_LEVEL_NOTE:
case MYSQL_ERROR::WARN_LEVEL_WARN:
got_warning= 1;
break;
case MYSQL_ERROR::WARN_LEVEL_ERROR:
break;
default:
DBUG_ASSERT(FALSE);
}
if (handle_condition(sql_errno, sqlstate, level, msg, &cond))
DBUG_RETURN(cond);
if (level == MYSQL_ERROR::WARN_LEVEL_ERROR)
{
is_slave_error= 1; // needed to catch query errors during replication
/*
thd->lex->current_select == 0 if lex structure is not inited
(not query command (COM_QUERY))
*/
if (lex->current_select &&
lex->current_select->no_error && !is_fatal_error)
{
DBUG_PRINT("error",
("Error converted to warning: current_select: no_error %d "
"fatal_error: %d",
(lex->current_select ?
lex->current_select->no_error : 0),
(int) is_fatal_error));
}
else
{
if (! stmt_da->is_error())
{
set_row_count_func(-1);
stmt_da->set_error_status(this, sql_errno, msg, sqlstate);
}
}
}
/*
If a continue handler is found, the error message will be cleared
by the stored procedures code.
*/
if (!is_fatal_error && spcont &&
spcont->handle_condition(this, sql_errno, sqlstate, level, msg, &cond))
{
/*
Do not push any warnings, a handled error must be completely
silenced.
*/
DBUG_RETURN(cond);
}
/* Un-handled conditions */
cond= raise_condition_no_handler(sql_errno, sqlstate, level, msg);
DBUG_RETURN(cond);
}
MYSQL_ERROR*
THD::raise_condition_no_handler(uint sql_errno,
const char* sqlstate,
MYSQL_ERROR::enum_warning_level level,
const char* msg)
{
MYSQL_ERROR *cond= NULL;
DBUG_ENTER("THD::raise_condition_no_handler");
query_cache_abort(&query_cache_tls);
/* FIXME: broken special case */
if (no_warnings_for_error && (level == MYSQL_ERROR::WARN_LEVEL_ERROR))
DBUG_RETURN(NULL);
cond= warning_info->push_warning(this, sql_errno, sqlstate, level, msg);
DBUG_RETURN(cond);
}
extern "C"
void *thd_alloc(MYSQL_THD thd, unsigned int size)
{
return thd->alloc(size);
}
extern "C"
void *thd_calloc(MYSQL_THD thd, unsigned int size)
{
return thd->calloc(size);
}
extern "C"
char *thd_strdup(MYSQL_THD thd, const char *str)
{
return thd->strdup(str);
}
extern "C"
char *thd_strmake(MYSQL_THD thd, const char *str, unsigned int size)
{
return thd->strmake(str, size);
}
extern "C"
LEX_STRING *thd_make_lex_string(THD *thd, LEX_STRING *lex_str,
const char *str, unsigned int size,
int allocate_lex_string)
{
return thd->make_lex_string(lex_str, str, size,
(bool) allocate_lex_string);
}
extern "C"
void *thd_memdup(MYSQL_THD thd, const void* str, unsigned int size)
{
return thd->memdup(str, size);
}
extern "C"
void thd_get_xid(const MYSQL_THD thd, MYSQL_XID *xid)
{
*xid = *(MYSQL_XID *) &thd->transaction.xid_state.xid;
}
#ifdef _WIN32
extern "C" THD *_current_thd_noinline(void)
{
return my_pthread_getspecific_ptr(THD*,THR_THD);
}
#endif
/*
Init common variables that has to be reset on start and on change_user
*/
void THD::init(void)
{
mysql_mutex_lock(&LOCK_global_system_variables);
plugin_thdvar_init(this);
/*
variables= global_system_variables above has reset
variables.pseudo_thread_id to 0. We need to correct it here to
avoid temporary tables replication failure.
*/
variables.pseudo_thread_id= thread_id;
mysql_mutex_unlock(&LOCK_global_system_variables);
server_status= SERVER_STATUS_AUTOCOMMIT;
if (variables.sql_mode & MODE_NO_BACKSLASH_ESCAPES)
server_status|= SERVER_STATUS_NO_BACKSLASH_ESCAPES;
transaction.all.modified_non_trans_table=
transaction.stmt.modified_non_trans_table= FALSE;
open_options=ha_open_options;
update_lock_default= (variables.low_priority_updates ?
TL_WRITE_LOW_PRIORITY :
TL_WRITE);
session_tx_isolation= (enum_tx_isolation) variables.tx_isolation;
update_charset();
reset_current_stmt_binlog_format_row();
bzero((char *) &status_var, sizeof(status_var));
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
sql_log_bin_toplevel= variables.option_bits & OPTION_BIN_LOG;
#if defined(ENABLED_DEBUG_SYNC)
/* Initialize the Debug Sync Facility. See debug_sync.cc. */
debug_sync_init_thread(this);
#endif /* defined(ENABLED_DEBUG_SYNC) */
}
/*
Init THD for query processing.
This has to be called once before we call mysql_parse.
See also comments in sql_class.h.
*/
void THD::init_for_queries()
{
set_time();
ha_enable_transaction(this,TRUE);
reset_root_defaults(mem_root, variables.query_alloc_block_size,
variables.query_prealloc_size);
reset_root_defaults(&transaction.mem_root,
variables.trans_alloc_block_size,
variables.trans_prealloc_size);
transaction.xid_state.xid.null();
transaction.xid_state.in_thd=1;
}
/*
Do what's needed when one invokes change user
SYNOPSIS
change_user()
IMPLEMENTATION
Reset all resources that are connection specific
*/
void THD::change_user(void)
{
mysql_mutex_lock(&LOCK_status);
add_to_status(&global_status_var, &status_var);
mysql_mutex_unlock(&LOCK_status);
cleanup();
killed= NOT_KILLED;
2002-11-21 21:25:53 +01:00
cleanup_done= 0;
init();
stmt_map.reset();
my_hash_init(&user_vars, system_charset_info, USER_VARS_HASH_SIZE, 0, 0,
(my_hash_get_key) get_var_key,
(my_hash_free_key) free_user_var, 0);
sp_cache_clear(&sp_proc_cache);
sp_cache_clear(&sp_func_cache);
}
/* Do operations that may take a long time */
void THD::cleanup(void)
2000-07-31 21:29:14 +02:00
{
DBUG_ENTER("THD::cleanup");
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
DBUG_ASSERT(cleanup_done == 0);
2007-04-05 21:34:33 +02:00
killed= KILL_CONNECTION;
2005-04-06 18:43:35 +02:00
#ifdef ENABLE_WHEN_BINLOG_WILL_BE_ABLE_TO_PREPARE
if (transaction.xid_state.xa_state == XA_PREPARED)
{
#error xid_state in the cache should be replaced by the allocated value
}
2005-04-06 18:43:35 +02:00
#endif
{
transaction.xid_state.xa_state= XA_NOTR;
trans_rollback(this);
xid_cache_delete(&transaction.xid_state);
}
Backport of revno ## 2617.31.1, 2617.31.3, 2617.31.4, 2617.31.5, 2617.31.12, 2617.31.15, 2617.31.15, 2617.31.16, 2617.43.1 - initial changeset that introduced the fix for Bug#989 and follow up fixes for all test suite failures introduced in the initial changeset. ------------------------------------------------------------ revno: 2617.31.1 committer: Davi Arnaut <Davi.Arnaut@Sun.COM> branch nick: 4284-6.0 timestamp: Fri 2009-03-06 19:17:00 -0300 message: Bug#989: If DROP TABLE while there's an active transaction, wrong binlog order WL#4284: Transactional DDL locking Currently the MySQL server does not keep metadata locks on schema objects for the duration of a transaction, thus failing to guarantee the integrity of the schema objects being used during the transaction and to protect then from concurrent DDL operations. This also poses a problem for replication as a DDL operation might be replicated even thought there are active transactions using the object being modified. The solution is to defer the release of metadata locks until a active transaction is either committed or rolled back. This prevents other statements from modifying the table for the entire duration of the transaction. This provides commitment ordering for guaranteeing serializability across multiple transactions. - Incompatible change: If MySQL's metadata locking system encounters a lock conflict, the usual schema is to use the try and back-off technique to avoid deadlocks -- this schema consists in releasing all locks and trying to acquire them all in one go. But in a transactional context this algorithm can't be utilized as its not possible to release locks acquired during the course of the transaction without breaking the transaction commitments. To avoid deadlocks in this case, the ER_LOCK_DEADLOCK will be returned if a lock conflict is encountered during a transaction. Let's consider an example: A transaction has two statements that modify table t1, then table t2, and then commits. The first statement of the transaction will acquire a shared metadata lock on table t1, and it will be kept utill COMMIT to ensure serializability. At the moment when the second statement attempts to acquire a shared metadata lock on t2, a concurrent ALTER or DROP statement might have locked t2 exclusively. The prescription of the current locking protocol is that the acquirer of the shared lock backs off -- gives up all his current locks and retries. This implies that the entire multi-statement transaction has to be rolled back. - Incompatible change: FLUSH commands such as FLUSH PRIVILEGES and FLUSH TABLES WITH READ LOCK won't cause locked tables to be implicitly unlocked anymore.
2009-12-05 00:02:48 +01:00
locked_tables_list.unlock_locked_tables(this);
A prerequisite patch for the fix for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". Introduce a notion of a sentinel to MDL_context. A sentinel is a ticket that separates all tickets in the context into two groups: before and after it. Currently we can have (and need) only one designated sentinel -- it separates all locks taken by LOCK TABLE or HANDLER statement, which must survive COMMIT and ROLLBACK and all other locks, which must be released at COMMIT or ROLLBACK. The tricky part is maintaining the sentinel up to date when someone release its corresponding ticket. This can happen, e.g. if someone issues DROP TABLE under LOCK TABLES (generally, see all calls to release_all_locks_for_name()). MDL_context::release_ticket() is modified to take care of it. ****** A fix and a test case for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". An attempt to mix HANDLER SQL statements, which are transaction- agnostic, an open multi-statement transaction, and DDL against the involved tables (in a concurrent connection) could lead to a deadlock. The deadlock would occur when HANDLER OPEN or HANDLER READ would have to wait on a conflicting metadata lock. If the connection that issued HANDLER statement also had other metadata locks (say, acquired in scope of a transaction), a classical deadlock situation of mutual wait could occur. Incompatible change: entering LOCK TABLES mode automatically closes all open HANDLERs in the current connection. Incompatible change: previously an attempt to wait on a lock in a connection that has an open HANDLER statement could wait indefinitely/deadlock. After this patch, an error ER_LOCK_DEADLOCK is produced. The idea of the fix is to merge thd->handler_mdl_context with the main mdl_context of the connection, used for transactional locks. This makes deadlock detection possible, since all waits with locks are "visible" and available to analysis in a single MDL context of the connection. Since HANDLER locks and transactional locks have a different life cycle -- HANDLERs are explicitly open and closed, and so are HANDLER locks, explicitly acquired and released, whereas transactional locks "accumulate" till the end of a transaction and are released only with COMMIT, ROLLBACK and ROLLBACK TO SAVEPOINT, a concept of "sentinel" was introduced to MDL_context. All locks, HANDLER and others, reside in the same linked list. However, a selected element of the list separates locks with different life cycle. HANDLER locks always reside at the end of the list, after the sentinel. Transactional locks are prepended to the beginning of the list, before the sentinel. Thus, ROLLBACK, COMMIT or ROLLBACK TO SAVEPOINT, only release those locks that reside before the sentinel. HANDLER locks must be released explicitly as part of HANDLER CLOSE statement, or an implicit close. The same approach with sentinel is also employed for LOCK TABLES locks. Since HANDLER and LOCK TABLES statement has never worked together, the implementation is made simple and only maintains one sentinel, which is used either for HANDLER locks, or for LOCK TABLES locks.
2009-12-22 17:09:15 +01:00
mysql_ha_cleanup(this);
Implement new type-of-operation-aware metadata locks. Add a wait-for graph based deadlock detector to the MDL subsystem. Fixes bug #46272 "MySQL 5.4.4, new MDL: unnecessary deadlock" and bug #37346 "innodb does not detect deadlock between update and alter table". The first bug manifested itself as an unwarranted abort of a transaction with ER_LOCK_DEADLOCK error by a concurrent ALTER statement, when this transaction tried to repeat use of a table, which it has already used in a similar fashion before ALTER started. The second bug showed up as a deadlock between table-level locks and InnoDB row locks, which was "detected" only after innodb_lock_wait_timeout timeout. A transaction would start using the table and modify a few rows. Then ALTER TABLE would come in, and start copying rows into a temporary table. Eventually it would stumble on the modified records and get blocked on a row lock. The first transaction would try to do more updates, and get blocked on thr_lock.c lock. This situation of circular wait would only get resolved by a timeout. Both these bugs stemmed from inadequate solutions to the problem of deadlocks occurring between different locking subsystems. In the first case we tried to avoid deadlocks between metadata locking and table-level locking subsystems, when upgrading shared metadata lock to exclusive one. Transactions holding the shared lock on the table and waiting for some table-level lock used to be aborted too aggressively. We also allowed ALTER TABLE to start in presence of transactions that modify the subject table. ALTER TABLE acquires TL_WRITE_ALLOW_READ lock at start, and that block all writes against the table (naturally, we don't want any writes to be lost when switching the old and the new table). TL_WRITE_ALLOW_READ lock, in turn, would block the started transaction on thr_lock.c lock, should they do more updates. This, again, lead to the need to abort such transactions. The second bug occurred simply because we didn't have any mechanism to detect deadlocks between the table-level locks in thr_lock.c and row-level locks in InnoDB, other than innodb_lock_wait_timeout. This patch solves both these problems by moving lock conflicts which are causing these deadlocks into the metadata locking subsystem, thus making it possible to avoid or detect such deadlocks inside MDL. To do this we introduce new type-of-operation-aware metadata locks, which allow MDL subsystem to know not only the fact that transaction has used or is going to use some object but also what kind of operation it has carried out or going to carry out on the object. This, along with the addition of a special kind of upgradable metadata lock, allows ALTER TABLE to wait until all transactions which has updated the table to go away. This solves the second issue. Another special type of upgradable metadata lock is acquired by LOCK TABLE WRITE. This second lock type allows to solve the first issue, since abortion of table-level locks in event of DDL under LOCK TABLES becomes also unnecessary. Below follows the list of incompatible changes introduced by this patch: - From now on, ALTER TABLE and CREATE/DROP TRIGGER SQL (i.e. those statements that acquire TL_WRITE_ALLOW_READ lock) wait for all transactions which has *updated* the table to complete. - From now on, LOCK TABLES ... WRITE, REPAIR/OPTIMIZE TABLE (i.e. all statements which acquire TL_WRITE table-level lock) wait for all transaction which *updated or read* from the table to complete. As a consequence, innodb_table_locks=0 option no longer applies to LOCK TABLES ... WRITE. - DROP DATABASE, DROP TABLE, RENAME TABLE no longer abort statements or transactions which use tables being dropped or renamed, and instead wait for these transactions to complete. - Since LOCK TABLES WRITE now takes a special metadata lock, not compatible with with reads or writes against the subject table and transaction-wide, thr_lock.c deadlock avoidance algorithm that used to ensure absence of deadlocks between LOCK TABLES WRITE and other statements is no longer sufficient, even for MyISAM. The wait-for graph based deadlock detector of MDL subsystem may sometimes be necessary and is involved. This may lead to ER_LOCK_DEADLOCK error produced for multi-statement transactions even if these only use MyISAM: session 1: session 2: begin; update t1 ... lock table t2 write, t1 write; -- gets a lock on t2, blocks on t1 update t2 ... (ER_LOCK_DEADLOCK) - Finally, support of LOW_PRIORITY option for LOCK TABLES ... WRITE was abandoned. LOCK TABLE ... LOW_PRIORITY WRITE from now on has the same priority as the usual LOCK TABLE ... WRITE. SELECT HIGH PRIORITY no longer trumps LOCK TABLE ... WRITE in the wait queue. - We do not take upgradable metadata locks on implicitly locked tables. So if one has, say, a view v1 that uses table t1, and issues: LOCK TABLE v1 WRITE; FLUSH TABLE t1; -- (or just 'FLUSH TABLES'), an error is produced. In order to be able to perform DDL on a table under LOCK TABLES, the table must be locked explicitly in the LOCK TABLES list.
2010-02-01 12:43:06 +01:00
DBUG_ASSERT(open_tables == NULL);
Backport of revno ## 2617.31.1, 2617.31.3, 2617.31.4, 2617.31.5, 2617.31.12, 2617.31.15, 2617.31.15, 2617.31.16, 2617.43.1 - initial changeset that introduced the fix for Bug#989 and follow up fixes for all test suite failures introduced in the initial changeset. ------------------------------------------------------------ revno: 2617.31.1 committer: Davi Arnaut <Davi.Arnaut@Sun.COM> branch nick: 4284-6.0 timestamp: Fri 2009-03-06 19:17:00 -0300 message: Bug#989: If DROP TABLE while there's an active transaction, wrong binlog order WL#4284: Transactional DDL locking Currently the MySQL server does not keep metadata locks on schema objects for the duration of a transaction, thus failing to guarantee the integrity of the schema objects being used during the transaction and to protect then from concurrent DDL operations. This also poses a problem for replication as a DDL operation might be replicated even thought there are active transactions using the object being modified. The solution is to defer the release of metadata locks until a active transaction is either committed or rolled back. This prevents other statements from modifying the table for the entire duration of the transaction. This provides commitment ordering for guaranteeing serializability across multiple transactions. - Incompatible change: If MySQL's metadata locking system encounters a lock conflict, the usual schema is to use the try and back-off technique to avoid deadlocks -- this schema consists in releasing all locks and trying to acquire them all in one go. But in a transactional context this algorithm can't be utilized as its not possible to release locks acquired during the course of the transaction without breaking the transaction commitments. To avoid deadlocks in this case, the ER_LOCK_DEADLOCK will be returned if a lock conflict is encountered during a transaction. Let's consider an example: A transaction has two statements that modify table t1, then table t2, and then commits. The first statement of the transaction will acquire a shared metadata lock on table t1, and it will be kept utill COMMIT to ensure serializability. At the moment when the second statement attempts to acquire a shared metadata lock on t2, a concurrent ALTER or DROP statement might have locked t2 exclusively. The prescription of the current locking protocol is that the acquirer of the shared lock backs off -- gives up all his current locks and retries. This implies that the entire multi-statement transaction has to be rolled back. - Incompatible change: FLUSH commands such as FLUSH PRIVILEGES and FLUSH TABLES WITH READ LOCK won't cause locked tables to be implicitly unlocked anymore.
2009-12-05 00:02:48 +01:00
/*
If the thread was in the middle of an ongoing transaction (rolled
back a few lines above) or under LOCK TABLES (unlocked the tables
and left the mode a few lines above), there will be outstanding
metadata locks. Release them.
*/
A prerequisite patch for the fix for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". Introduce a notion of a sentinel to MDL_context. A sentinel is a ticket that separates all tickets in the context into two groups: before and after it. Currently we can have (and need) only one designated sentinel -- it separates all locks taken by LOCK TABLE or HANDLER statement, which must survive COMMIT and ROLLBACK and all other locks, which must be released at COMMIT or ROLLBACK. The tricky part is maintaining the sentinel up to date when someone release its corresponding ticket. This can happen, e.g. if someone issues DROP TABLE under LOCK TABLES (generally, see all calls to release_all_locks_for_name()). MDL_context::release_ticket() is modified to take care of it. ****** A fix and a test case for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". An attempt to mix HANDLER SQL statements, which are transaction- agnostic, an open multi-statement transaction, and DDL against the involved tables (in a concurrent connection) could lead to a deadlock. The deadlock would occur when HANDLER OPEN or HANDLER READ would have to wait on a conflicting metadata lock. If the connection that issued HANDLER statement also had other metadata locks (say, acquired in scope of a transaction), a classical deadlock situation of mutual wait could occur. Incompatible change: entering LOCK TABLES mode automatically closes all open HANDLERs in the current connection. Incompatible change: previously an attempt to wait on a lock in a connection that has an open HANDLER statement could wait indefinitely/deadlock. After this patch, an error ER_LOCK_DEADLOCK is produced. The idea of the fix is to merge thd->handler_mdl_context with the main mdl_context of the connection, used for transactional locks. This makes deadlock detection possible, since all waits with locks are "visible" and available to analysis in a single MDL context of the connection. Since HANDLER locks and transactional locks have a different life cycle -- HANDLERs are explicitly open and closed, and so are HANDLER locks, explicitly acquired and released, whereas transactional locks "accumulate" till the end of a transaction and are released only with COMMIT, ROLLBACK and ROLLBACK TO SAVEPOINT, a concept of "sentinel" was introduced to MDL_context. All locks, HANDLER and others, reside in the same linked list. However, a selected element of the list separates locks with different life cycle. HANDLER locks always reside at the end of the list, after the sentinel. Transactional locks are prepended to the beginning of the list, before the sentinel. Thus, ROLLBACK, COMMIT or ROLLBACK TO SAVEPOINT, only release those locks that reside before the sentinel. HANDLER locks must be released explicitly as part of HANDLER CLOSE statement, or an implicit close. The same approach with sentinel is also employed for LOCK TABLES locks. Since HANDLER and LOCK TABLES statement has never worked together, the implementation is made simple and only maintains one sentinel, which is used either for HANDLER locks, or for LOCK TABLES locks.
2009-12-22 17:09:15 +01:00
mdl_context.release_transactional_locks();
Backport of revno ## 2617.31.1, 2617.31.3, 2617.31.4, 2617.31.5, 2617.31.12, 2617.31.15, 2617.31.15, 2617.31.16, 2617.43.1 - initial changeset that introduced the fix for Bug#989 and follow up fixes for all test suite failures introduced in the initial changeset. ------------------------------------------------------------ revno: 2617.31.1 committer: Davi Arnaut <Davi.Arnaut@Sun.COM> branch nick: 4284-6.0 timestamp: Fri 2009-03-06 19:17:00 -0300 message: Bug#989: If DROP TABLE while there's an active transaction, wrong binlog order WL#4284: Transactional DDL locking Currently the MySQL server does not keep metadata locks on schema objects for the duration of a transaction, thus failing to guarantee the integrity of the schema objects being used during the transaction and to protect then from concurrent DDL operations. This also poses a problem for replication as a DDL operation might be replicated even thought there are active transactions using the object being modified. The solution is to defer the release of metadata locks until a active transaction is either committed or rolled back. This prevents other statements from modifying the table for the entire duration of the transaction. This provides commitment ordering for guaranteeing serializability across multiple transactions. - Incompatible change: If MySQL's metadata locking system encounters a lock conflict, the usual schema is to use the try and back-off technique to avoid deadlocks -- this schema consists in releasing all locks and trying to acquire them all in one go. But in a transactional context this algorithm can't be utilized as its not possible to release locks acquired during the course of the transaction without breaking the transaction commitments. To avoid deadlocks in this case, the ER_LOCK_DEADLOCK will be returned if a lock conflict is encountered during a transaction. Let's consider an example: A transaction has two statements that modify table t1, then table t2, and then commits. The first statement of the transaction will acquire a shared metadata lock on table t1, and it will be kept utill COMMIT to ensure serializability. At the moment when the second statement attempts to acquire a shared metadata lock on t2, a concurrent ALTER or DROP statement might have locked t2 exclusively. The prescription of the current locking protocol is that the acquirer of the shared lock backs off -- gives up all his current locks and retries. This implies that the entire multi-statement transaction has to be rolled back. - Incompatible change: FLUSH commands such as FLUSH PRIVILEGES and FLUSH TABLES WITH READ LOCK won't cause locked tables to be implicitly unlocked anymore.
2009-12-05 00:02:48 +01:00
Implement new type-of-operation-aware metadata locks. Add a wait-for graph based deadlock detector to the MDL subsystem. Fixes bug #46272 "MySQL 5.4.4, new MDL: unnecessary deadlock" and bug #37346 "innodb does not detect deadlock between update and alter table". The first bug manifested itself as an unwarranted abort of a transaction with ER_LOCK_DEADLOCK error by a concurrent ALTER statement, when this transaction tried to repeat use of a table, which it has already used in a similar fashion before ALTER started. The second bug showed up as a deadlock between table-level locks and InnoDB row locks, which was "detected" only after innodb_lock_wait_timeout timeout. A transaction would start using the table and modify a few rows. Then ALTER TABLE would come in, and start copying rows into a temporary table. Eventually it would stumble on the modified records and get blocked on a row lock. The first transaction would try to do more updates, and get blocked on thr_lock.c lock. This situation of circular wait would only get resolved by a timeout. Both these bugs stemmed from inadequate solutions to the problem of deadlocks occurring between different locking subsystems. In the first case we tried to avoid deadlocks between metadata locking and table-level locking subsystems, when upgrading shared metadata lock to exclusive one. Transactions holding the shared lock on the table and waiting for some table-level lock used to be aborted too aggressively. We also allowed ALTER TABLE to start in presence of transactions that modify the subject table. ALTER TABLE acquires TL_WRITE_ALLOW_READ lock at start, and that block all writes against the table (naturally, we don't want any writes to be lost when switching the old and the new table). TL_WRITE_ALLOW_READ lock, in turn, would block the started transaction on thr_lock.c lock, should they do more updates. This, again, lead to the need to abort such transactions. The second bug occurred simply because we didn't have any mechanism to detect deadlocks between the table-level locks in thr_lock.c and row-level locks in InnoDB, other than innodb_lock_wait_timeout. This patch solves both these problems by moving lock conflicts which are causing these deadlocks into the metadata locking subsystem, thus making it possible to avoid or detect such deadlocks inside MDL. To do this we introduce new type-of-operation-aware metadata locks, which allow MDL subsystem to know not only the fact that transaction has used or is going to use some object but also what kind of operation it has carried out or going to carry out on the object. This, along with the addition of a special kind of upgradable metadata lock, allows ALTER TABLE to wait until all transactions which has updated the table to go away. This solves the second issue. Another special type of upgradable metadata lock is acquired by LOCK TABLE WRITE. This second lock type allows to solve the first issue, since abortion of table-level locks in event of DDL under LOCK TABLES becomes also unnecessary. Below follows the list of incompatible changes introduced by this patch: - From now on, ALTER TABLE and CREATE/DROP TRIGGER SQL (i.e. those statements that acquire TL_WRITE_ALLOW_READ lock) wait for all transactions which has *updated* the table to complete. - From now on, LOCK TABLES ... WRITE, REPAIR/OPTIMIZE TABLE (i.e. all statements which acquire TL_WRITE table-level lock) wait for all transaction which *updated or read* from the table to complete. As a consequence, innodb_table_locks=0 option no longer applies to LOCK TABLES ... WRITE. - DROP DATABASE, DROP TABLE, RENAME TABLE no longer abort statements or transactions which use tables being dropped or renamed, and instead wait for these transactions to complete. - Since LOCK TABLES WRITE now takes a special metadata lock, not compatible with with reads or writes against the subject table and transaction-wide, thr_lock.c deadlock avoidance algorithm that used to ensure absence of deadlocks between LOCK TABLES WRITE and other statements is no longer sufficient, even for MyISAM. The wait-for graph based deadlock detector of MDL subsystem may sometimes be necessary and is involved. This may lead to ER_LOCK_DEADLOCK error produced for multi-statement transactions even if these only use MyISAM: session 1: session 2: begin; update t1 ... lock table t2 write, t1 write; -- gets a lock on t2, blocks on t1 update t2 ... (ER_LOCK_DEADLOCK) - Finally, support of LOW_PRIORITY option for LOCK TABLES ... WRITE was abandoned. LOCK TABLE ... LOW_PRIORITY WRITE from now on has the same priority as the usual LOCK TABLE ... WRITE. SELECT HIGH PRIORITY no longer trumps LOCK TABLE ... WRITE in the wait queue. - We do not take upgradable metadata locks on implicitly locked tables. So if one has, say, a view v1 that uses table t1, and issues: LOCK TABLE v1 WRITE; FLUSH TABLE t1; -- (or just 'FLUSH TABLES'), an error is produced. In order to be able to perform DDL on a table under LOCK TABLES, the table must be locked explicitly in the LOCK TABLES list.
2010-02-01 12:43:06 +01:00
/* Release the global read lock, if acquired. */
if (global_read_lock.is_acquired())
global_read_lock.unlock_global_read_lock(this);
/* All metadata locks must have been released by now. */
DBUG_ASSERT(!mdl_context.has_locks());
#if defined(ENABLED_DEBUG_SYNC)
/* End the Debug Sync Facility. See debug_sync.cc. */
debug_sync_end_thread(this);
#endif /* defined(ENABLED_DEBUG_SYNC) */
delete_dynamic(&user_var_events);
my_hash_free(&user_vars);
2000-07-31 21:29:14 +02:00
close_temporary_tables(this);
sp_cache_clear(&sp_proc_cache);
sp_cache_clear(&sp_func_cache);
if (ull)
{
mysql_mutex_lock(&LOCK_user_locks);
item_user_lock_release(ull);
mysql_mutex_unlock(&LOCK_user_locks);
ull= NULL;
}
cleanup_done=1;
DBUG_VOID_RETURN;
}
THD::~THD()
{
THD_CHECK_SENTRY(this);
DBUG_ENTER("~THD()");
/* Ensure that no one is using THD */
mysql_mutex_lock(&LOCK_thd_data);
mysql_mutex_unlock(&LOCK_thd_data);
add_to_status(&global_status_var, &status_var);
/* Close connection */
2005-04-04 00:50:05 +02:00
#ifndef EMBEDDED_LIBRARY
if (net.vio)
{
vio_delete(net.vio);
2005-04-04 00:50:05 +02:00
net_end(&net);
}
#endif
stmt_map.reset(); /* close all prepared statements */
DBUG_ASSERT(lock_info.n_cursors == 0);
if (!cleanup_done)
cleanup();
2005-01-16 13:16:23 +01:00
mdl_context.destroy();
2005-04-04 00:50:05 +02:00
ha_close_connection(this);
mysql_audit_release(this);
plugin_thdvar_cleanup(this);
2000-07-31 21:29:14 +02:00
DBUG_PRINT("info", ("freeing security context"));
main_security_ctx.destroy();
safeFree(db);
free_root(&transaction.mem_root,MYF(0));
2000-07-31 21:29:14 +02:00
mysys_var=0; // Safety (shouldn't be needed)
mysql_mutex_destroy(&LOCK_thd_data);
#ifndef DBUG_OFF
2003-12-04 20:08:26 +01:00
dbug_sentry= THD_SENTRY_GONE;
#endif
#ifndef EMBEDDED_LIBRARY
if (rli_fake)
{
delete rli_fake;
rli_fake= NULL;
}
mysql_audit_free_thd(this);
#endif
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
free_root(&main_mem_root, MYF(0));
2000-07-31 21:29:14 +02:00
DBUG_VOID_RETURN;
}
/*
2006-01-10 17:56:23 +01:00
Add all status variables to another status variable array
SYNOPSIS
add_to_status()
to_var add to this array
from_var from this array
NOTES
This function assumes that all variables are long/ulong.
If this assumption will change, then we have to explictely add
the other variables after the while loop
*/
void add_to_status(STATUS_VAR *to_var, STATUS_VAR *from_var)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
ulong *end= (ulong*) ((uchar*) to_var +
offsetof(STATUS_VAR, last_system_status_var) +
sizeof(ulong));
ulong *to= (ulong*) to_var, *from= (ulong*) from_var;
while (to != end)
*(to++)+= *(from++);
}
/*
Add the difference between two status variable arrays to another one.
SYNOPSIS
add_diff_to_status
to_var add to this array
from_var from this array
dec_var minus this array
NOTE
This function assumes that all variables are long/ulong.
*/
void add_diff_to_status(STATUS_VAR *to_var, STATUS_VAR *from_var,
STATUS_VAR *dec_var)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
ulong *end= (ulong*) ((uchar*) to_var + offsetof(STATUS_VAR,
last_system_status_var) +
sizeof(ulong));
ulong *to= (ulong*) to_var, *from= (ulong*) from_var, *dec= (ulong*) dec_var;
while (to != end)
*(to++)+= *(from++) - *(dec++);
}
void THD::awake(THD::killed_state state_to_set)
{
DBUG_ENTER("THD::awake");
DBUG_PRINT("enter", ("this: 0x%lx", (long) this));
THD_CHECK_SENTRY(this);
mysql_mutex_assert_owner(&LOCK_thd_data);
killed= state_to_set;
if (state_to_set != THD::KILL_QUERY)
{
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
thr_alarm_kill(thread_id);
if (!slave_thread)
thread_scheduler.post_kill_notification(this);
#ifdef SIGNAL_WITH_VIO_CLOSE
if (this != current_thd)
{
/*
In addition to a signal, let's close the socket of the thread that
is being killed. This is to make sure it does not block if the
signal is lost. This needs to be done only on platforms where
signals are not a reliable interruption mechanism.
If we're killing ourselves, we know that we're not blocked, so this
hack is not used.
*/
close_active_vio();
}
#endif
}
if (mysys_var)
{
mysql_mutex_lock(&mysys_var->mutex);
if (!system_thread) // Don't abort locks
mysys_var->abort=1;
/*
This broadcast could be up in the air if the victim thread
exits the cond in the time between read and broadcast, but that is
ok since all we want to do is to make the victim thread get out
of waiting on current_cond.
If we see a non-zero current_cond: it cannot be an old value (because
then exit_cond() should have run and it can't because we have mutex); so
it is the true value but maybe current_mutex is not yet non-zero (we're
in the middle of enter_cond() and there is a "memory order
inversion"). So we test the mutex too to not lock 0.
Note that there is a small chance we fail to kill. If victim has locked
current_mutex, but hasn't yet entered enter_cond() (which means that
current_cond and current_mutex are 0), then the victim will not get
a signal and it may wait "forever" on the cond (until
we issue a second KILL or the status it's waiting for happens).
It's true that we have set its thd->killed but it may not
see it immediately and so may have time to reach the cond_wait().
*/
if (mysys_var->current_cond && mysys_var->current_mutex)
{
mysql_mutex_lock(mysys_var->current_mutex);
mysql_cond_broadcast(mysys_var->current_cond);
mysql_mutex_unlock(mysys_var->current_mutex);
}
mysql_mutex_unlock(&mysys_var->mutex);
}
DBUG_VOID_RETURN;
}
/*
Remember the location of thread info, the structure needed for
sql_alloc() and the structure for the net buffer
*/
2000-07-31 21:29:14 +02:00
bool THD::store_globals()
{
/*
Assert that thread_stack is initialized: it's necessary to be able
to track stack overrun.
*/
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
DBUG_ASSERT(thread_stack);
if (my_pthread_setspecific_ptr(THR_THD, this) ||
2002-10-02 16:55:12 +02:00
my_pthread_setspecific_ptr(THR_MALLOC, &mem_root))
return 1;
mysys_var=my_thread_var;
/*
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
Let mysqld define the thread id (not mysys)
This allows us to move THD to different threads if needed.
*/
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
mysys_var->id= thread_id;
real_id= pthread_self(); // For debugging
/*
We have to call thr_lock_info_init() again here as THD may have been
created in another thread
*/
thr_lock_info_init(&lock_info);
return 0;
2000-07-31 21:29:14 +02:00
}
/*
Cleanup after query.
SYNOPSIS
THD::cleanup_after_query()
DESCRIPTION
2006-10-03 15:07:30 +02:00
This function is used to reset thread data to its default state.
NOTE
This function is not suitable for setting thread data to some
non-default values, as there is only one replication thread, so
different master threads may overwrite data of each other on
slave.
*/
2006-10-03 15:07:30 +02:00
void THD::cleanup_after_query()
{
/*
Reset rand_used so that detection of calls to rand() will save random
seeds if needed by the slave.
Do not reset rand_used if inside a stored function or trigger because
only the call to these operations is logged. Thus only the calling
statement needs to detect rand() calls made by its substatements. These
substatements must not set rand_used to 0 because it would remove the
detection of rand() by the calling statement.
*/
if (!in_sub_stmt) /* stored functions and triggers are a special case */
{
/* Forget those values, for next binlogger: */
stmt_depends_on_first_successful_insert_id_in_prev_stmt= 0;
auto_inc_intervals_in_cur_stmt_for_binlog.empty();
rand_used= 0;
}
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
if (first_successful_insert_id_in_cur_stmt > 0)
{
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
/* set what LAST_INSERT_ID() will return */
first_successful_insert_id_in_prev_stmt=
first_successful_insert_id_in_cur_stmt;
first_successful_insert_id_in_cur_stmt= 0;
2006-08-01 06:49:43 +02:00
substitute_null_with_insert_id= TRUE;
}
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
arg_of_last_insert_id_function= 0;
/* Free Items that were created during this execution */
free_items();
/* Reset where. */
where= THD::DEFAULT_WHERE;
/* reset table map for multi-table update */
table_map_for_update= 0;
}
2006-10-03 15:07:30 +02:00
/**
2007-10-16 21:37:31 +02:00
Create a LEX_STRING in this connection.
@param lex_str pointer to LEX_STRING object to be initialized
@param str initializer to be copied into lex_str
@param length length of str, in bytes
@param allocate_lex_string if TRUE, allocate new LEX_STRING object,
instead of using lex_str value
@return NULL on failure, or pointer to the LEX_STRING object
*/
LEX_STRING *THD::make_lex_string(LEX_STRING *lex_str,
const char* str, uint length,
bool allocate_lex_string)
{
if (allocate_lex_string)
if (!(lex_str= (LEX_STRING *)alloc(sizeof(LEX_STRING))))
return 0;
if (!(lex_str->str= strmake_root(mem_root, str, length)))
return 0;
lex_str->length= length;
return lex_str;
}
/*
Convert a string to another character set
SYNOPSIS
convert_string()
to Store new allocated string here
to_cs New character set for allocated string
from String to convert
from_length Length of string to convert
from_cs Original character set
NOTES
to will be 0-terminated to make it easy to pass to system funcs
RETURN
0 ok
1 End of memory.
In this case to->str will point to 0 and to->length will be 0.
*/
bool THD::convert_string(LEX_STRING *to, CHARSET_INFO *to_cs,
const char *from, uint from_length,
CHARSET_INFO *from_cs)
{
DBUG_ENTER("convert_string");
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t new_length= to_cs->mbmaxlen * from_length;
uint dummy_errors;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (!(to->str= (char*) alloc(new_length+1)))
{
to->length= 0; // Safety fix
DBUG_RETURN(1); // EOM
}
to->length= copy_and_convert((char*) to->str, new_length, to_cs,
from, from_length, from_cs, &dummy_errors);
to->str[to->length]=0; // Safety
DBUG_RETURN(0);
}
/*
Convert string from source character set to target character set inplace.
SYNOPSIS
THD::convert_string
DESCRIPTION
Convert string using convert_buffer - buffer for character set
conversion shared between all protocols.
RETURN
0 ok
!0 out of memory
*/
bool THD::convert_string(String *s, CHARSET_INFO *from_cs, CHARSET_INFO *to_cs)
{
uint dummy_errors;
if (convert_buffer.copy(s->ptr(), s->length(), from_cs, to_cs, &dummy_errors))
return TRUE;
/* If convert_buffer >> s copying is more efficient long term */
if (convert_buffer.alloced_length() >= convert_buffer.length() * 2 ||
!s->is_alloced())
{
return s->copy(convert_buffer);
}
s->swap(convert_buffer);
return FALSE;
}
/*
Update some cache variables when character set changes
*/
void THD::update_charset()
{
uint32 not_used;
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
charset_is_system_charset=
!String::needs_conversion(0,
variables.character_set_client,
system_charset_info,
&not_used);
charset_is_collation_connection=
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
!String::needs_conversion(0,
variables.character_set_client,
variables.collation_connection,
&not_used);
charset_is_character_set_filesystem=
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
!String::needs_conversion(0,
variables.character_set_client,
variables.character_set_filesystem,
&not_used);
}
/* routings to adding tables to list of changed in transaction tables */
inline static void list_include(CHANGED_TABLE_LIST** prev,
CHANGED_TABLE_LIST* curr,
CHANGED_TABLE_LIST* new_table)
{
if (new_table)
{
*prev = new_table;
(*prev)->next = curr;
}
}
/* add table to list of changed in transaction tables */
void THD::add_changed_table(TABLE *table)
{
DBUG_ENTER("THD::add_changed_table(table)");
DBUG_ASSERT(in_multi_stmt_transaction() && table->file->has_transactions());
add_changed_table(table->s->table_cache_key.str,
(long) table->s->table_cache_key.length);
2002-09-19 19:10:06 +02:00
DBUG_VOID_RETURN;
}
void THD::add_changed_table(const char *key, long key_length)
{
DBUG_ENTER("THD::add_changed_table(key)");
CHANGED_TABLE_LIST **prev_changed = &transaction.changed_tables;
CHANGED_TABLE_LIST *curr = transaction.changed_tables;
for (; curr; prev_changed = &(curr->next), curr = curr->next)
{
int cmp = (long)curr->key_length - (long)key_length;
if (cmp < 0)
{
list_include(prev_changed, curr, changed_table_dup(key, key_length));
DBUG_PRINT("info",
2007-03-22 19:32:07 +01:00
("key_length: %ld %u", key_length,
(*prev_changed)->key_length));
DBUG_VOID_RETURN;
}
else if (cmp == 0)
{
cmp = memcmp(curr->key, key, curr->key_length);
if (cmp < 0)
{
list_include(prev_changed, curr, changed_table_dup(key, key_length));
DBUG_PRINT("info",
("key_length: %ld %u", key_length,
(*prev_changed)->key_length));
DBUG_VOID_RETURN;
}
else if (cmp == 0)
{
DBUG_PRINT("info", ("already in list"));
DBUG_VOID_RETURN;
}
}
}
*prev_changed = changed_table_dup(key, key_length);
2007-03-22 19:32:07 +01:00
DBUG_PRINT("info", ("key_length: %ld %u", key_length,
(*prev_changed)->key_length));
DBUG_VOID_RETURN;
}
CHANGED_TABLE_LIST* THD::changed_table_dup(const char *key, long key_length)
{
CHANGED_TABLE_LIST* new_table =
(CHANGED_TABLE_LIST*) trans_alloc(ALIGN_SIZE(sizeof(CHANGED_TABLE_LIST))+
key_length + 1);
if (!new_table)
{
my_error(EE_OUTOFMEMORY, MYF(ME_BELL),
ALIGN_SIZE(sizeof(TABLE_LIST)) + key_length + 1);
killed= KILL_CONNECTION;
return 0;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
new_table->key= ((char*)new_table)+ ALIGN_SIZE(sizeof(CHANGED_TABLE_LIST));
new_table->next = 0;
new_table->key_length = key_length;
::memcpy(new_table->key, key, key_length);
return new_table;
}
int THD::send_explain_fields(select_result *result)
{
List<Item> field_list;
Item *item;
CHARSET_INFO *cs= system_charset_info;
field_list.push_back(new Item_return_int("id",3, MYSQL_TYPE_LONGLONG));
field_list.push_back(new Item_empty_string("select_type", 19, cs));
field_list.push_back(item= new Item_empty_string("table", NAME_CHAR_LEN, cs));
item->maybe_null= 1;
2005-12-22 10:29:00 +01:00
if (lex->describe & DESCRIBE_PARTITIONS)
{
/* Maximum length of string that make_used_partitions_str() can produce */
item= new Item_empty_string("partitions", MAX_PARTITIONS * (1 + FN_LEN),
cs);
field_list.push_back(item);
2005-12-22 10:29:00 +01:00
item->maybe_null= 1;
}
field_list.push_back(item= new Item_empty_string("type", 10, cs));
item->maybe_null= 1;
field_list.push_back(item=new Item_empty_string("possible_keys",
NAME_CHAR_LEN*MAX_KEY, cs));
item->maybe_null=1;
field_list.push_back(item=new Item_empty_string("key", NAME_CHAR_LEN, cs));
item->maybe_null=1;
2003-11-13 15:52:02 +01:00
field_list.push_back(item=new Item_empty_string("key_len",
NAME_CHAR_LEN*MAX_KEY));
item->maybe_null=1;
field_list.push_back(item=new Item_empty_string("ref",
NAME_CHAR_LEN*MAX_REF_PARTS,
cs));
item->maybe_null=1;
field_list.push_back(item= new Item_return_int("rows", 10,
MYSQL_TYPE_LONGLONG));
if (lex->describe & DESCRIBE_EXTENDED)
{
field_list.push_back(item= new Item_float("filtered", 0.1234, 2, 4));
item->maybe_null=1;
}
item->maybe_null= 1;
field_list.push_back(new Item_empty_string("Extra", 255, cs));
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.
2009-10-21 22:02:06 +02:00
return (result->send_result_set_metadata(field_list,
Protocol::SEND_NUM_ROWS | Protocol::SEND_EOF));
}
#ifdef SIGNAL_WITH_VIO_CLOSE
void THD::close_active_vio()
{
2003-03-12 00:40:06 +01:00
DBUG_ENTER("close_active_vio");
mysql_mutex_assert_owner(&LOCK_thd_data);
2003-01-28 18:03:05 +01:00
#ifndef EMBEDDED_LIBRARY
if (active_vio)
{
vio_close(active_vio);
active_vio = 0;
}
2003-01-28 18:03:05 +01:00
#endif
DBUG_VOID_RETURN;
}
#endif
struct Item_change_record: public ilink
{
Item **place;
Item *old_value;
/* Placement new was hidden by `new' in ilink (TODO: check): */
static void *operator new(size_t size, void *mem) { return mem; }
static void operator delete(void *ptr, size_t size) {}
static void operator delete(void *ptr, void *mem) { /* never called */ }
};
/*
Register an item tree tree transformation, performed by the query
optimizer. We need a pointer to runtime_memroot because it may be !=
thd->mem_root (due to possible set_n_backup_active_arena called for thd).
*/
void THD::nocheck_register_item_tree_change(Item **place, Item *old_value,
MEM_ROOT *runtime_memroot)
{
Item_change_record *change;
/*
Now we use one node per change, which adds some memory overhead,
but still is rather fast as we use alloc_root for allocations.
A list of item tree changes of an average query should be short.
*/
void *change_mem= alloc_root(runtime_memroot, sizeof(*change));
if (change_mem == 0)
{
/*
OOM, thd->fatal_error() is called by the error handler of the
memroot. Just return.
*/
return;
}
change= new (change_mem) Item_change_record;
change->place= place;
change->old_value= old_value;
A fix and test case for Bug#5987 "subselect in bool function crashes server (prepared statements)": the bug was that all boolean items always recovered its original arguments at statement cleanup stage. This collided with Item_subselect::select_transformer, which tries to permanently change the item tree to use a transformed subselect instead of original one. So we had this call sequence for prepare: mysql_stmt_prepare -> JOIN::prepare -> Item_subselect::fix_fields -> the item tree gets transformed -> Item_bool_rowready_func2::cleanup, item tree is recovered to original state, while it shouldn't have been; mysql_stmt_execute -> attempts to execute a broken tree -> crash. Now instead of bluntly recovering all arguments of bool functions in Item_bool_rowready_func2::cleanup, we recover only those which were changed, and do it in one place. There still would exist a possibility for a collision with subselect tranformation, if permanent and temporary changes were performed at the same stage. But fortunately subselect transformation is always done first, so it doesn't conflict with the optimization done by propogate_cond_constants. Now we have: mysql_stmt_prepare -> JOIN::prepare -> subselect transformation permanently changes the tree -> cleanup doesn't recover anything, because nothing was registered for recovery. mysql_stmt_execute -> JOIN::prepare (the tree is already transformed, so it doesn't change), JOIN::optimize -> propogate_cond_constants -> temporary changes the item tree with constants -> JOIN::execute -> cleanup -> the changes done by propogate_cond_constants are recovered, as they were registered for recovery.
2004-10-10 00:39:22 +02:00
change_list.append(change);
}
void THD::rollback_item_tree_changes()
{
I_List_iterator<Item_change_record> it(change_list);
Item_change_record *change;
DBUG_ENTER("rollback_item_tree_changes");
while ((change= it++))
*change->place= change->old_value;
/* We can forget about changes memory: it's allocated in runtime memroot */
change_list.empty();
DBUG_VOID_RETURN;
}
2000-07-31 21:29:14 +02:00
/*****************************************************************************
** Functions to provide a interface to select results
*****************************************************************************/
select_result::select_result()
{
thd=current_thd;
}
void select_result::send_error(uint errcode,const char *err)
{
my_message(errcode, err, MYF(0));
}
void select_result::cleanup()
{
/* do nothing */
}
bool select_result::check_simple_select() const
{
my_error(ER_SP_BAD_CURSOR_QUERY, MYF(0));
return TRUE;
}
static String default_line_term("\n",default_charset_info);
static String default_escaped("\\",default_charset_info);
static String default_field_term("\t",default_charset_info);
2009-10-12 08:22:53 +02:00
static String default_xml_row_term("<row>", default_charset_info);
2000-07-31 21:29:14 +02:00
2009-10-12 08:22:53 +02:00
sql_exchange::sql_exchange(char *name, bool flag,
enum enum_filetype filetype_arg)
2000-07-31 21:29:14 +02:00
:file_name(name), opt_enclosed(0), dumpfile(flag), skip_lines(0)
{
2009-10-12 08:22:53 +02:00
filetype= filetype_arg;
2000-07-31 21:29:14 +02:00
field_term= &default_field_term;
enclosed= line_start= &my_empty_string;
2009-10-12 08:22:53 +02:00
line_term= filetype == FILETYPE_CSV ?
&default_line_term : &default_xml_row_term;
2000-07-31 21:29:14 +02:00
escaped= &default_escaped;
cs= NULL;
2000-07-31 21:29:14 +02:00
}
bool sql_exchange::escaped_given(void)
{
return escaped != &default_escaped;
}
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.
2009-10-21 22:02:06 +02:00
bool select_send::send_result_set_metadata(List<Item> &list, uint flags)
2000-07-31 21:29:14 +02:00
{
bool res;
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.
2009-10-21 22:02:06 +02:00
if (!(res= thd->protocol->send_result_set_metadata(&list, flags)))
is_result_set_started= 1;
return res;
2000-07-31 21:29:14 +02:00
}
void select_send::abort()
{
DBUG_ENTER("select_send::abort");
if (is_result_set_started && thd->spcont)
{
/*
We're executing a stored procedure, have an open result
set and an SQL exception condition. In this situation we
must abort the current statement, silence the error and
start executing the continue/exit handler if one is found.
Before aborting the statement, let's end the open result set, as
otherwise the client will hang due to the violation of the
client/server protocol.
*/
thd->spcont->end_partial_result_set= TRUE;
}
DBUG_VOID_RETURN;
}
/**
Cleanup an instance of this class for re-use
at next execution of a prepared statement/
stored procedure statement.
*/
void select_send::cleanup()
{
is_result_set_started= FALSE;
}
2000-07-31 21:29:14 +02:00
/* Send data to client. Returns 0 if ok */
bool select_send::send_data(List<Item> &items)
{
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.
2009-10-21 22:02:06 +02:00
Protocol *protocol= thd->protocol;
DBUG_ENTER("select_send::send_data");
if (unit->offset_limit_cnt)
2000-07-31 21:29:14 +02:00
{ // using limit offset,count
unit->offset_limit_cnt--;
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.
2009-10-21 22:02:06 +02:00
DBUG_RETURN(FALSE);
2000-07-31 21:29:14 +02:00
}
2003-03-16 18:17:54 +01:00
/*
We may be passing the control from mysqld to the client: release the
InnoDB adaptive hash S-latch to avoid thread deadlocks if it was reserved
by thd
*/
ha_release_temporary_latches(thd);
protocol->prepare_for_resend();
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.
2009-10-21 22:02:06 +02:00
if (protocol->send_result_set_row(&items))
{
protocol->remove_last_row();
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.
2009-10-21 22:02:06 +02:00
DBUG_RETURN(TRUE);
}
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.
2009-10-21 22:02:06 +02:00
thd->sent_row_count++;
if (thd->vio_ok())
DBUG_RETURN(protocol->write());
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.
2009-10-21 22:02:06 +02:00
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
bool select_send::send_eof()
{
/*
We may be passing the control from mysqld to the client: release the
InnoDB adaptive hash S-latch to avoid thread deadlocks if it was reserved
by thd
*/
ha_release_temporary_latches(thd);
2000-07-31 21:29:14 +02:00
/* Unlock tables before sending packet to gain some speed */
if (thd->lock && ! thd->locked_tables_mode)
2000-07-31 21:29:14 +02:00
{
mysql_unlock_tables(thd, thd->lock);
thd->lock=0;
2000-07-31 21:29:14 +02:00
}
/*
Don't send EOF if we're in error condition (which implies we've already
sent or are sending an error)
*/
if (thd->is_error())
return TRUE;
::my_eof(thd);
is_result_set_started= 0;
return FALSE;
2000-07-31 21:29:14 +02:00
}
/************************************************************************
Handling writing to file
************************************************************************/
2000-07-31 21:29:14 +02:00
void select_to_file::send_error(uint errcode,const char *err)
{
my_message(errcode, err, MYF(0));
if (file > 0)
{
(void) end_io_cache(&cache);
mysql_file_close(file, MYF(0));
/* Delete file on error */
mysql_file_delete(key_select_to_file, path, MYF(0));
file= -1;
}
}
2000-07-31 21:29:14 +02:00
bool select_to_file::send_eof()
{
int error= test(end_io_cache(&cache));
if (mysql_file_close(file, MYF(MY_WME)))
error= 1;
if (!error)
{
::my_ok(thd,row_count);
}
file= -1;
return error;
}
void select_to_file::cleanup()
{
/* In case of error send_eof() may be not called: close the file here. */
if (file >= 0)
{
(void) end_io_cache(&cache);
mysql_file_close(file, MYF(0));
file= -1;
}
path[0]= '\0';
row_count= 0;
}
select_to_file::~select_to_file()
2000-07-31 21:29:14 +02:00
{
if (file >= 0)
{ // This only happens in case of error
(void) end_io_cache(&cache);
mysql_file_close(file, MYF(0));
2000-07-31 21:29:14 +02:00
file= -1;
}
}
/***************************************************************************
** Export of select to textfile
***************************************************************************/
select_export::~select_export()
{
thd->sent_row_count=row_count;
2000-07-31 21:29:14 +02:00
}
/*
Create file with IO cache
SYNOPSIS
create_file()
thd Thread handle
path File name
exchange Excange class
cache IO cache
RETURN
>= 0 File handle
-1 Error
*/
static File create_file(THD *thd, char *path, sql_exchange *exchange,
IO_CACHE *cache)
2000-07-31 21:29:14 +02:00
{
File file;
uint option= MY_UNPACK_FILENAME | MY_RELATIVE_PATH;
2000-07-31 21:29:14 +02:00
#ifdef DONT_ALLOW_FULL_LOAD_DATA_PATHS
option|= MY_REPLACE_DIR; // Force use of db directory
2000-07-31 21:29:14 +02:00
#endif
2004-09-13 18:49:41 +02:00
if (!dirname_length(exchange->file_name))
2004-09-07 08:55:34 +02:00
{
strxnmov(path, FN_REFLEN-1, mysql_real_data_home, thd->db ? thd->db : "",
NullS);
2004-09-07 08:55:34 +02:00
(void) fn_format(path, exchange->file_name, path, "", option);
}
else
(void) fn_format(path, exchange->file_name, mysql_real_data_home, "", option);
if (opt_secure_file_priv &&
strncmp(opt_secure_file_priv, path, strlen(opt_secure_file_priv)))
{
/* Write only allowed to dir or subdir specified by secure_file_priv */
my_error(ER_OPTION_PREVENTS_STATEMENT, MYF(0), "--secure-file-priv");
return -1;
}
if (!access(path, F_OK))
2000-07-31 21:29:14 +02:00
{
my_error(ER_FILE_EXISTS_ERROR, MYF(0), exchange->file_name);
return -1;
2000-07-31 21:29:14 +02:00
}
/* Create the file world readable */
if ((file= mysql_file_create(key_select_to_file,
path, 0666, O_WRONLY|O_EXCL, MYF(MY_WME))) < 0)
return file;
2000-07-31 21:29:14 +02:00
#ifdef HAVE_FCHMOD
(void) fchmod(file, 0666); // Because of umask()
2000-07-31 21:29:14 +02:00
#else
(void) chmod(path, 0666);
2000-07-31 21:29:14 +02:00
#endif
if (init_io_cache(cache, file, 0L, WRITE_CACHE, 0L, 1, MYF(MY_WME)))
2000-07-31 21:29:14 +02:00
{
mysql_file_close(file, MYF(0));
/* Delete file on error, it was just created */
mysql_file_delete(key_select_to_file, path, MYF(0));
return -1;
2000-07-31 21:29:14 +02:00
}
return file;
}
int
select_export::prepare(List<Item> &list, SELECT_LEX_UNIT *u)
{
bool blob_flag=0;
bool string_results= FALSE, non_string_results= FALSE;
unit= u;
if ((uint) strlen(exchange->file_name) + NAME_LEN >= FN_REFLEN)
strmake(path,exchange->file_name,FN_REFLEN-1);
write_cs= exchange->cs ? exchange->cs : &my_charset_bin;
if ((file= create_file(thd, path, exchange, &cache)) < 0)
return 1;
2000-07-31 21:29:14 +02:00
/* Check if there is any blobs in data */
{
List_iterator_fast<Item> li(list);
2000-07-31 21:29:14 +02:00
Item *item;
while ((item=li++))
{
if (item->max_length >= MAX_BLOB_WIDTH)
{
blob_flag=1;
break;
}
if (item->result_type() == STRING_RESULT)
string_results= TRUE;
else
non_string_results= TRUE;
2000-07-31 21:29:14 +02:00
}
}
if (exchange->escaped->numchars() > 1 || exchange->enclosed->numchars() > 1)
{
my_error(ER_WRONG_FIELD_TERMINATORS, MYF(0));
return TRUE;
}
if (exchange->escaped->length() > 1 || exchange->enclosed->length() > 1 ||
!my_isascii(exchange->escaped->ptr()[0]) ||
!my_isascii(exchange->enclosed->ptr()[0]) ||
!exchange->field_term->is_ascii() || !exchange->line_term->is_ascii() ||
!exchange->line_start->is_ascii())
{
/*
Current LOAD DATA INFILE recognizes field/line separators "as is" without
converting from client charset to data file charset. So, it is supposed,
that input file of LOAD DATA INFILE consists of data in one charset and
separators in other charset. For the compatibility with that [buggy]
behaviour SELECT INTO OUTFILE implementation has been saved "as is" too,
but the new warning message has been added:
Non-ASCII separator arguments are not fully supported
*/
push_warning(thd, MYSQL_ERROR::WARN_LEVEL_WARN,
WARN_NON_ASCII_SEPARATOR_NOT_IMPLEMENTED,
ER(WARN_NON_ASCII_SEPARATOR_NOT_IMPLEMENTED));
}
2000-07-31 21:29:14 +02:00
field_term_length=exchange->field_term->length();
field_term_char= field_term_length ?
(int) (uchar) (*exchange->field_term)[0] : INT_MAX;
2000-07-31 21:29:14 +02:00
if (!exchange->line_term->length())
exchange->line_term=exchange->field_term; // Use this if it exists
field_sep_char= (exchange->enclosed->length() ?
(int) (uchar) (*exchange->enclosed)[0] : field_term_char);
if (exchange->escaped->length() && (exchange->escaped_given() ||
!(thd->variables.sql_mode & MODE_NO_BACKSLASH_ESCAPES)))
escape_char= (int) (uchar) (*exchange->escaped)[0];
else
escape_char= -1;
is_ambiguous_field_sep= test(strchr(ESCAPE_CHARS, field_sep_char));
is_unsafe_field_sep= test(strchr(NUMERIC_CHARS, field_sep_char));
2000-07-31 21:29:14 +02:00
line_sep_char= (exchange->line_term->length() ?
(int) (uchar) (*exchange->line_term)[0] : INT_MAX);
2000-07-31 21:29:14 +02:00
if (!field_term_length)
exchange->opt_enclosed=0;
if (!exchange->enclosed->length())
exchange->opt_enclosed=1; // A little quicker loop
fixed_row_size= (!field_term_length && !exchange->enclosed->length() &&
!blob_flag);
if ((is_ambiguous_field_sep && exchange->enclosed->is_empty() &&
(string_results || is_unsafe_field_sep)) ||
(exchange->opt_enclosed && non_string_results &&
field_term_length && strchr(NUMERIC_CHARS, field_term_char)))
{
push_warning(thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_AMBIGUOUS_FIELD_TERM, ER(ER_AMBIGUOUS_FIELD_TERM));
is_ambiguous_field_term= TRUE;
}
else
is_ambiguous_field_term= FALSE;
2000-07-31 21:29:14 +02:00
return 0;
}
#define NEED_ESCAPING(x) ((int) (uchar) (x) == escape_char || \
(enclosed ? (int) (uchar) (x) == field_sep_char \
: (int) (uchar) (x) == field_term_char) || \
(int) (uchar) (x) == line_sep_char || \
!(x))
2000-07-31 21:29:14 +02:00
bool select_export::send_data(List<Item> &items)
{
DBUG_ENTER("select_export::send_data");
2000-07-31 21:29:14 +02:00
char buff[MAX_FIELD_WIDTH],null_buff[2],space[MAX_FIELD_WIDTH];
char cvt_buff[MAX_FIELD_WIDTH];
String cvt_str(cvt_buff, sizeof(cvt_buff), write_cs);
2000-07-31 21:29:14 +02:00
bool space_inited=0;
2003-01-29 14:31:20 +01:00
String tmp(buff,sizeof(buff),&my_charset_bin),*res;
2000-07-31 21:29:14 +02:00
tmp.length(0);
if (unit->offset_limit_cnt)
2000-07-31 21:29:14 +02:00
{ // using limit offset,count
unit->offset_limit_cnt--;
2000-07-31 21:29:14 +02:00
DBUG_RETURN(0);
}
row_count++;
Item *item;
uint used_length=0,items_left=items.elements;
List_iterator_fast<Item> li(items);
2000-07-31 21:29:14 +02:00
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) exchange->line_start->ptr(),
2000-07-31 21:29:14 +02:00
exchange->line_start->length()))
goto err;
while ((item=li++))
{
Item_result result_type=item->result_type();
bool enclosed = (exchange->enclosed->length() &&
(!exchange->opt_enclosed || result_type == STRING_RESULT));
2000-07-31 21:29:14 +02:00
res=item->str_result(&tmp);
if (res && !my_charset_same(write_cs, res->charset()) &&
!my_charset_same(write_cs, &my_charset_bin))
{
const char *well_formed_error_pos;
const char *cannot_convert_error_pos;
const char *from_end_pos;
const char *error_pos;
uint32 bytes;
bytes= well_formed_copy_nchars(write_cs, cvt_buff, sizeof(cvt_buff),
res->charset(), res->ptr(), res->length(),
sizeof(cvt_buff),
&well_formed_error_pos,
&cannot_convert_error_pos,
&from_end_pos);
error_pos= well_formed_error_pos ? well_formed_error_pos
: cannot_convert_error_pos;
if (error_pos)
{
char printable_buff[32];
convert_to_printable(printable_buff, sizeof(printable_buff),
error_pos, res->ptr() + res->length() - error_pos,
res->charset(), 6);
push_warning_printf(thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_TRUNCATED_WRONG_VALUE_FOR_FIELD,
ER(ER_TRUNCATED_WRONG_VALUE_FOR_FIELD),
"string", printable_buff,
item->name, row_count);
}
cvt_str.length(bytes);
res= &cvt_str;
}
if (res && enclosed)
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) exchange->enclosed->ptr(),
2000-07-31 21:29:14 +02:00
exchange->enclosed->length()))
goto err;
}
if (!res)
{ // NULL
if (!fixed_row_size)
{
if (escape_char != -1) // Use \N syntax
{
null_buff[0]=escape_char;
null_buff[1]='N';
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) null_buff,2))
2000-07-31 21:29:14 +02:00
goto err;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
else if (my_b_write(&cache,(uchar*) "NULL",4))
2000-07-31 21:29:14 +02:00
goto err;
}
else
{
used_length=0; // Fill with space
}
}
else
{
if (fixed_row_size)
used_length=min(res->length(),item->max_length);
else
used_length=res->length();
if ((result_type == STRING_RESULT || is_unsafe_field_sep) &&
escape_char != -1)
2000-07-31 21:29:14 +02:00
{
char *pos, *start, *end;
CHARSET_INFO *res_charset= res->charset();
CHARSET_INFO *character_set_client= thd->variables.
character_set_client;
bool check_second_byte= (res_charset == &my_charset_bin) &&
character_set_client->
escape_with_backslash_is_dangerous;
DBUG_ASSERT(character_set_client->mbmaxlen == 2 ||
!character_set_client->escape_with_backslash_is_dangerous);
2000-07-31 21:29:14 +02:00
for (start=pos=(char*) res->ptr(),end=pos+used_length ;
pos != end ;
pos++)
{
#ifdef USE_MB
2002-12-20 14:57:24 +01:00
if (use_mb(res_charset))
2000-07-31 21:29:14 +02:00
{
int l;
2002-12-20 14:57:24 +01:00
if ((l=my_ismbchar(res_charset, pos, end)))
2000-07-31 21:29:14 +02:00
{
pos += l-1;
continue;
}
}
#endif
/*
Special case when dumping BINARY/VARBINARY/BLOB values
for the clients with character sets big5, cp932, gbk and sjis,
which can have the escape character (0x5C "\" by default)
as the second byte of a multi-byte sequence.
If
- pos[0] is a valid multi-byte head (e.g 0xEE) and
- pos[1] is 0x00, which will be escaped as "\0",
then we'll get "0xEE + 0x5C + 0x30" in the output file.
If this file is later loaded using this sequence of commands:
mysql> create table t1 (a varchar(128)) character set big5;
mysql> LOAD DATA INFILE 'dump.txt' INTO TABLE t1;
then 0x5C will be misinterpreted as the second byte
of a multi-byte character "0xEE + 0x5C", instead of
escape character for 0x00.
To avoid this confusion, we'll escape the multi-byte
head character too, so the sequence "0xEE + 0x00" will be
dumped as "0x5C + 0xEE + 0x5C + 0x30".
Note, in the condition below we only check if
mbcharlen is equal to 2, because there are no
character sets with mbmaxlen longer than 2
and with escape_with_backslash_is_dangerous set.
DBUG_ASSERT before the loop makes that sure.
*/
if ((NEED_ESCAPING(*pos) ||
(check_second_byte &&
my_mbcharlen(character_set_client, (uchar) *pos) == 2 &&
pos + 1 < end &&
NEED_ESCAPING(pos[1]))) &&
/*
Don't escape field_term_char by doubling - doubling is only
valid for ENCLOSED BY characters:
*/
(enclosed || !is_ambiguous_field_term ||
(int) (uchar) *pos != field_term_char))
{
2000-07-31 21:29:14 +02:00
char tmp_buff[2];
tmp_buff[0]= ((int) (uchar) *pos == field_sep_char &&
is_ambiguous_field_sep) ?
field_sep_char : escape_char;
2000-07-31 21:29:14 +02:00
tmp_buff[1]= *pos ? *pos : '0';
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) start,(uint) (pos-start)) ||
my_b_write(&cache,(uchar*) tmp_buff,2))
2000-07-31 21:29:14 +02:00
goto err;
start=pos+1;
}
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) start,(uint) (pos-start)))
2000-07-31 21:29:14 +02:00
goto err;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
else if (my_b_write(&cache,(uchar*) res->ptr(),used_length))
2000-07-31 21:29:14 +02:00
goto err;
}
if (fixed_row_size)
{ // Fill with space
if (item->max_length > used_length)
{
/* QQ: Fix by adding a my_b_fill() function */
if (!space_inited)
{
space_inited=1;
bfill(space,sizeof(space),' ');
}
uint length=item->max_length-used_length;
for (; length > sizeof(space) ; length-=sizeof(space))
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) space,sizeof(space)))
2000-07-31 21:29:14 +02:00
goto err;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) space,length))
2000-07-31 21:29:14 +02:00
goto err;
}
}
if (res && enclosed)
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache, (uchar*) exchange->enclosed->ptr(),
exchange->enclosed->length()))
goto err;
2000-07-31 21:29:14 +02:00
}
if (--items_left)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache, (uchar*) exchange->field_term->ptr(),
field_term_length))
goto err;
2000-07-31 21:29:14 +02:00
}
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) exchange->line_term->ptr(),
2000-07-31 21:29:14 +02:00
exchange->line_term->length()))
goto err;
DBUG_RETURN(0);
err:
DBUG_RETURN(1);
}
/***************************************************************************
** Dump of select to a binary file
***************************************************************************/
int
select_dump::prepare(List<Item> &list __attribute__((unused)),
SELECT_LEX_UNIT *u)
2000-07-31 21:29:14 +02:00
{
unit= u;
return (int) ((file= create_file(thd, path, exchange, &cache)) < 0);
2000-07-31 21:29:14 +02:00
}
bool select_dump::send_data(List<Item> &items)
{
List_iterator_fast<Item> li(items);
2000-07-31 21:29:14 +02:00
char buff[MAX_FIELD_WIDTH];
2003-01-29 14:31:20 +01:00
String tmp(buff,sizeof(buff),&my_charset_bin),*res;
2000-07-31 21:29:14 +02:00
tmp.length(0);
Item *item;
DBUG_ENTER("select_dump::send_data");
2000-07-31 21:29:14 +02:00
if (unit->offset_limit_cnt)
2000-07-31 21:29:14 +02:00
{ // using limit offset,count
unit->offset_limit_cnt--;
2000-07-31 21:29:14 +02:00
DBUG_RETURN(0);
}
if (row_count++ > 1)
{
2004-11-12 13:34:00 +01:00
my_message(ER_TOO_MANY_ROWS, ER(ER_TOO_MANY_ROWS), MYF(0));
2000-07-31 21:29:14 +02:00
goto err;
}
while ((item=li++))
{
res=item->str_result(&tmp);
if (!res) // If NULL
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_b_write(&cache,(uchar*) "",1))
goto err;
2000-07-31 21:29:14 +02:00
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
else if (my_b_write(&cache,(uchar*) res->ptr(),res->length()))
2000-07-31 21:29:14 +02:00
{
my_error(ER_ERROR_ON_WRITE, MYF(0), path, my_errno);
2000-07-31 21:29:14 +02:00
goto err;
}
}
DBUG_RETURN(0);
err:
DBUG_RETURN(1);
}
select_subselect::select_subselect(Item_subselect *item_arg)
{
item= item_arg;
}
2002-12-19 20:15:09 +01:00
bool select_singlerow_subselect::send_data(List<Item> &items)
{
2002-12-19 20:15:09 +01:00
DBUG_ENTER("select_singlerow_subselect::send_data");
Item_singlerow_subselect *it= (Item_singlerow_subselect *)item;
2002-11-18 21:41:57 +01:00
if (it->assigned())
{
my_message(ER_SUBQUERY_NO_1_ROW, ER(ER_SUBQUERY_NO_1_ROW), MYF(0));
2002-05-28 21:38:17 +02:00
DBUG_RETURN(1);
}
if (unit->offset_limit_cnt)
{ // Using limit offset,count
2002-05-28 21:38:17 +02:00
unit->offset_limit_cnt--;
DBUG_RETURN(0);
}
List_iterator_fast<Item> li(items);
Item *val_item;
for (uint i= 0; (val_item= li++); i++)
it->store(i, val_item);
it->assigned(1);
2002-05-28 21:38:17 +02:00
DBUG_RETURN(0);
}
void select_max_min_finder_subselect::cleanup()
{
DBUG_ENTER("select_max_min_finder_subselect::cleanup");
cache= 0;
DBUG_VOID_RETURN;
}
bool select_max_min_finder_subselect::send_data(List<Item> &items)
{
DBUG_ENTER("select_max_min_finder_subselect::send_data");
Item_maxmin_subselect *it= (Item_maxmin_subselect *)item;
List_iterator_fast<Item> li(items);
Item *val_item= li++;
it->register_value();
if (it->assigned())
{
cache->store(val_item);
if ((this->*op)())
it->store(0, cache);
}
else
{
if (!cache)
{
cache= Item_cache::get_cache(val_item);
switch (val_item->result_type())
{
case REAL_RESULT:
op= &select_max_min_finder_subselect::cmp_real;
break;
case INT_RESULT:
op= &select_max_min_finder_subselect::cmp_int;
break;
case STRING_RESULT:
op= &select_max_min_finder_subselect::cmp_str;
break;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
op= &select_max_min_finder_subselect::cmp_decimal;
break;
case ROW_RESULT:
// This case should never be choosen
DBUG_ASSERT(0);
op= 0;
}
}
cache->store(val_item);
it->store(0, cache);
}
it->assigned(1);
DBUG_RETURN(0);
}
bool select_max_min_finder_subselect::cmp_real()
{
Item *maxmin= ((Item_singlerow_subselect *)item)->element_index(0);
2004-11-11 19:39:35 +01:00
double val1= cache->val_real(), val2= maxmin->val_real();
if (fmax)
return (cache->null_value && !maxmin->null_value) ||
(!cache->null_value && !maxmin->null_value &&
val1 > val2);
return (maxmin->null_value && !cache->null_value) ||
(!cache->null_value && !maxmin->null_value &&
val1 < val2);
}
bool select_max_min_finder_subselect::cmp_int()
{
Item *maxmin= ((Item_singlerow_subselect *)item)->element_index(0);
longlong val1= cache->val_int(), val2= maxmin->val_int();
if (fmax)
return (cache->null_value && !maxmin->null_value) ||
(!cache->null_value && !maxmin->null_value &&
val1 > val2);
return (maxmin->null_value && !cache->null_value) ||
(!cache->null_value && !maxmin->null_value &&
val1 < val2);
}
2005-02-08 23:50:45 +01:00
bool select_max_min_finder_subselect::cmp_decimal()
{
Item *maxmin= ((Item_singlerow_subselect *)item)->element_index(0);
2005-02-08 23:50:45 +01:00
my_decimal cval, *cvalue= cache->val_decimal(&cval);
my_decimal mval, *mvalue= maxmin->val_decimal(&mval);
if (fmax)
return (cache->null_value && !maxmin->null_value) ||
(!cache->null_value && !maxmin->null_value &&
my_decimal_cmp(cvalue, mvalue) > 0) ;
return (maxmin->null_value && !cache->null_value) ||
(!cache->null_value && !maxmin->null_value &&
my_decimal_cmp(cvalue,mvalue) < 0);
2005-02-08 23:50:45 +01:00
}
bool select_max_min_finder_subselect::cmp_str()
{
String *val1, *val2, buf1, buf2;
Item *maxmin= ((Item_singlerow_subselect *)item)->element_index(0);
/*
as far as both operand is Item_cache buf1 & buf2 will not be used,
but added for safety
*/
val1= cache->val_str(&buf1);
val2= maxmin->val_str(&buf1);
if (fmax)
return (cache->null_value && !maxmin->null_value) ||
(!cache->null_value && !maxmin->null_value &&
sortcmp(val1, val2, cache->collation.collation) > 0) ;
return (maxmin->null_value && !cache->null_value) ||
(!cache->null_value && !maxmin->null_value &&
sortcmp(val1, val2, cache->collation.collation) < 0);
}
bool select_exists_subselect::send_data(List<Item> &items)
{
DBUG_ENTER("select_exists_subselect::send_data");
Item_exists_subselect *it= (Item_exists_subselect *)item;
if (unit->offset_limit_cnt)
{ // Using limit offset,count
unit->offset_limit_cnt--;
DBUG_RETURN(0);
}
it->value= 1;
it->assigned(1);
DBUG_RETURN(0);
}
2002-10-11 20:49:10 +02:00
/***************************************************************************
Dump of select to variables
2002-10-11 20:49:10 +02:00
***************************************************************************/
2002-10-16 15:55:08 +02:00
int select_dumpvar::prepare(List<Item> &list, SELECT_LEX_UNIT *u)
2002-10-11 20:49:10 +02:00
{
unit= u;
2002-10-16 15:55:08 +02:00
if (var_list.elements != list.elements)
2002-10-11 20:49:10 +02:00
{
2004-11-12 13:34:00 +01:00
my_message(ER_WRONG_NUMBER_OF_COLUMNS_IN_SELECT,
ER(ER_WRONG_NUMBER_OF_COLUMNS_IN_SELECT), MYF(0));
2002-10-16 15:55:08 +02:00
return 1;
}
2002-10-16 15:55:08 +02:00
return 0;
}
bool select_dumpvar::check_simple_select() const
{
my_error(ER_SP_BAD_CURSOR_SELECT, MYF(0));
return TRUE;
}
void select_dumpvar::cleanup()
{
row_count= 0;
}
Query_arena::Type Query_arena::type() const
{
DBUG_ASSERT(0); /* Should never be called */
return STATEMENT;
}
void Query_arena::free_items()
{
Item *next;
DBUG_ENTER("Query_arena::free_items");
/* This works because items are allocated with sql_alloc() */
for (; free_list; free_list= next)
{
next= free_list->next;
free_list->delete_self();
}
/* Postcondition: free_list is 0 */
DBUG_VOID_RETURN;
}
void Query_arena::set_query_arena(Query_arena *set)
{
mem_root= set->mem_root;
free_list= set->free_list;
state= set->state;
}
void Query_arena::cleanup_stmt()
{
DBUG_ASSERT(! "Query_arena::cleanup_stmt() not implemented");
}
/*
2007-04-13 19:23:02 +02:00
Statement functions
*/
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
Statement::Statement(LEX *lex_arg, MEM_ROOT *mem_root_arg,
enum enum_state state_arg, ulong id_arg)
:Query_arena(mem_root_arg, state_arg),
2005-06-22 21:12:25 +02:00
id(id_arg),
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
mark_used_columns(MARK_COLUMNS_READ),
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
lex(lex_arg),
cursor(0),
db(NULL),
db_length(0)
{
query_string.length= 0;
query_string.str= NULL;
name.str= NULL;
}
Query_arena::Type Statement::type() const
{
return STATEMENT;
}
void Statement::set_statement(Statement *stmt)
{
id= stmt->id;
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
mark_used_columns= stmt->mark_used_columns;
lex= stmt->lex;
query_string= stmt->query_string;
cursor= stmt->cursor;
}
void
Statement::set_n_backup_statement(Statement *stmt, Statement *backup)
{
DBUG_ENTER("Statement::set_n_backup_statement");
backup->set_statement(this);
set_statement(stmt);
DBUG_VOID_RETURN;
}
void Statement::restore_backup_statement(Statement *stmt, Statement *backup)
{
DBUG_ENTER("Statement::restore_backup_statement");
stmt->set_statement(this);
set_statement(backup);
DBUG_VOID_RETURN;
}
/** Assign a new value to thd->query. */
void Statement::set_query_inner(char *query_arg, uint32 query_length_arg)
{
query_string.str= query_arg;
query_string.length= query_length_arg;
}
void THD::end_statement()
{
A fix for Bug#26750 "valgrind leak in sp_head" (and post-review fixes). The legend: on a replication slave, in case a trigger creation was filtered out because of application of replicate-do-table/ replicate-ignore-table rule, the parsed definition of a trigger was not cleaned up properly. LEX::sphead member was left around and leaked memory. Until the actual implementation of support of replicate-ignore-table rules for triggers by the patch for Bug 24478 it was never the case that "case SQLCOM_CREATE_TRIGGER" was not executed once a trigger was parsed, so the deletion of lex->sphead there worked and the memory did not leak. The fix: The real cause of the bug is that there is no 1 or 2 places where we can clean up the main LEX after parse. And the reason we can not have just one or two places where we clean up the LEX is asymmetric behaviour of MYSQLparse in case of success or error. One of the root causes of this behaviour is the code in Item::Item() constructor. There, a newly created item adds itself to THD::free_list - a single-linked list of Items used in a statement. Yuck. This code is unaware that we may have more than one statement active at a time, and always assumes that the free_list of the current statement is located in THD::free_list. One day we need to be able to explicitly allocate an item in a given Query_arena. Thus, when parsing a definition of a stored procedure, like CREATE PROCEDURE p1() BEGIN SELECT a FROM t1; SELECT b FROM t1; END; we actually need to reset THD::mem_root, THD::free_list and THD::lex to parse the nested procedure statement (SELECT *). The actual reset and restore is implemented in semantic actions attached to sp_proc_stmt grammar rule. The problem is that in case of a parsing error inside a nested statement Bison generated parser would abort immediately, without executing the restore part of the semantic action. This would leave THD in an in-the-middle-of-parsing state. This is why we couldn't have had a single place where we clean up the LEX after MYSQLparse - in case of an error we needed to do a clean up immediately, in case of success a clean up could have been delayed. This left the door open for a memory leak. One of the following possibilities were considered when working on a fix: - patch the replication logic to do the clean up. Rejected as breaks module borders, replication code should not need to know the gory details of clean up procedure after CREATE TRIGGER. - wrap MYSQLparse with a function that would do a clean up. Rejected as ideally we should fix the problem when it happens, not adjust for it outside of the problematic code. - make sure MYSQLparse cleans up after itself by invoking the clean up functionality in the appropriate places before return. Implemented in this patch. - use %destructor rule for sp_proc_stmt to restore THD - cleaner than the prevoius approach, but rejected because needs a careful analysis of the side effects, and this patch is for 5.0, and long term we need to use the next alternative anyway - make sure that sp_proc_stmt doesn't juggle with THD - this is a large work that will affect many modules. Cleanup: move main_lex and main_mem_root from Statement to its only two descendants Prepared_statement and THD. This ensures that when a Statement instance was created for purposes of statement backup, we do not involve LEX constructor/destructor, which is fairly expensive. In order to track that the transformation produces equivalent functionality please check the respective constructors and destructors of Statement, Prepared_statement and THD - these members were used only there. This cleanup is unrelated to the patch.
2007-03-07 10:24:46 +01:00
/* Cleanup SQL processing state to reuse this statement in next query. */
lex_end(lex);
delete lex->result;
lex->result= 0;
/* Note that free_list is freed in cleanup_after_query() */
/*
Don't free mem_root, as mem_root is freed in the end of dispatch_command
(once for any command).
*/
}
void THD::set_n_backup_active_arena(Query_arena *set, Query_arena *backup)
{
DBUG_ENTER("THD::set_n_backup_active_arena");
DBUG_ASSERT(backup->is_backup_arena == FALSE);
backup->set_query_arena(this);
set_query_arena(set);
#ifndef DBUG_OFF
backup->is_backup_arena= TRUE;
#endif
DBUG_VOID_RETURN;
}
void THD::restore_active_arena(Query_arena *set, Query_arena *backup)
{
DBUG_ENTER("THD::restore_active_arena");
DBUG_ASSERT(backup->is_backup_arena);
set->set_query_arena(this);
set_query_arena(backup);
#ifndef DBUG_OFF
backup->is_backup_arena= FALSE;
#endif
DBUG_VOID_RETURN;
}
Statement::~Statement()
{
}
C_MODE_START
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
static uchar *
get_statement_id_as_hash_key(const uchar *record, size_t *key_length,
my_bool not_used __attribute__((unused)))
{
const Statement *statement= (const Statement *) record;
*key_length= sizeof(statement->id);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
return (uchar *) &((const Statement *) statement)->id;
}
static void delete_statement_as_hash_key(void *key)
{
delete (Statement *) key;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
static uchar *get_stmt_name_hash_key(Statement *entry, size_t *length,
my_bool not_used __attribute__((unused)))
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
*length= entry->name.length;
return (uchar*) entry->name.str;
}
C_MODE_END
Statement_map::Statement_map() :
last_found_statement(0)
{
enum
{
START_STMT_HASH_SIZE = 16,
START_NAME_HASH_SIZE = 16
};
my_hash_init(&st_hash, &my_charset_bin, START_STMT_HASH_SIZE, 0, 0,
get_statement_id_as_hash_key,
delete_statement_as_hash_key, MYF(0));
my_hash_init(&names_hash, system_charset_info, START_NAME_HASH_SIZE, 0, 0,
(my_hash_get_key) get_stmt_name_hash_key,
NULL,MYF(0));
}
/*
Insert a new statement to the thread-local statement map.
DESCRIPTION
If there was an old statement with the same name, replace it with the
new one. Otherwise, check if max_prepared_stmt_count is not reached yet,
increase prepared_stmt_count, and insert the new statement. It's okay
to delete an old statement and fail to insert the new one.
POSTCONDITIONS
All named prepared statements are also present in names_hash.
Statement names in names_hash are unique.
The statement is added only if prepared_stmt_count < max_prepard_stmt_count
last_found_statement always points to a valid statement or is 0
RETURN VALUE
0 success
1 error: out of resources or max_prepared_stmt_count limit has been
reached. An error is sent to the client, the statement is deleted.
*/
int Statement_map::insert(THD *thd, Statement *statement)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_hash_insert(&st_hash, (uchar*) statement))
{
/*
Delete is needed only in case of an insert failure. In all other
cases hash_delete will also delete the statement.
*/
delete statement;
my_error(ER_OUT_OF_RESOURCES, MYF(0));
goto err_st_hash;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (statement->name.str && my_hash_insert(&names_hash, (uchar*) statement))
{
my_error(ER_OUT_OF_RESOURCES, MYF(0));
goto err_names_hash;
}
mysql_mutex_lock(&LOCK_prepared_stmt_count);
/*
We don't check that prepared_stmt_count is <= max_prepared_stmt_count
because we would like to allow to lower the total limit
of prepared statements below the current count. In that case
no new statements can be added until prepared_stmt_count drops below
the limit.
*/
if (prepared_stmt_count >= max_prepared_stmt_count)
{
mysql_mutex_unlock(&LOCK_prepared_stmt_count);
my_error(ER_MAX_PREPARED_STMT_COUNT_REACHED, MYF(0),
max_prepared_stmt_count);
goto err_max;
}
prepared_stmt_count++;
mysql_mutex_unlock(&LOCK_prepared_stmt_count);
last_found_statement= statement;
return 0;
err_max:
if (statement->name.str)
my_hash_delete(&names_hash, (uchar*) statement);
err_names_hash:
my_hash_delete(&st_hash, (uchar*) statement);
err_st_hash:
return 1;
}
void Statement_map::close_transient_cursors()
{
#ifdef TO_BE_IMPLEMENTED
Statement *stmt;
while ((stmt= transient_cursor_list.head()))
stmt->close_cursor(); /* deletes itself from the list */
#endif
}
void Statement_map::erase(Statement *statement)
{
if (statement == last_found_statement)
last_found_statement= 0;
if (statement->name.str)
my_hash_delete(&names_hash, (uchar *) statement);
my_hash_delete(&st_hash, (uchar *) statement);
mysql_mutex_lock(&LOCK_prepared_stmt_count);
DBUG_ASSERT(prepared_stmt_count > 0);
prepared_stmt_count--;
mysql_mutex_unlock(&LOCK_prepared_stmt_count);
}
void Statement_map::reset()
{
/* Must be first, hash_free will reset st_hash.records */
mysql_mutex_lock(&LOCK_prepared_stmt_count);
DBUG_ASSERT(prepared_stmt_count >= st_hash.records);
prepared_stmt_count-= st_hash.records;
mysql_mutex_unlock(&LOCK_prepared_stmt_count);
my_hash_reset(&names_hash);
my_hash_reset(&st_hash);
last_found_statement= 0;
}
Statement_map::~Statement_map()
{
/* Must go first, hash_free will reset st_hash.records */
mysql_mutex_lock(&LOCK_prepared_stmt_count);
DBUG_ASSERT(prepared_stmt_count >= st_hash.records);
prepared_stmt_count-= st_hash.records;
mysql_mutex_unlock(&LOCK_prepared_stmt_count);
my_hash_free(&names_hash);
my_hash_free(&st_hash);
}
2002-10-16 15:55:08 +02:00
bool select_dumpvar::send_data(List<Item> &items)
{
List_iterator_fast<my_var> var_li(var_list);
List_iterator<Item> it(items);
Item *item;
my_var *mv;
DBUG_ENTER("select_dumpvar::send_data");
2002-10-16 15:55:08 +02:00
2003-01-18 15:23:37 +01:00
if (unit->offset_limit_cnt)
{ // using limit offset,count
2003-01-18 15:23:37 +01:00
unit->offset_limit_cnt--;
DBUG_RETURN(0);
}
2002-10-16 15:55:08 +02:00
if (row_count++)
{
2004-11-12 13:34:00 +01:00
my_message(ER_TOO_MANY_ROWS, ER(ER_TOO_MANY_ROWS), MYF(0));
2002-10-16 15:55:08 +02:00
DBUG_RETURN(1);
}
while ((mv= var_li++) && (item= it++))
2003-10-24 20:14:26 +02:00
{
if (mv->local)
{
if (thd->spcont->set_variable(thd, mv->offset, &item))
DBUG_RETURN(1);
}
else
{
Item_func_set_user_var *suv= new Item_func_set_user_var(mv->s, item);
if (suv->fix_fields(thd, 0))
DBUG_RETURN (1);
suv->save_item_result(item);
if (suv->update())
DBUG_RETURN (1);
}
2003-10-24 20:14:26 +02:00
}
DBUG_RETURN(thd->is_error());
2002-10-11 20:49:10 +02:00
}
bool select_dumpvar::send_eof()
{
if (! row_count)
2004-11-24 17:22:29 +01:00
push_warning(thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_SP_FETCH_NO_DATA, ER(ER_SP_FETCH_NO_DATA));
::my_ok(thd,row_count);
return 0;
2002-10-11 20:49:10 +02:00
}
/****************************************************************************
TMP_TABLE_PARAM
****************************************************************************/
void TMP_TABLE_PARAM::init()
{
DBUG_ENTER("TMP_TABLE_PARAM::init");
DBUG_PRINT("enter", ("this: 0x%lx", (ulong)this));
field_count= sum_func_count= func_count= hidden_field_count= 0;
group_parts= group_length= group_null_parts= 0;
quick_group= 1;
table_charset= 0;
precomputed_group_by= 0;
DBUG_VOID_RETURN;
}
void thd_increment_bytes_sent(ulong length)
{
2005-03-09 19:22:30 +01:00
THD *thd=current_thd;
2005-03-09 20:49:44 +01:00
if (likely(thd != 0))
2005-03-09 19:22:30 +01:00
{ /* current_thd==0 when close_connection() calls net_send_error() */
thd->status_var.bytes_sent+= length;
}
}
void thd_increment_bytes_received(ulong length)
{
current_thd->status_var.bytes_received+= length;
}
void thd_increment_net_big_packet_count(ulong length)
{
current_thd->status_var.net_big_packet_count+= length;
}
void THD::set_status_var_init()
{
bzero((char*) &status_var, sizeof(status_var));
}
void Security_context::init()
{
host= user= priv_user= ip= 0;
host_or_ip= "connecting host";
priv_host[0]= '\0';
master_access= 0;
#ifndef NO_EMBEDDED_ACCESS_CHECKS
db_access= NO_ACCESS;
#endif
}
void Security_context::destroy()
{
// If not pointer to constant
if (host != my_localhost)
safeFree(host);
if (user != delayed_user)
safeFree(user);
safeFree(ip);
}
void Security_context::skip_grants()
{
/* privileges for the user are unknown everything is allowed */
host_or_ip= (char *)"";
master_access= ~NO_ACCESS;
priv_user= (char *)"";
*priv_host= '\0';
}
bool Security_context::set_user(char *user_arg)
{
safeFree(user);
user= my_strdup(user_arg, MYF(0));
return user == 0;
}
#ifndef NO_EMBEDDED_ACCESS_CHECKS
/**
Initialize this security context from the passed in credentials
and activate it in the current thread.
2007-08-16 16:07:48 +02:00
@param thd
@param definer_user
@param definer_host
@param db
@param[out] backup Save a pointer to the current security context
in the thread. In case of success it points to the
saved old context, otherwise it points to NULL.
During execution of a statement, multiple security contexts may
be needed:
- the security context of the authenticated user, used as the
default security context for all top-level statements
- in case of a view or a stored program, possibly the security
context of the definer of the routine, if the object is
defined with SQL SECURITY DEFINER option.
The currently "active" security context is parameterized in THD
member security_ctx. By default, after a connection is
established, this member points at the "main" security context
- the credentials of the authenticated user.
Later, if we would like to execute some sub-statement or a part
of a statement under credentials of a different user, e.g.
definer of a procedure, we authenticate this user in a local
instance of Security_context by means of this method (and
ultimately by means of acl_getroot_no_password), and make the
local instance active in the thread by re-setting
thd->security_ctx pointer.
Note, that the life cycle and memory management of the "main" and
temporary security contexts are different.
For the main security context, the memory for user/host/ip is
allocated on system heap, and the THD class frees this memory in
its destructor. The only case when contents of the main security
context may change during its life time is when someone issued
CHANGE USER command.
Memory management of a "temporary" security context is
responsibility of the module that creates it.
@retval TRUE there is no user with the given credentials. The erro
is reported in the thread.
@retval FALSE success
*/
bool
Security_context::
change_security_context(THD *thd,
LEX_STRING *definer_user,
LEX_STRING *definer_host,
LEX_STRING *db,
Security_context **backup)
{
bool needs_change;
DBUG_ENTER("Security_context::change_security_context");
DBUG_ASSERT(definer_user->str && definer_host->str);
*backup= NULL;
/*
The current security context may have NULL members
if we have just started the thread and not authenticated
any user. This use case is currently in events worker thread.
*/
needs_change= (thd->security_ctx->priv_user == NULL ||
strcmp(definer_user->str, thd->security_ctx->priv_user) ||
thd->security_ctx->priv_host == NULL ||
my_strcasecmp(system_charset_info, definer_host->str,
thd->security_ctx->priv_host));
if (needs_change)
{
if (acl_getroot_no_password(this, definer_user->str, definer_host->str,
definer_host->str, db->str))
{
my_error(ER_NO_SUCH_USER, MYF(0), definer_user->str,
definer_host->str);
DBUG_RETURN(TRUE);
}
*backup= thd->security_ctx;
thd->security_ctx= this;
}
DBUG_RETURN(FALSE);
}
void
Security_context::restore_security_context(THD *thd,
Security_context *backup)
{
if (backup)
thd->security_ctx= backup;
}
#endif
bool Security_context::user_matches(Security_context *them)
{
return ((user != NULL) && (them->user != NULL) &&
!strcmp(user, them->user));
}
/****************************************************************************
Handling of open and locked tables states.
This is used when we want to open/lock (and then close) some tables when
we already have a set of tables open and locked. We use these methods for
access to mysql.proc table to find definitions of stored routines.
****************************************************************************/
void THD::reset_n_backup_open_tables_state(Open_tables_backup *backup)
{
DBUG_ENTER("reset_n_backup_open_tables_state");
backup->set_open_tables_state(this);
backup->mdl_system_tables_svp= mdl_context.mdl_savepoint();
Initial import of WL#3726 "DDL locking for all metadata objects". Backport of: ------------------------------------------------------------ revno: 2630.4.1 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Fri 2008-05-23 17:54:03 +0400 message: WL#3726 "DDL locking for all metadata objects". After review fixes in progress. ------------------------------------------------------------ This is the first patch in series. It transforms the metadata locking subsystem to use a dedicated module (mdl.h,cc). No significant changes in the locking protocol. The import passes the test suite with the exception of deprecated/removed 6.0 features, and MERGE tables. The latter are subject to a fix by WL#4144. Unfortunately, the original changeset comments got lost in a merge, thus this import has its own (largely insufficient) comments. This patch fixes Bug#25144 "replication / binlog with view breaks". Warning: this patch introduces an incompatible change: Under LOCK TABLES, it's no longer possible to FLUSH a table that was not locked for WRITE. Under LOCK TABLES, it's no longer possible to DROP a table or VIEW that was not locked for WRITE. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.2 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 14:03:45 +0400 message: WL#3726 "DDL locking for all metadata objects". After review fixes in progress. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.3 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 14:08:51 +0400 message: WL#3726 "DDL locking for all metadata objects" Fixed failing Windows builds by adding mdl.cc to the lists of files needed to build server/libmysqld on Windows. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.4 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 21:57:58 +0400 message: WL#3726 "DDL locking for all metadata objects". Fix for assert failures in kill.test which occured when one tried to kill ALTER TABLE statement on merge table while it was waiting in wait_while_table_is_used() for other connections to close this table. These assert failures stemmed from the fact that cleanup code in this case assumed that temporary table representing new version of table was open with adding to THD::temporary_tables list while code which were opening this temporary table wasn't always fulfilling this. This patch changes code that opens new version of table to always do this linking in. It also streamlines cleanup process for cases when error occurs while we have new version of table open. ****** WL#3726 "DDL locking for all metadata objects" Add libmysqld/mdl.cc to .bzrignore. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.6 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sun 2008-05-25 00:33:22 +0400 message: WL#3726 "DDL locking for all metadata objects". Addition to the fix of assert failures in kill.test caused by changes for this worklog. Make sure we close the new table only once.
2009-11-30 16:55:03 +01:00
reset_open_tables_state(this);
state_flags|= Open_tables_state::BACKUPS_AVAIL;
DBUG_VOID_RETURN;
}
void THD::restore_backup_open_tables_state(Open_tables_backup *backup)
{
DBUG_ENTER("restore_backup_open_tables_state");
mdl_context.rollback_to_savepoint(backup->mdl_system_tables_svp);
/*
Before we will throw away current open tables state we want
to be sure that it was properly cleaned up.
*/
DBUG_ASSERT(open_tables == 0 && temporary_tables == 0 &&
A prerequisite patch for the fix for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". Introduce a notion of a sentinel to MDL_context. A sentinel is a ticket that separates all tickets in the context into two groups: before and after it. Currently we can have (and need) only one designated sentinel -- it separates all locks taken by LOCK TABLE or HANDLER statement, which must survive COMMIT and ROLLBACK and all other locks, which must be released at COMMIT or ROLLBACK. The tricky part is maintaining the sentinel up to date when someone release its corresponding ticket. This can happen, e.g. if someone issues DROP TABLE under LOCK TABLES (generally, see all calls to release_all_locks_for_name()). MDL_context::release_ticket() is modified to take care of it. ****** A fix and a test case for Bug#46224 "HANDLER statements within a transaction might lead to deadlocks". An attempt to mix HANDLER SQL statements, which are transaction- agnostic, an open multi-statement transaction, and DDL against the involved tables (in a concurrent connection) could lead to a deadlock. The deadlock would occur when HANDLER OPEN or HANDLER READ would have to wait on a conflicting metadata lock. If the connection that issued HANDLER statement also had other metadata locks (say, acquired in scope of a transaction), a classical deadlock situation of mutual wait could occur. Incompatible change: entering LOCK TABLES mode automatically closes all open HANDLERs in the current connection. Incompatible change: previously an attempt to wait on a lock in a connection that has an open HANDLER statement could wait indefinitely/deadlock. After this patch, an error ER_LOCK_DEADLOCK is produced. The idea of the fix is to merge thd->handler_mdl_context with the main mdl_context of the connection, used for transactional locks. This makes deadlock detection possible, since all waits with locks are "visible" and available to analysis in a single MDL context of the connection. Since HANDLER locks and transactional locks have a different life cycle -- HANDLERs are explicitly open and closed, and so are HANDLER locks, explicitly acquired and released, whereas transactional locks "accumulate" till the end of a transaction and are released only with COMMIT, ROLLBACK and ROLLBACK TO SAVEPOINT, a concept of "sentinel" was introduced to MDL_context. All locks, HANDLER and others, reside in the same linked list. However, a selected element of the list separates locks with different life cycle. HANDLER locks always reside at the end of the list, after the sentinel. Transactional locks are prepended to the beginning of the list, before the sentinel. Thus, ROLLBACK, COMMIT or ROLLBACK TO SAVEPOINT, only release those locks that reside before the sentinel. HANDLER locks must be released explicitly as part of HANDLER CLOSE statement, or an implicit close. The same approach with sentinel is also employed for LOCK TABLES locks. Since HANDLER and LOCK TABLES statement has never worked together, the implementation is made simple and only maintains one sentinel, which is used either for HANDLER locks, or for LOCK TABLES locks.
2009-12-22 17:09:15 +01:00
derived_tables == 0 &&
lock == 0 &&
locked_tables_mode == LTM_NONE &&
m_reprepare_observer == NULL);
Initial import of WL#3726 "DDL locking for all metadata objects". Backport of: ------------------------------------------------------------ revno: 2630.4.1 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Fri 2008-05-23 17:54:03 +0400 message: WL#3726 "DDL locking for all metadata objects". After review fixes in progress. ------------------------------------------------------------ This is the first patch in series. It transforms the metadata locking subsystem to use a dedicated module (mdl.h,cc). No significant changes in the locking protocol. The import passes the test suite with the exception of deprecated/removed 6.0 features, and MERGE tables. The latter are subject to a fix by WL#4144. Unfortunately, the original changeset comments got lost in a merge, thus this import has its own (largely insufficient) comments. This patch fixes Bug#25144 "replication / binlog with view breaks". Warning: this patch introduces an incompatible change: Under LOCK TABLES, it's no longer possible to FLUSH a table that was not locked for WRITE. Under LOCK TABLES, it's no longer possible to DROP a table or VIEW that was not locked for WRITE. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.2 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 14:03:45 +0400 message: WL#3726 "DDL locking for all metadata objects". After review fixes in progress. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.3 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 14:08:51 +0400 message: WL#3726 "DDL locking for all metadata objects" Fixed failing Windows builds by adding mdl.cc to the lists of files needed to build server/libmysqld on Windows. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.4 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sat 2008-05-24 21:57:58 +0400 message: WL#3726 "DDL locking for all metadata objects". Fix for assert failures in kill.test which occured when one tried to kill ALTER TABLE statement on merge table while it was waiting in wait_while_table_is_used() for other connections to close this table. These assert failures stemmed from the fact that cleanup code in this case assumed that temporary table representing new version of table was open with adding to THD::temporary_tables list while code which were opening this temporary table wasn't always fulfilling this. This patch changes code that opens new version of table to always do this linking in. It also streamlines cleanup process for cases when error occurs while we have new version of table open. ****** WL#3726 "DDL locking for all metadata objects" Add libmysqld/mdl.cc to .bzrignore. ****** Backport of: ------------------------------------------------------------ revno: 2630.4.6 committer: Dmitry Lenev <dlenev@mysql.com> branch nick: mysql-6.0-3726-w timestamp: Sun 2008-05-25 00:33:22 +0400 message: WL#3726 "DDL locking for all metadata objects". Addition to the fix of assert failures in kill.test caused by changes for this worklog. Make sure we close the new table only once.
2009-11-30 16:55:03 +01:00
set_open_tables_state(backup);
DBUG_VOID_RETURN;
}
/**
Check the killed state of a user thread
@param thd user thread
@retval 0 the user thread is active
@retval 1 the user thread has been killed
*/
extern "C" int thd_killed(const MYSQL_THD thd)
{
return(thd->killed);
}
/**
Return the thread id of a user thread
@param thd user thread
@return thread id
*/
extern "C" unsigned long thd_get_thread_id(const MYSQL_THD thd)
{
return((unsigned long)thd->thread_id);
}
#ifdef INNODB_COMPATIBILITY_HOOKS
extern "C" struct charset_info_st *thd_charset(MYSQL_THD thd)
{
return(thd->charset());
}
/**
OBSOLETE : there's no way to ensure the string is null terminated.
Use thd_query_string instead()
*/
extern "C" char **thd_query(MYSQL_THD thd)
{
return(&thd->query_string.str);
}
/**
Get the current query string for the thread.
@param The MySQL internal thread pointer
@return query string and length. May be non-null-terminated.
*/
extern "C" LEX_STRING * thd_query_string (MYSQL_THD thd)
{
return(&thd->query_string);
}
extern "C" int thd_slave_thread(const MYSQL_THD thd)
{
return(thd->slave_thread);
}
extern "C" int thd_non_transactional_update(const MYSQL_THD thd)
{
return(thd->transaction.all.modified_non_trans_table);
}
extern "C" int thd_binlog_format(const MYSQL_THD thd)
{
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
if (mysql_bin_log.is_open() && (thd->variables.option_bits & OPTION_BIN_LOG))
return (int) thd->variables.binlog_format;
else
return BINLOG_FORMAT_UNSPEC;
}
extern "C" void thd_mark_transaction_to_rollback(MYSQL_THD thd, bool all)
{
mark_transaction_to_rollback(thd, all);
}
extern "C" bool thd_binlog_filter_ok(const MYSQL_THD thd)
{
return binlog_filter->db_ok(thd->db);
}
#endif // INNODB_COMPATIBILITY_HOOKS */
/****************************************************************************
Handling of statement states in functions and triggers.
This is used to ensure that the function/trigger gets a clean state
to work with and does not cause any side effects of the calling statement.
It also allows most stored functions and triggers to replicate even
if they are used items that would normally be stored in the binary
replication (like last_insert_id() etc...)
The following things is done
- Disable binary logging for the duration of the statement
- Disable multi-result-sets for the duration of the statement
- Value of last_insert_id() is saved and restored
- Value set by 'SET INSERT_ID=#' is reset and restored
- Value for found_rows() is reset and restored
- examined_row_count is added to the total
- cuted_fields is added to the total
- new savepoint level is created and destroyed
NOTES:
Seed for random() is saved for the first! usage of RAND()
We reset examined_row_count and cuted_fields and add these to the
result to ensure that if we have a bug that would reset these within
a function, we are not loosing any rows from the main statement.
We do not reset value of last_insert_id().
****************************************************************************/
void THD::reset_sub_statement_state(Sub_statement_state *backup,
uint new_state)
{
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
#ifndef EMBEDDED_LIBRARY
/* BUG#33029, if we are replicating from a buggy master, reset
auto_inc_intervals_forced to prevent substatement
(triggers/functions) from using erroneous INSERT_ID value
*/
if (rpl_master_erroneous_autoinc(this))
{
DBUG_ASSERT(backup->auto_inc_intervals_forced.nb_elements() == 0);
auto_inc_intervals_forced.swap(&backup->auto_inc_intervals_forced);
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
}
#endif
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
backup->option_bits= variables.option_bits;
backup->in_sub_stmt= in_sub_stmt;
backup->enable_slow_log= enable_slow_log;
backup->limit_found_rows= limit_found_rows;
backup->examined_row_count= examined_row_count;
backup->sent_row_count= sent_row_count;
backup->cuted_fields= cuted_fields;
backup->client_capabilities= client_capabilities;
backup->savepoints= transaction.savepoints;
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
backup->first_successful_insert_id_in_prev_stmt=
first_successful_insert_id_in_prev_stmt;
backup->first_successful_insert_id_in_cur_stmt=
first_successful_insert_id_in_cur_stmt;
if ((!lex->requires_prelocking() || is_update_query(lex->sql_command)) &&
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
!is_current_stmt_binlog_format_row())
{
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
variables.option_bits&= ~OPTION_BIN_LOG;
}
2007-02-26 20:06:10 +01:00
2010-01-12 13:07:09 +01:00
if ((backup->option_bits & OPTION_BIN_LOG) &&
is_update_query(lex->sql_command) &&
!is_current_stmt_binlog_format_row())
2007-02-26 20:06:10 +01:00
mysql_bin_log.start_union_events(this, this->query_id);
/* Disable result sets */
client_capabilities &= ~CLIENT_MULTI_RESULTS;
in_sub_stmt|= new_state;
examined_row_count= 0;
sent_row_count= 0;
cuted_fields= 0;
transaction.savepoints= 0;
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
first_successful_insert_id_in_cur_stmt= 0;
}
void THD::restore_sub_statement_state(Sub_statement_state *backup)
{
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
#ifndef EMBEDDED_LIBRARY
/* BUG#33029, if we are replicating from a buggy master, restore
auto_inc_intervals_forced so that the top statement can use the
INSERT_ID value set before this statement.
*/
if (rpl_master_erroneous_autoinc(this))
{
backup->auto_inc_intervals_forced.swap(&auto_inc_intervals_forced);
DBUG_ASSERT(backup->auto_inc_intervals_forced.nb_elements() == 0);
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
}
#endif
/*
To save resources we want to release savepoints which were created
during execution of function or trigger before leaving their savepoint
level. It is enough to release first savepoint set on this level since
all later savepoints will be released automatically.
*/
if (transaction.savepoints)
{
SAVEPOINT *sv;
for (sv= transaction.savepoints; sv->prev; sv= sv->prev)
{}
/* ha_release_savepoint() never returns error. */
(void)ha_release_savepoint(this, sv);
}
transaction.savepoints= backup->savepoints;
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
variables.option_bits= backup->option_bits;
in_sub_stmt= backup->in_sub_stmt;
enable_slow_log= backup->enable_slow_log;
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
first_successful_insert_id_in_prev_stmt=
backup->first_successful_insert_id_in_prev_stmt;
first_successful_insert_id_in_cur_stmt=
backup->first_successful_insert_id_in_cur_stmt;
limit_found_rows= backup->limit_found_rows;
sent_row_count= backup->sent_row_count;
client_capabilities= backup->client_capabilities;
/*
If we've left sub-statement mode, reset the fatal error flag.
Otherwise keep the current value, to propagate it up the sub-statement
stack.
*/
if (!in_sub_stmt)
is_fatal_sub_stmt_error= FALSE;
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
if ((variables.option_bits & OPTION_BIN_LOG) && is_update_query(lex->sql_command) &&
2010-01-12 13:07:09 +01:00
!is_current_stmt_binlog_format_row())
2007-02-26 20:06:10 +01:00
mysql_bin_log.stop_union_events(this);
2007-02-23 18:58:56 +01:00
/*
The following is added to the old values as we are interested in the
total complexity of the query
*/
examined_row_count+= backup->examined_row_count;
cuted_fields+= backup->cuted_fields;
}
void THD::set_statement(Statement *stmt)
{
mysql_mutex_lock(&LOCK_thd_data);
Statement::set_statement(stmt);
mysql_mutex_unlock(&LOCK_thd_data);
}
/** Assign a new value to thd->query. */
void THD::set_query(char *query_arg, uint32 query_length_arg)
{
mysql_mutex_lock(&LOCK_thd_data);
set_query_inner(query_arg, query_length_arg);
mysql_mutex_unlock(&LOCK_thd_data);
}
/** Assign a new value to thd->query and thd->query_id. */
void THD::set_query_and_id(char *query_arg, uint32 query_length_arg,
query_id_t new_query_id)
{
mysql_mutex_lock(&LOCK_thd_data);
set_query_inner(query_arg, query_length_arg);
query_id= new_query_id;
mysql_mutex_unlock(&LOCK_thd_data);
}
/** Assign a new value to thd->query_id. */
void THD::set_query_id(query_id_t new_query_id)
{
mysql_mutex_lock(&LOCK_thd_data);
query_id= new_query_id;
mysql_mutex_unlock(&LOCK_thd_data);
}
/**
Leave explicit LOCK TABLES or prelocked mode and restore value of
transaction sentinel in MDL subsystem.
*/
void THD::leave_locked_tables_mode()
{
locked_tables_mode= LTM_NONE;
/* Make sure we don't release the global read lock when leaving LTM. */
mdl_context.reset_trans_sentinel(global_read_lock.global_shared_lock());
/* Also ensure that we don't release metadata locks for open HANDLERs. */
if (handler_tables_hash.records)
mysql_ha_move_tickets_after_trans_sentinel(this);
}
/**
Mark transaction to rollback and mark error as fatal to a sub-statement.
@param thd Thread handle
@param all TRUE <=> rollback main transaction.
*/
void mark_transaction_to_rollback(THD *thd, bool all)
{
if (thd)
{
thd->is_fatal_sub_stmt_error= TRUE;
thd->transaction_rollback_request= all;
/*
Aborted transactions can not be IGNOREd.
Switch off the IGNORE flag for the current
SELECT_LEX. This should allow my_error()
to report the error and abort the execution
flow, even in presence
of IGNORE clause.
*/
if (thd->lex->current_select)
thd->lex->current_select->no_error= FALSE;
}
}
/***************************************************************************
Handling of XA id cacheing
***************************************************************************/
mysql_mutex_t LOCK_xid_cache;
HASH xid_cache;
extern "C" uchar *xid_get_hash_key(const uchar *, size_t *, my_bool);
extern "C" void xid_free_hash(void *);
uchar *xid_get_hash_key(const uchar *ptr, size_t *length,
my_bool not_used __attribute__((unused)))
{
*length=((XID_STATE*)ptr)->xid.key_length();
return ((XID_STATE*)ptr)->xid.key();
}
void xid_free_hash(void *ptr)
{
if (!((XID_STATE*)ptr)->in_thd)
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
my_free((uchar*)ptr, MYF(0));
}
#ifdef HAVE_PSI_INTERFACE
static PSI_mutex_key key_LOCK_xid_cache;
static PSI_mutex_info all_xid_mutexes[]=
{
{ &key_LOCK_xid_cache, "LOCK_xid_cache", PSI_FLAG_GLOBAL}
};
static void init_xid_psi_keys(void)
{
const char* category= "sql";
int count;
if (PSI_server == NULL)
return;
count= array_elements(all_xid_mutexes);
PSI_server->register_mutex(category, all_xid_mutexes, count);
}
#endif /* HAVE_PSI_INTERFACE */
bool xid_cache_init()
{
#ifdef HAVE_PSI_INTERFACE
init_xid_psi_keys();
#endif
mysql_mutex_init(key_LOCK_xid_cache, &LOCK_xid_cache, MY_MUTEX_INIT_FAST);
return my_hash_init(&xid_cache, &my_charset_bin, 100, 0, 0,
xid_get_hash_key, xid_free_hash, 0) != 0;
}
void xid_cache_free()
{
if (my_hash_inited(&xid_cache))
{
my_hash_free(&xid_cache);
mysql_mutex_destroy(&LOCK_xid_cache);
}
}
XID_STATE *xid_cache_search(XID *xid)
{
mysql_mutex_lock(&LOCK_xid_cache);
XID_STATE *res=(XID_STATE *)my_hash_search(&xid_cache, xid->key(),
xid->key_length());
mysql_mutex_unlock(&LOCK_xid_cache);
return res;
}
bool xid_cache_insert(XID *xid, enum xa_states xa_state)
{
XID_STATE *xs;
my_bool res;
mysql_mutex_lock(&LOCK_xid_cache);
if (my_hash_search(&xid_cache, xid->key(), xid->key_length()))
res=0;
else if (!(xs=(XID_STATE *)my_malloc(sizeof(*xs), MYF(MY_WME))))
res=1;
else
{
xs->xa_state=xa_state;
xs->xid.set(xid);
xs->in_thd=0;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
res=my_hash_insert(&xid_cache, (uchar*)xs);
}
mysql_mutex_unlock(&LOCK_xid_cache);
return res;
}
bool xid_cache_insert(XID_STATE *xid_state)
{
mysql_mutex_lock(&LOCK_xid_cache);
DBUG_ASSERT(my_hash_search(&xid_cache, xid_state->xid.key(),
xid_state->xid.key_length())==0);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
my_bool res=my_hash_insert(&xid_cache, (uchar*)xid_state);
mysql_mutex_unlock(&LOCK_xid_cache);
return res;
}
void xid_cache_delete(XID_STATE *xid_state)
{
mysql_mutex_lock(&LOCK_xid_cache);
my_hash_delete(&xid_cache, (uchar *)xid_state);
mysql_mutex_unlock(&LOCK_xid_cache);
}
/**
Decide on logging format to use for the statement and issue errors
or warnings as needed. The decision depends on the following
parameters:
- The logging mode, i.e., the value of binlog_format. Can be
statement, mixed, or row.
- The type of statement. There are three types of statements:
"normal" safe statements; unsafe statements; and row injections.
An unsafe statement is one that, if logged in statement format,
might produce different results when replayed on the slave (e.g.,
INSERT DELAYED). A row injection is either a BINLOG statement, or
a row event executed by the slave's SQL thread.
- The capabilities of tables modified by the statement. The
*capabilities vector* for a table is a set of flags associated
with the table. Currently, it only includes two flags: *row
capability flag* and *statement capability flag*.
The row capability flag is set if and only if the engine can
handle row-based logging. The statement capability flag is set if
and only if the table can handle statement-based logging.
Decision table for logging format
---------------------------------
The following table summarizes how the format and generated
warning/error depends on the tables' capabilities, the statement
type, and the current binlog_format.
Row capable N NNNNNNNNN YYYYYYYYY YYYYYYYYY
Statement capable N YYYYYYYYY NNNNNNNNN YYYYYYYYY
Statement type * SSSUUUIII SSSUUUIII SSSUUUIII
binlog_format * SMRSMRSMR SMRSMRSMR SMRSMRSMR
Logged format - SS-S----- -RR-RR-RR SRRSRR-RR
Warning/Error 1 --2732444 5--5--6-- ---7--6--
Legend
------
Row capable: N - Some table not row-capable, Y - All tables row-capable
Stmt capable: N - Some table not stmt-capable, Y - All tables stmt-capable
Statement type: (S)afe, (U)nsafe, or Row (I)njection
binlog_format: (S)TATEMENT, (M)IXED, or (R)OW
Logged format: (S)tatement or (R)ow
Warning/Error: Warnings and error messages are as follows:
1. Error: Cannot execute statement: binlogging impossible since both
row-incapable engines and statement-incapable engines are
involved.
2. Error: Cannot execute statement: binlogging impossible since
BINLOG_FORMAT = ROW and at least one table uses a storage engine
limited to statement-logging.
3. Error: Cannot execute statement: binlogging of unsafe statement
is impossible when storage engine is limited to statement-logging
and BINLOG_FORMAT = MIXED.
4. Error: Cannot execute row injection: binlogging impossible since
at least one table uses a storage engine limited to
statement-logging.
5. Error: Cannot execute statement: binlogging impossible since
BINLOG_FORMAT = STATEMENT and at least one table uses a storage
engine limited to row-logging.
6. Error: Cannot execute row injection: binlogging impossible since
BINLOG_FORMAT = STATEMENT.
7. Warning: Unsafe statement binlogged in statement format since
BINLOG_FORMAT = STATEMENT.
In addition, we can produce the following error (not depending on
the variables of the decision diagram):
8. Error: Cannot execute statement: binlogging impossible since more
than one engine is involved and at least one engine is
self-logging.
For each error case above, the statement is prevented from being
logged, we report an error, and roll back the statement. For
warnings, we set the thd->binlog_flags variable: the warning will be
printed only if the statement is successfully logged.
@see THD::binlog_query
@param[in] thd Client thread
@param[in] tables Tables involved in the query
@retval 0 No error; statement can be logged.
@retval -1 One of the error conditions above applies (1, 2, 4, 5, or 6).
*/
int THD::decide_logging_format(TABLE_LIST *tables)
{
DBUG_ENTER("THD::decide_logging_format");
DBUG_PRINT("info", ("query: %s", query()));
DBUG_PRINT("info", ("variables.binlog_format: %u",
variables.binlog_format));
DBUG_PRINT("info", ("lex->get_stmt_unsafe_flags(): 0x%x",
lex->get_stmt_unsafe_flags()));
/*
We should not decide logging format if the binlog is closed or
binlogging is off, or if the statement is filtered out from the
binlog by filtering rules.
*/
2010-01-12 13:07:09 +01:00
if (mysql_bin_log.is_open() && (variables.option_bits & OPTION_BIN_LOG) &&
!(variables.binlog_format == BINLOG_FORMAT_STMT &&
!binlog_filter->db_ok(db)))
{
/*
Compute one bit field with the union of all the engine
capabilities, and one with the intersection of all the engine
capabilities.
*/
handler::Table_flags flags_write_some_set= 0;
handler::Table_flags flags_some_set= 0;
handler::Table_flags flags_write_all_set=
HA_BINLOG_ROW_CAPABLE | HA_BINLOG_STMT_CAPABLE;
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
/*
If different types of engines are about to be updated.
For example: Innodb and Falcon; Innodb and MyIsam.
*/
my_bool multi_write_engine= FALSE;
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
/*
If different types of engines are about to be accessed
and any of them is about to be updated. For example:
Innodb and Falcon; Innodb and MyIsam.
*/
my_bool multi_access_engine= FALSE;
/*
If non-transactional and transactional engines are about
to be accessed and any of them is about to be updated.
For example: Innodb and MyIsam.
*/
my_bool trans_non_trans_access_engines= FALSE;
/*
If all engines that are about to be updated are
transactional.
*/
my_bool all_trans_write_engines= TRUE;
TABLE* prev_write_table= NULL;
TABLE* prev_access_table= NULL;
#ifndef DBUG_OFF
{
static const char *prelocked_mode_name[] = {
"NON_PRELOCKED",
"PRELOCKED",
"PRELOCKED_UNDER_LOCK_TABLES",
};
DBUG_PRINT("debug", ("prelocked_mode: %s",
2010-02-04 23:08:08 +01:00
prelocked_mode_name[locked_tables_mode]));
}
#endif
/*
Get the capabilities vector for all involved storage engines and
mask out the flags for the binary log.
*/
for (TABLE_LIST *table= tables; table; table= table->next_global)
{
if (table->placeholder())
continue;
if (table->table->s->table_category == TABLE_CATEGORY_PERFORMANCE)
lex->set_stmt_unsafe(LEX::BINLOG_STMT_UNSAFE_SYSTEM_TABLE);
handler::Table_flags const flags= table->table->file->ha_table_flags();
DBUG_PRINT("info", ("table: %s; ha_table_flags: 0x%llx",
table->table_name, flags));
if (table->lock_type >= TL_WRITE_ALLOW_WRITE)
{
if (prev_write_table && prev_write_table->file->ht !=
table->table->file->ht)
multi_write_engine= TRUE;
BUG#51894 Replication failure with SBR on DROP TEMPORARY TABLE inside a transaction BUG#52616 Temp table prevents switch binlog format from STATEMENT to ROW Before the WL#2687 and BUG#46364, every non-transactional change that happened after a transactional change was written to trx-cache and flushed upon committing the transaction. WL#2687 and BUG#46364 changed this behavior and non-transactional changes are now written to the binary log upon committing the statement. A binary log event is identified as transactional or non-transactional through a flag in the Log_event which is set taking into account the underlie storage engine on what it is stems from. In the current bug, this flag was not being set properly when the DROP TEMPORARY TABLE was executed. However, while fixing this bug we figured out that changes to temporary tables should be always written to the trx-cache if there is an on-going transaction. Otherwise, binlog events in the reversed order would be produced. Regarding concurrency, keeping changes to temporary tables in the trx-cache is also safe as temporary tables are only visible to the owner connection. In this patch, we classify the following statements as unsafe: 1 - INSERT INTO t_myisam SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_myisam 3 - CREATE TEMPORARY TABLE t_myisam_temp SELECT * FROM t_myisam On the other hand, the following statements are classified as safe: 1 - INSERT INTO t_innodb SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_innodb The patch also guarantees that transactions that have a DROP TEMPORARY are always written to the binary log regardless of the mode and the outcome: commit or rollback. In particular, the DROP TEMPORARY is extended with the IF EXISTS clause when the current statement logging format is set to row. Finally, the patch allows to switch from STATEMENT to MIXED/ROW when there are temporary tables but the contrary is not possible.
2010-04-20 11:10:43 +02:00
/*
Every temporary table must be always written to the binary
log in transaction boundaries and as such we artificially
classify them as transactional.
Indirectly, this avoids classifying a temporary table created
on a non-transactional engine as unsafe when it is modified
after any transactional table:
BEGIN;
INSERT INTO innodb_t VALUES (1);
INSERT INTO myisam_t_temp VALUES (1);
COMMIT;
BINARY LOG:
BEGIN;
INSERT INTO innodb_t VALUES (1);
INSERT INTO myisam_t_temp VALUES (1);
COMMIT;
*/
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
all_trans_write_engines= all_trans_write_engines &&
BUG#51894 Replication failure with SBR on DROP TEMPORARY TABLE inside a transaction BUG#52616 Temp table prevents switch binlog format from STATEMENT to ROW Before the WL#2687 and BUG#46364, every non-transactional change that happened after a transactional change was written to trx-cache and flushed upon committing the transaction. WL#2687 and BUG#46364 changed this behavior and non-transactional changes are now written to the binary log upon committing the statement. A binary log event is identified as transactional or non-transactional through a flag in the Log_event which is set taking into account the underlie storage engine on what it is stems from. In the current bug, this flag was not being set properly when the DROP TEMPORARY TABLE was executed. However, while fixing this bug we figured out that changes to temporary tables should be always written to the trx-cache if there is an on-going transaction. Otherwise, binlog events in the reversed order would be produced. Regarding concurrency, keeping changes to temporary tables in the trx-cache is also safe as temporary tables are only visible to the owner connection. In this patch, we classify the following statements as unsafe: 1 - INSERT INTO t_myisam SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_myisam 3 - CREATE TEMPORARY TABLE t_myisam_temp SELECT * FROM t_myisam On the other hand, the following statements are classified as safe: 1 - INSERT INTO t_innodb SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_innodb The patch also guarantees that transactions that have a DROP TEMPORARY are always written to the binary log regardless of the mode and the outcome: commit or rollback. In particular, the DROP TEMPORARY is extended with the IF EXISTS clause when the current statement logging format is set to row. Finally, the patch allows to switch from STATEMENT to MIXED/ROW when there are temporary tables but the contrary is not possible.
2010-04-20 11:10:43 +02:00
(table->table->file->has_transactions() ||
table->table->s->tmp_table);
prev_write_table= table->table;
flags_write_all_set &= flags;
flags_write_some_set |= flags;
}
flags_some_set |= flags;
BUG#51894 Replication failure with SBR on DROP TEMPORARY TABLE inside a transaction BUG#52616 Temp table prevents switch binlog format from STATEMENT to ROW Before the WL#2687 and BUG#46364, every non-transactional change that happened after a transactional change was written to trx-cache and flushed upon committing the transaction. WL#2687 and BUG#46364 changed this behavior and non-transactional changes are now written to the binary log upon committing the statement. A binary log event is identified as transactional or non-transactional through a flag in the Log_event which is set taking into account the underlie storage engine on what it is stems from. In the current bug, this flag was not being set properly when the DROP TEMPORARY TABLE was executed. However, while fixing this bug we figured out that changes to temporary tables should be always written to the trx-cache if there is an on-going transaction. Otherwise, binlog events in the reversed order would be produced. Regarding concurrency, keeping changes to temporary tables in the trx-cache is also safe as temporary tables are only visible to the owner connection. In this patch, we classify the following statements as unsafe: 1 - INSERT INTO t_myisam SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_myisam 3 - CREATE TEMPORARY TABLE t_myisam_temp SELECT * FROM t_myisam On the other hand, the following statements are classified as safe: 1 - INSERT INTO t_innodb SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_innodb The patch also guarantees that transactions that have a DROP TEMPORARY are always written to the binary log regardless of the mode and the outcome: commit or rollback. In particular, the DROP TEMPORARY is extended with the IF EXISTS clause when the current statement logging format is set to row. Finally, the patch allows to switch from STATEMENT to MIXED/ROW when there are temporary tables but the contrary is not possible.
2010-04-20 11:10:43 +02:00
/*
The mixture of non-transactional and transactional tables must
identified and classified as unsafe. However, a temporary table
must be always handled as a transactional table. Based on that,
we have the following statements classified as mixed and by
consequence as unsafe:
1: INSERT INTO myisam_t SELECT * FROM innodb_t;
2: INSERT INTO innodb_t SELECT * FROM myisam_t;
3: INSERT INTO myisam_t SELECT * FROM myisam_t_temp;
4: INSERT INTO myisam_t_temp SELECT * FROM myisam_t;
5: CREATE TEMPORARY TABLE myisam_t_temp SELECT * FROM mysiam_t;
The following statements are not considered mixed and as such
are safe:
1: INSERT INTO innodb_t SELECT * FROM myisam_t_temp;
2: INSERT INTO myisam_t_temp SELECT * FROM innodb_t_temp;
*/
if (!trans_non_trans_access_engines && prev_access_table &&
(lex->sql_command != SQLCOM_CREATE_TABLE ||
(lex->sql_command == SQLCOM_CREATE_TABLE &&
(lex->create_info.options & HA_LEX_CREATE_TMP_TABLE))))
{
BUG#51894 Replication failure with SBR on DROP TEMPORARY TABLE inside a transaction BUG#52616 Temp table prevents switch binlog format from STATEMENT to ROW Before the WL#2687 and BUG#46364, every non-transactional change that happened after a transactional change was written to trx-cache and flushed upon committing the transaction. WL#2687 and BUG#46364 changed this behavior and non-transactional changes are now written to the binary log upon committing the statement. A binary log event is identified as transactional or non-transactional through a flag in the Log_event which is set taking into account the underlie storage engine on what it is stems from. In the current bug, this flag was not being set properly when the DROP TEMPORARY TABLE was executed. However, while fixing this bug we figured out that changes to temporary tables should be always written to the trx-cache if there is an on-going transaction. Otherwise, binlog events in the reversed order would be produced. Regarding concurrency, keeping changes to temporary tables in the trx-cache is also safe as temporary tables are only visible to the owner connection. In this patch, we classify the following statements as unsafe: 1 - INSERT INTO t_myisam SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_myisam 3 - CREATE TEMPORARY TABLE t_myisam_temp SELECT * FROM t_myisam On the other hand, the following statements are classified as safe: 1 - INSERT INTO t_innodb SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_innodb The patch also guarantees that transactions that have a DROP TEMPORARY are always written to the binary log regardless of the mode and the outcome: commit or rollback. In particular, the DROP TEMPORARY is extended with the IF EXISTS clause when the current statement logging format is set to row. Finally, the patch allows to switch from STATEMENT to MIXED/ROW when there are temporary tables but the contrary is not possible.
2010-04-20 11:10:43 +02:00
my_bool prev_trans;
my_bool act_trans;
if (prev_access_table->s->tmp_table || table->table->s->tmp_table)
{
prev_trans= prev_access_table->s->tmp_table ? TRUE :
prev_access_table->file->has_transactions();
act_trans= table->table->s->tmp_table ? TRUE :
table->table->file->has_transactions();
}
else
{
prev_trans= prev_access_table->file->has_transactions();
act_trans= table->table->file->has_transactions();
}
trans_non_trans_access_engines= (prev_trans != act_trans);
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
multi_access_engine= TRUE;
}
thread_temporary_used |= table->table->s->tmp_table;
prev_access_table= table->table;
}
DBUG_PRINT("info", ("flags_write_all_set: 0x%llx", flags_write_all_set));
DBUG_PRINT("info", ("flags_write_some_set: 0x%llx", flags_write_some_set));
DBUG_PRINT("info", ("flags_some_set: 0x%llx", flags_some_set));
DBUG_PRINT("info", ("multi_write_engine: %d", multi_write_engine));
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
DBUG_PRINT("info", ("multi_access_engine: %d", multi_access_engine));
DBUG_PRINT("info", ("trans_non_trans_access_engines: %d",
trans_non_trans_access_engines));
int error= 0;
int unsafe_flags;
/*
Set the statement as unsafe if:
. it is a mixed statement, i.e. access transactional and non-transactional
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
tables, and update any of them;
or:
. an early statement updated a transactional table;
. and, the current statement updates a non-transactional table.
Any mixed statement is classified as unsafe to ensure that mixed mode is
completely safe. Consider the following example to understand why we
decided to do this:
Note that mixed statements such as
1: INSERT INTO myisam_t SELECT * FROM innodb_t;
2: INSERT INTO innodb_t SELECT * FROM myisam_t;
BUG#51894 Replication failure with SBR on DROP TEMPORARY TABLE inside a transaction BUG#52616 Temp table prevents switch binlog format from STATEMENT to ROW Before the WL#2687 and BUG#46364, every non-transactional change that happened after a transactional change was written to trx-cache and flushed upon committing the transaction. WL#2687 and BUG#46364 changed this behavior and non-transactional changes are now written to the binary log upon committing the statement. A binary log event is identified as transactional or non-transactional through a flag in the Log_event which is set taking into account the underlie storage engine on what it is stems from. In the current bug, this flag was not being set properly when the DROP TEMPORARY TABLE was executed. However, while fixing this bug we figured out that changes to temporary tables should be always written to the trx-cache if there is an on-going transaction. Otherwise, binlog events in the reversed order would be produced. Regarding concurrency, keeping changes to temporary tables in the trx-cache is also safe as temporary tables are only visible to the owner connection. In this patch, we classify the following statements as unsafe: 1 - INSERT INTO t_myisam SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_myisam 3 - CREATE TEMPORARY TABLE t_myisam_temp SELECT * FROM t_myisam On the other hand, the following statements are classified as safe: 1 - INSERT INTO t_innodb SELECT * FROM t_myisam_temp 2 - INSERT INTO t_myisam_temp SELECT * FROM t_innodb The patch also guarantees that transactions that have a DROP TEMPORARY are always written to the binary log regardless of the mode and the outcome: commit or rollback. In particular, the DROP TEMPORARY is extended with the IF EXISTS clause when the current statement logging format is set to row. Finally, the patch allows to switch from STATEMENT to MIXED/ROW when there are temporary tables but the contrary is not possible.
2010-04-20 11:10:43 +02:00
3: INSERT INTO myisam_t SELECT * FROM myisam_t_temp;
4: INSERT INTO myisam_t_temp SELECT * FROM myisam_t;
5: CREATE TEMPORARY TABLE myisam_t_temp SELECT * FROM mysiam_t;
are classified as unsafe to ensure that in mixed mode the execution is
completely safe and equivalent to the row mode. Consider the following
statements and sessions (connections) to understand the reason:
con1: INSERT INTO innodb_t VALUES (1);
con1: INSERT INTO innodb_t VALUES (100);
con1: BEGIN
con2: INSERT INTO myisam_t SELECT * FROM innodb_t;
con1: INSERT INTO innodb_t VALUES (200);
con1: COMMIT;
The point is that the concurrent statements may be written into the binary log
in a way different from the execution. For example,
BINARY LOG:
con2: BEGIN;
con2: INSERT INTO myisam_t SELECT * FROM innodb_t;
con2: COMMIT;
con1: BEGIN
con1: INSERT INTO innodb_t VALUES (200);
con1: COMMIT;
....
or
BINARY LOG:
con1: BEGIN
con1: INSERT INTO innodb_t VALUES (200);
con1: COMMIT;
con2: BEGIN;
con2: INSERT INTO myisam_t SELECT * FROM innodb_t;
con2: COMMIT;
Clearly, this may become a problem in STMT mode and setting the statement
as unsafe will make rows to be written into the binary log in MIXED mode
and as such the problem will not stand.
In STMT mode, although such statement is classified as unsafe, i.e.
INSERT INTO myisam_t SELECT * FROM innodb_t;
there is no enough information to avoid writing it outside the boundaries
of a transaction. This is not a problem if we are considering snapshot
isolation level but if we have pure repeatable read or serializable the
lock history on the slave will be different from the master.
*/
if (trans_non_trans_access_engines)
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
lex->set_stmt_unsafe(LEX::BINLOG_STMT_UNSAFE_MIXED_STATEMENT);
else if (trans_has_updated_trans_table(this) && !all_trans_write_engines)
lex->set_stmt_unsafe(LEX::BINLOG_STMT_UNSAFE_NONTRANS_AFTER_TRANS);
/*
If more than one engine is involved in the statement and at
least one is doing it's own logging (is *self-logging*), the
statement cannot be logged atomically, so we generate an error
rather than allowing the binlog to become corrupt.
*/
if (multi_write_engine &&
(flags_write_some_set & HA_HAS_OWN_BINLOGGING))
my_error((error= ER_BINLOG_MULTIPLE_ENGINES_AND_SELF_LOGGING_ENGINE),
MYF(0));
BUG#51291 Unfortunate effect around variable binlog_direct_non_transactional_updates BUG#46364 introduced the flag binlog_direct_non_transactional_updates which would make N-changes to be written to the binary log upon committing the statement when "ON". On the other hand, when "OFF" the option was supposed to mimic the behavior in 5.1. However, the implementation was not mimicking the behavior correctly and the following bugs popped up: Case #1: N-changes executed within a transaction would go into the S-cache. When later in the same transaction a T-change occurs, N-changes following it were written to the T-cache instead of the S-cache. In some cases, this raises problems. For example, a Table_map_log_event being written initially into the S-cache, together with the initial N-changes, would be absent from the T-cache. This would log N-changes orphaned from a Table_map_log_event (thence discarded at the slave). (MIXED and ROW) Case #2: When rolling back a transaction, the N-changes that might be in the T-cache were disregarded and truncated along with the T-changes. (MIXED and ROW) Case #3: When a MIXED statement (TN) is ahead of any other T-changes in the transaction and it fails, it is kept in the T-cache until the transaction ends. This is not the case in 5.1 or Betony (5.5.2). In these, the failed TN statement would be written to the binlog at the same instant it had failed and not deferred until transaction end. (SBR) To fix these problems, we have decided to do what follows: For Case #1 and #2, we circumvent them: 1. by not letting binlog_direct_non_transactional_updates affect MIXED and RBR. These modes will keep the behavior provided by WL#2687. Although this will make Celosia to behave differently from 5.1, an execution will be always safe under such modes in the sense that slaves will never go out sync. In 5.1, using either MIXED or ROW while mixing N-statements and T-statements was not safe. For Case #3, we don't actually fix it. We: 1. keep it and make all MIXED statements whether they end up failing or not or whether they are up front in the transaction or after some transactional change to always be stored in the T-cache. This means that it is written to the binary log on transaction commit/rollback only. 2. We make the warning message even more specific about the MIXED statement and SBR.
2010-03-31 15:22:47 +02:00
else if (multi_access_engine && flags_some_set & HA_HAS_OWN_BINLOGGING)
lex->set_stmt_unsafe(LEX::BINLOG_STMT_UNSAFE_MULTIPLE_ENGINES_AND_SELF_LOGGING_ENGINE);
/* both statement-only and row-only engines involved */
if ((flags_write_all_set & (HA_BINLOG_STMT_CAPABLE | HA_BINLOG_ROW_CAPABLE)) == 0)
{
/*
1. Error: Binary logging impossible since both row-incapable
engines and statement-incapable engines are involved
*/
my_error((error= ER_BINLOG_ROW_ENGINE_AND_STMT_ENGINE), MYF(0));
}
/* statement-only engines involved */
else if ((flags_write_all_set & HA_BINLOG_ROW_CAPABLE) == 0)
{
if (lex->is_stmt_row_injection())
{
/*
4. Error: Cannot execute row injection since table uses
storage engine limited to statement-logging
*/
my_error((error= ER_BINLOG_ROW_INJECTION_AND_STMT_ENGINE), MYF(0));
}
else if (variables.binlog_format == BINLOG_FORMAT_ROW)
{
/*
2. Error: Cannot modify table that uses a storage engine
limited to statement-logging when BINLOG_FORMAT = ROW
*/
my_error((error= ER_BINLOG_ROW_MODE_AND_STMT_ENGINE), MYF(0));
}
else if ((unsafe_flags= lex->get_stmt_unsafe_flags()) != 0)
{
/*
3. Error: Cannot execute statement: binlogging of unsafe
statement is impossible when storage engine is limited to
statement-logging and BINLOG_FORMAT = MIXED.
*/
for (int unsafe_type= 0;
unsafe_type < LEX::BINLOG_STMT_UNSAFE_COUNT;
unsafe_type++)
if (unsafe_flags & (1 << unsafe_type))
my_error((error= ER_BINLOG_UNSAFE_AND_STMT_ENGINE), MYF(0),
ER(LEX::binlog_stmt_unsafe_errcode[unsafe_type]));
}
/* log in statement format! */
}
/* no statement-only engines */
else
{
/* binlog_format = STATEMENT */
if (variables.binlog_format == BINLOG_FORMAT_STMT)
{
if (lex->is_stmt_row_injection())
{
/*
6. Error: Cannot execute row injection since
BINLOG_FORMAT = STATEMENT
*/
my_error((error= ER_BINLOG_ROW_INJECTION_AND_STMT_MODE), MYF(0));
}
else if ((flags_write_all_set & HA_BINLOG_STMT_CAPABLE) == 0)
{
/*
5. Error: Cannot modify table that uses a storage engine
limited to row-logging when binlog_format = STATEMENT
*/
my_error((error= ER_BINLOG_STMT_MODE_AND_ROW_ENGINE), MYF(0), "");
}
else if ((unsafe_flags= lex->get_stmt_unsafe_flags()) != 0)
{
/*
7. Warning: Unsafe statement logged as statement due to
binlog_format = STATEMENT
*/
binlog_unsafe_warning_flags|= unsafe_flags;
DBUG_PRINT("info", ("Scheduling warning to be issued by "
"binlog_query: '%s'",
ER(ER_BINLOG_UNSAFE_STATEMENT)));
DBUG_PRINT("info", ("binlog_unsafe_warning_flags: 0x%x",
binlog_unsafe_warning_flags));
}
/* log in statement format! */
}
/* No statement-only engines and binlog_format != STATEMENT.
I.e., nothing prevents us from row logging if needed. */
else
{
if (lex->is_stmt_unsafe() || lex->is_stmt_row_injection()
|| (flags_write_all_set & HA_BINLOG_STMT_CAPABLE) == 0)
{
/* log in row format! */
set_current_stmt_binlog_format_row_if_mixed();
}
}
}
if (error) {
DBUG_PRINT("info", ("decision: no logging since an error was generated"));
DBUG_RETURN(-1);
}
DBUG_PRINT("info", ("decision: logging in %s format",
is_current_stmt_binlog_format_row() ?
"ROW" : "STATEMENT"));
}
#ifndef DBUG_OFF
else
DBUG_PRINT("info", ("decision: no logging since "
"mysql_bin_log.is_open() = %d "
"and (options & OPTION_BIN_LOG) = 0x%llx "
"and binlog_format = %u "
"and binlog_filter->db_ok(db) = %d",
mysql_bin_log.is_open(),
2010-01-12 13:07:09 +01:00
(variables.option_bits & OPTION_BIN_LOG),
variables.binlog_format,
binlog_filter->db_ok(db)));
#endif
DBUG_RETURN(0);
}
/*
Implementation of interface to write rows to the binary log through the
thread. The thread is responsible for writing the rows it has
inserted/updated/deleted.
*/
#ifndef MYSQL_CLIENT
/*
Template member function for ensuring that there is an rows log
event of the apropriate type before proceeding.
PRE CONDITION:
- Events of type 'RowEventT' have the type code 'type_code'.
POST CONDITION:
If a non-NULL pointer is returned, the pending event for thread 'thd' will
be an event of type 'RowEventT' (which have the type code 'type_code')
will either empty or have enough space to hold 'needed' bytes. In
addition, the columns bitmap will be correct for the row, meaning that
the pending event will be flushed if the columns in the event differ from
the columns suppled to the function.
RETURNS
If no error, a non-NULL pending event (either one which already existed or
the newly created one).
If error, NULL.
*/
template <class RowsEventT> Rows_log_event*
THD::binlog_prepare_pending_rows_event(TABLE* table, uint32 serv_id,
MY_BITMAP const* cols,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t colcnt,
size_t needed,
2006-01-09 15:59:39 +01:00
bool is_transactional,
RowsEventT *hint __attribute__((unused)))
{
DBUG_ENTER("binlog_prepare_pending_rows_event");
/* Pre-conditions */
DBUG_ASSERT(table->s->table_map_id != ~0UL);
/* Fetch the type code for the RowsEventT template parameter */
int const type_code= RowsEventT::TYPE_CODE;
/*
There is no good place to set up the transactional data, so we
have to do it here.
*/
if (binlog_setup_trx_data())
DBUG_RETURN(NULL);
Rows_log_event* pending= binlog_get_pending_rows_event(is_transactional);
if (unlikely(pending && !pending->is_valid()))
DBUG_RETURN(NULL);
/*
Check if the current event is non-NULL and a write-rows
event. Also check if the table provided is mapped: if it is not,
then we have switched to writing to a new table.
If there is no pending event, we need to create one. If there is a pending
event, but it's not about the same table id, or not of the same type
(between Write, Update and Delete), or not the same affected columns, or
going to be too big, flush this event to disk and create a new pending
event.
*/
if (!pending ||
pending->server_id != serv_id ||
pending->get_table_id() != table->s->table_map_id ||
pending->get_type_code() != type_code ||
pending->get_data_size() + needed > opt_binlog_rows_event_max_size ||
pending->get_width() != colcnt ||
!bitmap_cmp(pending->get_cols(), cols))
{
/* Create a new RowsEventT... */
Rows_log_event* const
ev= new RowsEventT(this, table, table->s->table_map_id, cols,
is_transactional);
if (unlikely(!ev))
DBUG_RETURN(NULL);
ev->server_id= serv_id; // I don't like this, it's too easy to forget.
/*
flush the pending event and replace it with the newly created
event...
*/
if (unlikely(
mysql_bin_log.flush_and_set_pending_rows_event(this, ev,
is_transactional)))
{
delete ev;
DBUG_RETURN(NULL);
}
DBUG_RETURN(ev); /* This is the new pending event */
}
DBUG_RETURN(pending); /* This is the current pending event */
}
2006-02-02 15:22:31 +01:00
#ifdef HAVE_EXPLICIT_TEMPLATE_INSTANTIATION
/*
Instantiate the versions we need, we have -fno-implicit-template as
compiling option.
*/
template Rows_log_event*
2006-01-09 15:59:39 +01:00
THD::binlog_prepare_pending_rows_event(TABLE*, uint32, MY_BITMAP const*,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t, size_t, bool,
2006-01-09 15:59:39 +01:00
Write_rows_log_event*);
template Rows_log_event*
2006-01-09 15:59:39 +01:00
THD::binlog_prepare_pending_rows_event(TABLE*, uint32, MY_BITMAP const*,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t colcnt, size_t, bool,
2006-01-09 15:59:39 +01:00
Delete_rows_log_event *);
template Rows_log_event*
2006-01-09 15:59:39 +01:00
THD::binlog_prepare_pending_rows_event(TABLE*, uint32, MY_BITMAP const*,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t colcnt, size_t, bool,
2006-01-09 15:59:39 +01:00
Update_rows_log_event *);
2006-02-02 15:22:31 +01:00
#endif
#ifdef NOT_USED
static char const*
field_type_name(enum_field_types type)
{
switch (type) {
case MYSQL_TYPE_DECIMAL:
return "MYSQL_TYPE_DECIMAL";
case MYSQL_TYPE_TINY:
return "MYSQL_TYPE_TINY";
case MYSQL_TYPE_SHORT:
return "MYSQL_TYPE_SHORT";
case MYSQL_TYPE_LONG:
return "MYSQL_TYPE_LONG";
case MYSQL_TYPE_FLOAT:
return "MYSQL_TYPE_FLOAT";
case MYSQL_TYPE_DOUBLE:
return "MYSQL_TYPE_DOUBLE";
case MYSQL_TYPE_NULL:
return "MYSQL_TYPE_NULL";
case MYSQL_TYPE_TIMESTAMP:
return "MYSQL_TYPE_TIMESTAMP";
case MYSQL_TYPE_LONGLONG:
return "MYSQL_TYPE_LONGLONG";
case MYSQL_TYPE_INT24:
return "MYSQL_TYPE_INT24";
case MYSQL_TYPE_DATE:
return "MYSQL_TYPE_DATE";
case MYSQL_TYPE_TIME:
return "MYSQL_TYPE_TIME";
case MYSQL_TYPE_DATETIME:
return "MYSQL_TYPE_DATETIME";
case MYSQL_TYPE_YEAR:
return "MYSQL_TYPE_YEAR";
case MYSQL_TYPE_NEWDATE:
return "MYSQL_TYPE_NEWDATE";
case MYSQL_TYPE_VARCHAR:
return "MYSQL_TYPE_VARCHAR";
case MYSQL_TYPE_BIT:
return "MYSQL_TYPE_BIT";
case MYSQL_TYPE_NEWDECIMAL:
return "MYSQL_TYPE_NEWDECIMAL";
case MYSQL_TYPE_ENUM:
return "MYSQL_TYPE_ENUM";
case MYSQL_TYPE_SET:
return "MYSQL_TYPE_SET";
case MYSQL_TYPE_TINY_BLOB:
return "MYSQL_TYPE_TINY_BLOB";
case MYSQL_TYPE_MEDIUM_BLOB:
return "MYSQL_TYPE_MEDIUM_BLOB";
case MYSQL_TYPE_LONG_BLOB:
return "MYSQL_TYPE_LONG_BLOB";
case MYSQL_TYPE_BLOB:
return "MYSQL_TYPE_BLOB";
case MYSQL_TYPE_VAR_STRING:
return "MYSQL_TYPE_VAR_STRING";
case MYSQL_TYPE_STRING:
return "MYSQL_TYPE_STRING";
case MYSQL_TYPE_GEOMETRY:
return "MYSQL_TYPE_GEOMETRY";
}
return "Unknown";
}
#endif
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
namespace {
/**
Class to handle temporary allocation of memory for row data.
The responsibilities of the class is to provide memory for
packing one or two rows of packed data (depending on what
constructor is called).
In order to make the allocation more efficient for "simple" rows,
i.e., rows that do not contain any blobs, a pointer to the
allocated memory is of memory is stored in the table structure
for simple rows. If memory for a table containing a blob field
is requested, only memory for that is allocated, and subsequently
released when the object is destroyed.
*/
class Row_data_memory {
public:
/**
Build an object to keep track of a block-local piece of memory
for storing a row of data.
@param table
Table where the pre-allocated memory is stored.
@param length
Length of data that is needed, if the record contain blobs.
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
Row_data_memory(TABLE *table, size_t const len1)
: m_memory(0)
{
#ifndef DBUG_OFF
m_alloc_checked= FALSE;
#endif
allocate_memory(table, len1);
m_ptr[0]= has_memory() ? m_memory : 0;
m_ptr[1]= 0;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
Row_data_memory(TABLE *table, size_t const len1, size_t const len2)
: m_memory(0)
{
#ifndef DBUG_OFF
m_alloc_checked= FALSE;
#endif
allocate_memory(table, len1 + len2);
m_ptr[0]= has_memory() ? m_memory : 0;
m_ptr[1]= has_memory() ? m_memory + len1 : 0;
}
~Row_data_memory()
{
if (m_memory != 0 && m_release_memory_on_destruction)
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
my_free((uchar*) m_memory, MYF(MY_WME));
}
/**
Is there memory allocated?
@retval true There is memory allocated
@retval false Memory allocation failed
*/
bool has_memory() const {
#ifndef DBUG_OFF
m_alloc_checked= TRUE;
#endif
return m_memory != 0;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *slot(uint s)
{
DBUG_ASSERT(s < sizeof(m_ptr)/sizeof(*m_ptr));
DBUG_ASSERT(m_ptr[s] != 0);
DBUG_ASSERT(m_alloc_checked == TRUE);
return m_ptr[s];
}
private:
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
void allocate_memory(TABLE *const table, size_t const total_length)
{
if (table->s->blob_fields == 0)
{
/*
The maximum length of a packed record is less than this
length. We use this value instead of the supplied length
when allocating memory for records, since we don't know how
the memory will be used in future allocations.
Since table->s->reclength is for unpacked records, we have
to add two bytes for each field, which can potentially be
added to hold the length of a packed field.
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const maxlen= table->s->reclength + 2 * table->s->fields;
/*
Allocate memory for two records if memory hasn't been
allocated. We allocate memory for two records so that it can
be used when processing update rows as well.
*/
if (table->write_row_record == 0)
table->write_row_record=
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
(uchar *) alloc_root(&table->mem_root, 2 * maxlen);
m_memory= table->write_row_record;
m_release_memory_on_destruction= FALSE;
}
else
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
m_memory= (uchar *) my_malloc(total_length, MYF(MY_WME));
m_release_memory_on_destruction= TRUE;
}
}
#ifndef DBUG_OFF
mutable bool m_alloc_checked;
#endif
bool m_release_memory_on_destruction;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *m_memory;
uchar *m_ptr[2];
};
}
int THD::binlog_write_row(TABLE* table, bool is_trans,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
MY_BITMAP const* cols, size_t colcnt,
uchar const *record)
{
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
DBUG_ASSERT(is_current_stmt_binlog_format_row() && mysql_bin_log.is_open());
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
/*
Pack records into format for transfer. We are allocating more
memory than needed, but that doesn't matter.
*/
Row_data_memory memory(table, max_row_length(table, record));
if (!memory.has_memory())
return HA_ERR_OUT_OF_MEM;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *row_data= memory.slot(0);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const len= pack_row(table, cols, row_data, record);
Rows_log_event* const ev=
binlog_prepare_pending_rows_event(table, server_id, cols, colcnt,
len, is_trans,
static_cast<Write_rows_log_event*>(0));
if (unlikely(ev == 0))
return HA_ERR_OUT_OF_MEM;
return ev->add_row_data(row_data, len);
}
int THD::binlog_update_row(TABLE* table, bool is_trans,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
MY_BITMAP const* cols, size_t colcnt,
const uchar *before_record,
const uchar *after_record)
{
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
DBUG_ASSERT(is_current_stmt_binlog_format_row() && mysql_bin_log.is_open());
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const before_maxlen = max_row_length(table, before_record);
size_t const after_maxlen = max_row_length(table, after_record);
Row_data_memory row_data(table, before_maxlen, after_maxlen);
if (!row_data.has_memory())
return HA_ERR_OUT_OF_MEM;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *before_row= row_data.slot(0);
uchar *after_row= row_data.slot(1);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const before_size= pack_row(table, cols, before_row,
before_record);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const after_size= pack_row(table, cols, after_row,
after_record);
/*
Don't print debug messages when running valgrind since they can
trigger false warnings.
*/
#ifndef HAVE_purify
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
DBUG_DUMP("before_record", before_record, table->s->reclength);
DBUG_DUMP("after_record", after_record, table->s->reclength);
DBUG_DUMP("before_row", before_row, before_size);
DBUG_DUMP("after_row", after_row, after_size);
#endif
2006-01-09 15:59:39 +01:00
Rows_log_event* const ev=
binlog_prepare_pending_rows_event(table, server_id, cols, colcnt,
before_size + after_size, is_trans,
static_cast<Update_rows_log_event*>(0));
if (unlikely(ev == 0))
return HA_ERR_OUT_OF_MEM;
return
ev->add_row_data(before_row, before_size) ||
ev->add_row_data(after_row, after_size);
}
int THD::binlog_delete_row(TABLE* table, bool is_trans,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
MY_BITMAP const* cols, size_t colcnt,
uchar const *record)
{
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
DBUG_ASSERT(is_current_stmt_binlog_format_row() && mysql_bin_log.is_open());
/*
Pack records into format for transfer. We are allocating more
memory than needed, but that doesn't matter.
*/
Row_data_memory memory(table, max_row_length(table, record));
if (unlikely(!memory.has_memory()))
return HA_ERR_OUT_OF_MEM;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *row_data= memory.slot(0);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
size_t const len= pack_row(table, cols, row_data, record);
2006-01-09 15:59:39 +01:00
Rows_log_event* const ev=
binlog_prepare_pending_rows_event(table, server_id, cols, colcnt,
len, is_trans,
static_cast<Delete_rows_log_event*>(0));
if (unlikely(ev == 0))
return HA_ERR_OUT_OF_MEM;
return ev->add_row_data(row_data, len);
}
int THD::binlog_remove_pending_rows_event(bool clear_maps,
bool is_transactional)
{
DBUG_ENTER("THD::binlog_remove_pending_rows_event");
if (!mysql_bin_log.is_open())
DBUG_RETURN(0);
mysql_bin_log.remove_pending_rows_event(this, is_transactional);
if (clear_maps)
binlog_table_maps= 0;
DBUG_RETURN(0);
}
int THD::binlog_flush_pending_rows_event(bool stmt_end, bool is_transactional)
{
DBUG_ENTER("THD::binlog_flush_pending_rows_event");
/*
We shall flush the pending event even if we are not in row-based
mode: it might be the case that we left row-based mode before
flushing anything (e.g., if we have explicitly locked tables).
*/
if (!mysql_bin_log.is_open())
DBUG_RETURN(0);
/*
Mark the event as the last event of a statement if the stmt_end
flag is set.
*/
int error= 0;
if (Rows_log_event *pending= binlog_get_pending_rows_event(is_transactional))
{
if (stmt_end)
{
pending->set_flags(Rows_log_event::STMT_END_F);
pending->flags|= LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F;
binlog_table_maps= 0;
}
error= mysql_bin_log.flush_and_set_pending_rows_event(this, 0,
is_transactional);
}
DBUG_RETURN(error);
}
#if !defined(DBUG_OFF) && !defined(_lint)
static const char *
show_query_type(THD::enum_binlog_query_type qtype)
{
switch (qtype) {
case THD::ROW_QUERY_TYPE:
return "ROW";
case THD::STMT_QUERY_TYPE:
return "STMT";
case THD::QUERY_TYPE_COUNT:
default:
DBUG_ASSERT(0 <= qtype && qtype < THD::QUERY_TYPE_COUNT);
}
static char buf[64];
sprintf(buf, "UNKNOWN#%d", qtype);
return buf;
}
#endif
/**
Auxiliary method used by @c binlog_query() to raise warnings.
The type of warning and the type of unsafeness is stored in
THD::binlog_unsafe_warning_flags.
*/
void THD::issue_unsafe_warnings()
{
DBUG_ENTER("issue_unsafe_warnings");
/*
Ensure that binlog_unsafe_warning_flags is big enough to hold all
bits. This is actually a constant expression.
*/
DBUG_ASSERT(2 * LEX::BINLOG_STMT_UNSAFE_COUNT <=
sizeof(binlog_unsafe_warning_flags) * CHAR_BIT);
uint32 unsafe_type_flags= binlog_unsafe_warning_flags;
/*
Clear: (1) bits above BINLOG_STMT_UNSAFE_COUNT; (2) bits for
warnings that have been printed already.
*/
unsafe_type_flags &= (LEX::BINLOG_STMT_UNSAFE_ALL_FLAGS ^
(unsafe_type_flags >> LEX::BINLOG_STMT_UNSAFE_COUNT));
/* If all warnings have been printed already, return. */
if (unsafe_type_flags == 0)
DBUG_VOID_RETURN;
DBUG_PRINT("info", ("unsafe_type_flags: 0x%x", unsafe_type_flags));
/*
For each unsafe_type, check if the statement is unsafe in this way
and issue a warning.
*/
for (int unsafe_type=0;
unsafe_type < LEX::BINLOG_STMT_UNSAFE_COUNT;
unsafe_type++)
{
if ((unsafe_type_flags & (1 << unsafe_type)) != 0)
{
push_warning_printf(this, MYSQL_ERROR::WARN_LEVEL_NOTE,
ER_BINLOG_UNSAFE_STATEMENT,
ER(ER_BINLOG_UNSAFE_STATEMENT),
ER(LEX::binlog_stmt_unsafe_errcode[unsafe_type]));
if (global_system_variables.log_warnings)
{
char buf[MYSQL_ERRMSG_SIZE * 2];
sprintf(buf, ER(ER_BINLOG_UNSAFE_STATEMENT),
ER(LEX::binlog_stmt_unsafe_errcode[unsafe_type]));
sql_print_warning(ER(ER_MESSAGE_AND_STATEMENT), buf, query());
}
}
}
/*
Mark these unsafe types as already printed, to avoid printing
warnings for them again.
*/
binlog_unsafe_warning_flags|=
unsafe_type_flags << LEX::BINLOG_STMT_UNSAFE_COUNT;
DBUG_VOID_RETURN;
}
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
/**
Log the current query.
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
The query will be logged in either row format or statement format
depending on the value of @c current_stmt_binlog_format_row field and
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
the value of the @c qtype parameter.
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
This function must be called:
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
- After the all calls to ha_*_row() functions have been issued.
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
- After any writes to system tables. Rationale: if system tables
were written after a call to this function, and the master crashes
after the call to this function and before writing the system
tables, then the master and slave get out of sync.
- Before tables are unlocked and closed.
@see decide_logging_format
@retval 0 Success
@retval nonzero If there is a failure when writing the query (e.g.,
write failure), then the error code is returned.
*/
int THD::binlog_query(THD::enum_binlog_query_type qtype, char const *query_arg,
ulong query_len, bool is_trans, bool direct,
bool suppress_use, int errcode)
{
DBUG_ENTER("THD::binlog_query");
DBUG_PRINT("enter", ("qtype: %s query: '%s'",
show_query_type(qtype), query_arg));
DBUG_ASSERT(query_arg && mysql_bin_log.is_open());
/*
If we are not in prelocked mode, mysql_unlock_tables() will be
called after this binlog_query(), so we have to flush the pending
rows event with the STMT_END_F set to unlock all tables at the
slave side as well.
If we are in prelocked mode, the flushing will be done inside the
top-most close_thread_tables().
*/
if (this->locked_tables_mode <= LTM_LOCK_TABLES)
if (int error= binlog_flush_pending_rows_event(TRUE, is_trans))
DBUG_RETURN(error);
/*
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
Warnings for unsafe statements logged in statement format are
printed here instead of in decide_logging_format(). This is
because the warnings should be printed only if the statement is
actually logged. When executing decide_logging_format(), we cannot
know for sure if the statement will be logged.
*/
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
if (sql_log_bin_toplevel)
issue_unsafe_warnings();
switch (qtype) {
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
/*
ROW_QUERY_TYPE means that the statement may be logged either in
row format or in statement format. If
current_stmt_binlog_format is row, it means that the
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
statement has already been logged in row format and hence shall
not be logged again.
*/
case THD::ROW_QUERY_TYPE:
DBUG_PRINT("debug",
("is_current_stmt_binlog_format_row: %d",
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
is_current_stmt_binlog_format_row()));
if (is_current_stmt_binlog_format_row())
DBUG_RETURN(0);
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
/* Fall through */
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
/*
STMT_QUERY_TYPE means that the query must be logged in statement
format; it cannot be logged in row format. This is typically
used by DDL statements. It is an error to use this query type
if current_stmt_binlog_format_row is row.
@todo Currently there are places that call this method with
STMT_QUERY_TYPE and current_stmt_binlog_format is row. Fix those
places and add assert to ensure correct behavior. /Sven
*/
case THD::STMT_QUERY_TYPE:
/*
The MYSQL_LOG::write() function will set the STMT_END_F flag and
flush the pending rows event if necessary.
BUG#39934: Slave stops for engine that only support row-based logging General overview: The logic for switching to row format when binlog_format=MIXED had numerous flaws. The underlying problem was the lack of a consistent architecture. General purpose of this changeset: This changeset introduces an architecture for switching to row format when binlog_format=MIXED. It enforces the architecture where it has to. It leaves some bugs to be fixed later. It adds extensive tests to verify that unsafe statements work as expected and that appropriate errors are produced by problems with the selection of binlog format. It was not practical to split this into smaller pieces of work. Problem 1: To determine the logging mode, the code has to take several parameters into account (namely: (1) the value of binlog_format; (2) the capabilities of the engines; (3) the type of the current statement: normal, unsafe, or row injection). These parameters may conflict in several ways, namely: - binlog_format=STATEMENT for a row injection - binlog_format=STATEMENT for an unsafe statement - binlog_format=STATEMENT for an engine only supporting row logging - binlog_format=ROW for an engine only supporting statement logging - statement is unsafe and engine does not support row logging - row injection in a table that does not support statement logging - statement modifies one table that does not support row logging and one that does not support statement logging Several of these conflicts were not detected, or were detected with an inappropriate error message. The problem of BUG#39934 was that no appropriate error message was written for the case when an engine only supporting row logging executed a row injection with binlog_format=ROW. However, all above cases must be handled. Fix 1: Introduce new error codes (sql/share/errmsg.txt). Ensure that all conditions are detected and handled in decide_logging_format() Problem 2: The binlog format shall be determined once per statement, in decide_logging_format(). It shall not be changed before or after that. Before decide_logging_format() is called, all information necessary to determine the logging format must be available. This principle ensures that all unsafe statements are handled in a consistent way. However, this principle is not followed: thd->set_current_stmt_binlog_row_based_if_mixed() is called in several places, including from code executing UPDATE..LIMIT, INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and SET @@binlog_format. After Problem 1 was fixed, that caused inconsistencies where these unsafe statements would not print the appropriate warnings or errors for some of the conflicts. Fix 2: Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from code executed after decide_logging_format(). Compensate by calling the set_current_stmt_unsafe() at parse time. This way, all unsafe statements are detected by decide_logging_format(). Problem 3: INSERT DELAYED is not unsafe: it is logged in statement format even if binlog_format=MIXED, and no warning is printed even if binlog_format=STATEMENT. This is BUG#45825. Fix 3: Made INSERT DELAYED set itself to unsafe at parse time. This allows decide_logging_format() to detect that a warning should be printed or the binlog_format changed. Problem 4: LIMIT clause were not marked as unsafe when executed inside stored functions/triggers/views/prepared statements. This is BUG#45785. Fix 4: Make statements containing the LIMIT clause marked as unsafe at parse time, instead of at execution time. This allows propagating unsafe-ness to the view.
2009-07-14 21:31:19 +02:00
*/
{
Query_log_event qinfo(this, query_arg, query_len, is_trans, direct,
suppress_use, errcode);
qinfo.flags|= LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F;
/*
Binlog table maps will be irrelevant after a Query_log_event
(they are just removed on the slave side) so after the query
log event is written to the binary log, we pretend that no
table maps were written.
*/
int error= mysql_bin_log.write(&qinfo);
binlog_table_maps= 0;
DBUG_RETURN(error);
}
break;
case THD::QUERY_TYPE_COUNT:
default:
DBUG_ASSERT(0 <= qtype && qtype < QUERY_TYPE_COUNT);
}
DBUG_RETURN(0);
}
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
bool Discrete_intervals_list::append(ulonglong start, ulonglong val,
ulonglong incr)
{
DBUG_ENTER("Discrete_intervals_list::append");
/* first, see if this can be merged with previous */
if ((head == NULL) || tail->merge_if_contiguous(start, val, incr))
{
/* it cannot, so need to add a new interval */
Discrete_interval *new_interval= new Discrete_interval(start, val, incr);
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
DBUG_RETURN(append(new_interval));
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
}
DBUG_RETURN(0);
}
BUG#33029 5.0 to 5.1 replication fails on dup key when inserting using a trig in SP For all 5.0 and up to 5.1.12 exclusive, when a stored routine or trigger caused an INSERT into an AUTO_INCREMENT column, the generated AUTO_INCREMENT value should not be written into the binary log, which means if a statement does not generate AUTO_INCREMENT value itself, there will be no Intvar event (SET INSERT_ID) associated with it even if one of the stored routine or trigger caused generation of such a value. And meanwhile, when executing a stored routine or trigger, it would ignore the INSERT_ID value even if there is a INSERT_ID value available set by a SET INSERT_ID statement. Starting from MySQL 5.1.12, the generated AUTO_INCREMENT value is written into the binary log, and the value will be used if available when executing the stored routine or trigger. Prior fix of this bug in MySQL 5.0 and prior MySQL 5.1.12 (referenced as the buggy versions in the text below), when a statement that generates AUTO_INCREMENT value by the top statement was executed in the body of a SP, all statements in the SP after this statement would be treated as if they had generated AUTO_INCREMENT by the top statement. When a statement that did not generate AUTO_INCREMENT value by the top statement but by a function/trigger called by it, an erroneous Intvar event would be associated with the statement, this erroneous INSERT_ID value wouldn't cause problem when replicating between masters and slaves of 5.0.x or prior 5.1.12, because the erroneous INSERT_ID value was not used when executing functions/triggers. But when replicating from buggy versions to 5.1.12 or newer, which will use the INSERT_ID value in functions/triggers, the erroneous value will be used, which would cause duplicate entry error and cause the slave to stop. The patch for 5.1 fixed it to ignore the SET INSERT_ID value when executing functions/triggers if it is replicating from a master of buggy versions, another patch for 5.0 fixed it not to generate the erroneous Intvar event.
2008-03-14 04:35:41 +01:00
bool Discrete_intervals_list::append(Discrete_interval *new_interval)
{
DBUG_ENTER("Discrete_intervals_list::append");
if (unlikely(new_interval == NULL))
DBUG_RETURN(1);
DBUG_PRINT("info",("adding new auto_increment interval"));
if (head == NULL)
head= current= new_interval;
else
tail->next= new_interval;
tail= new_interval;
elements++;
DBUG_RETURN(0);
}
#endif /* !defined(MYSQL_CLIENT) */