mirror of
https://github.com/MariaDB/server.git
synced 2025-01-27 09:14:17 +01:00
bf0aa99aeb
A few different incorrect function type UBSAN issues have been grouped into this patch. The only real potentially undefined behavior is an error about show_func_mutex_instances_lost, which when invoked in sql_show.cc::show_status_array(), puts 5 arguments onto the stack; however, the implementing function only actually has 3 parameters (so only 3 would be popped). This was fixed by adding in the remaining parameters to satisfy the type mysql_show_var_func. The rest of the findings are pointer type mismatches that wouldn't lead to actual undefined behavior. The lf_hash_initializer function type definition is typedef void (*lf_hash_initializer)(LF_HASH *hash, void *dst, const void *src); but the MDL_lock and table cache's implementations of this function do not have that signature. The MDL_lock has specific MDL object parameters: static void lf_hash_initializer(LF_HASH *hash __attribute__((unused)), MDL_lock *lock, MDL_key *key_arg) and the table cache has specific TDC parameters: static void tdc_hash_initializer(LF_HASH *, TDC_element *element, LEX_STRING *key) leading to UBSAN runtime errors when invoking these functions. This patch fixes these type mis-matches by changing the implementing functions to use void * and const void * for their respective parameters, and later casting them to their expected type in the function body. Note too the functions tdc_hash_key and tc_purge_callback had a similar problem to tdc_hash_initializer and was fixed similarly. Reviewed By: ============ Sergei Golubchik <serg@mariadb.com>
545 lines
16 KiB
C++
545 lines
16 KiB
C++
/* Copyright (c) 2008, 2023, Oracle and/or its affiliates.
|
|
|
|
This program is free software; you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License, version 2.0,
|
|
as published by the Free Software Foundation.
|
|
|
|
This program is also distributed with certain software (including
|
|
but not limited to OpenSSL) that is licensed under separate terms,
|
|
as designated in a particular file or component or in included license
|
|
documentation. The authors of MySQL hereby grant you an additional
|
|
permission to link the program and your derivative works with the
|
|
separately licensed software that they have included with MySQL.
|
|
|
|
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, version 2.0, for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, write to the Free Software Foundation,
|
|
51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA */
|
|
|
|
/**
|
|
@file storage/perfschema/ha_perfschema.cc
|
|
Performance schema storage engine (implementation).
|
|
*/
|
|
|
|
#include "sql_plugin.h"
|
|
#include "my_pthread.h"
|
|
#include "ha_perfschema.h"
|
|
#include "pfs_engine_table.h"
|
|
#include "pfs_column_values.h"
|
|
#include "pfs_instr_class.h"
|
|
#include "pfs_instr.h"
|
|
#include "pfs_account.h"
|
|
#include "pfs_host.h"
|
|
#include "pfs_user.h"
|
|
#include "pfs_program.h"
|
|
#include "pfs_prepared_stmt.h"
|
|
#include "pfs_buffer_container.h"
|
|
|
|
handlerton *pfs_hton= NULL;
|
|
|
|
#define PFS_ENABLED() (pfs_initialized && (pfs_enabled || m_table_share->m_perpetual))
|
|
|
|
static handler* pfs_create_handler(handlerton *hton,
|
|
TABLE_SHARE *table,
|
|
MEM_ROOT *mem_root)
|
|
{
|
|
return new (mem_root) ha_perfschema(hton, table);
|
|
}
|
|
|
|
static int compare_database_names(const char *name1, const char *name2)
|
|
{
|
|
if (lower_case_table_names)
|
|
return strcasecmp(name1, name2);
|
|
return strcmp(name1, name2);
|
|
}
|
|
|
|
static const PFS_engine_table_share*
|
|
find_table_share(const char *db, const char *name)
|
|
{
|
|
DBUG_ENTER("find_table_share");
|
|
|
|
if (compare_database_names(db, PERFORMANCE_SCHEMA_str.str) != 0)
|
|
DBUG_RETURN(NULL);
|
|
|
|
const PFS_engine_table_share* result;
|
|
result= PFS_engine_table::find_engine_table_share(name);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
static int pfs_discover_table(handlerton *hton, THD *thd, TABLE_SHARE *share)
|
|
{
|
|
const PFS_engine_table_share *pfs_share;
|
|
|
|
if ((pfs_share= find_table_share(share->db.str, share->table_name.str)))
|
|
return share->init_from_sql_statement_string(thd, false,
|
|
pfs_share->sql.str,
|
|
pfs_share->sql.length);
|
|
return HA_ERR_NO_SUCH_TABLE;
|
|
}
|
|
|
|
static int pfs_discover_table_existence(handlerton *hton, const char *db,
|
|
const char *table_name)
|
|
{
|
|
return MY_TEST(find_table_share(db, table_name));
|
|
}
|
|
|
|
static int pfs_init_func(void *p)
|
|
{
|
|
DBUG_ENTER("pfs_init_func");
|
|
|
|
pfs_hton= reinterpret_cast<handlerton *> (p);
|
|
|
|
pfs_hton->create= pfs_create_handler;
|
|
pfs_hton->drop_table= [](handlerton *, const char*) { return -1; };
|
|
pfs_hton->show_status= pfs_show_status;
|
|
pfs_hton->flags= HTON_ALTER_NOT_SUPPORTED | HTON_TEMPORARY_NOT_SUPPORTED |
|
|
HTON_NO_PARTITION | HTON_NO_BINLOG_ROW_OPT;
|
|
|
|
/*
|
|
As long as the server implementation keeps using legacy_db_type,
|
|
as for example in mysql_truncate(),
|
|
we can not rely on the fact that different mysqld process will assign
|
|
consistently the same legacy_db_type for a given storage engine name.
|
|
In particular, using different --loose-skip-xxx options between
|
|
./mysqld --bootstrap
|
|
./mysqld
|
|
creates bogus .frm forms when bootstrapping the performance schema,
|
|
if we rely on ha_initialize_handlerton to assign a really dynamic value.
|
|
To fix this, a dedicated DB_TYPE is officially assigned to
|
|
the performance schema. See Bug#43039.
|
|
*/
|
|
pfs_hton->db_type= DB_TYPE_PERFORMANCE_SCHEMA;
|
|
pfs_hton->discover_table= pfs_discover_table;
|
|
pfs_hton->discover_table_existence= pfs_discover_table_existence;
|
|
pfs_hton->discover_table_names= pfs_discover_table_names;
|
|
|
|
PFS_engine_table_share::init_all_locks();
|
|
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
static int pfs_done_func(void *p)
|
|
{
|
|
DBUG_ENTER("pfs_done_func");
|
|
|
|
pfs_hton= NULL;
|
|
|
|
PFS_engine_table_share::delete_all_locks();
|
|
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
static int show_func_mutex_instances_lost(THD *thd, SHOW_VAR *var, void *buff,
|
|
struct system_status_var *status_var,
|
|
enum enum_var_type)
|
|
{
|
|
var->type= SHOW_LONG;
|
|
var->value= buff;
|
|
long *value= reinterpret_cast<long*>(buff);
|
|
*value= global_mutex_container.get_lost_counter();
|
|
return 0;
|
|
}
|
|
|
|
static struct st_mysql_show_var pfs_status_vars[]=
|
|
{
|
|
{"Performance_schema_mutex_classes_lost",
|
|
(char*) &mutex_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_rwlock_classes_lost",
|
|
(char*) &rwlock_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_cond_classes_lost",
|
|
(char*) &cond_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_thread_classes_lost",
|
|
(char*) &thread_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_file_classes_lost",
|
|
(char*) &file_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_socket_classes_lost",
|
|
(char*) &socket_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_memory_classes_lost",
|
|
(char*) &memory_class_lost, SHOW_LONG_NOFLUSH},
|
|
{"Performance_schema_mutex_instances_lost",
|
|
(char*) &show_func_mutex_instances_lost, SHOW_FUNC},
|
|
{"Performance_schema_rwlock_instances_lost",
|
|
(char*) &global_rwlock_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_cond_instances_lost",
|
|
(char*) &global_cond_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_thread_instances_lost",
|
|
(char*) &global_thread_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_file_instances_lost",
|
|
(char*) &global_file_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_file_handles_lost",
|
|
(char*) &file_handle_lost, SHOW_LONG},
|
|
{"Performance_schema_socket_instances_lost",
|
|
(char*) &global_socket_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_locker_lost",
|
|
(char*) &locker_lost, SHOW_LONG},
|
|
/* table shares, can be flushed */
|
|
{"Performance_schema_table_instances_lost",
|
|
(char*) &global_table_share_container.m_lost, SHOW_LONG},
|
|
/* table handles, can be flushed */
|
|
{"Performance_schema_table_handles_lost",
|
|
(char*) &global_table_container.m_lost, SHOW_LONG},
|
|
/* table lock stats, can be flushed */
|
|
{"Performance_schema_table_lock_stat_lost",
|
|
(char*) &global_table_share_lock_container.m_lost, SHOW_LONG},
|
|
/* table index stats, can be flushed */
|
|
{"Performance_schema_index_stat_lost",
|
|
(char*) &global_table_share_index_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_hosts_lost",
|
|
(char*) &global_host_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_users_lost",
|
|
(char*) &global_user_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_accounts_lost",
|
|
(char*) &global_account_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_stage_classes_lost",
|
|
(char*) &stage_class_lost, SHOW_LONG},
|
|
{"Performance_schema_statement_classes_lost",
|
|
(char*) &statement_class_lost, SHOW_LONG},
|
|
{"Performance_schema_digest_lost",
|
|
(char*) &digest_lost, SHOW_LONG},
|
|
{"Performance_schema_session_connect_attrs_lost",
|
|
(char*) &session_connect_attrs_lost, SHOW_LONG},
|
|
{"Performance_schema_program_lost",
|
|
(char*) &global_program_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_nested_statement_lost",
|
|
(char*) &nested_statement_lost, SHOW_LONG},
|
|
{"Performance_schema_prepared_statements_lost",
|
|
(char*) &global_prepared_stmt_container.m_lost, SHOW_LONG},
|
|
{"Performance_schema_metadata_lock_lost",
|
|
(char*) &global_mdl_container.m_lost, SHOW_LONG},
|
|
{NullS, NullS, SHOW_LONG}
|
|
};
|
|
|
|
struct st_mysql_storage_engine pfs_storage_engine=
|
|
{ MYSQL_HANDLERTON_INTERFACE_VERSION };
|
|
|
|
const char* pfs_engine_name= "PERFORMANCE_SCHEMA";
|
|
|
|
maria_declare_plugin(perfschema)
|
|
{
|
|
MYSQL_STORAGE_ENGINE_PLUGIN,
|
|
&pfs_storage_engine,
|
|
pfs_engine_name,
|
|
"Marc Alff, Oracle",
|
|
"Performance Schema",
|
|
PLUGIN_LICENSE_GPL,
|
|
pfs_init_func,
|
|
pfs_done_func,
|
|
0x0001,
|
|
pfs_status_vars,
|
|
NULL,
|
|
"5.7.31",
|
|
MariaDB_PLUGIN_MATURITY_STABLE
|
|
}
|
|
maria_declare_plugin_end;
|
|
|
|
ha_perfschema::ha_perfschema(handlerton *hton, TABLE_SHARE *share)
|
|
: handler(hton, share), m_table_share(NULL), m_table(NULL)
|
|
{}
|
|
|
|
ha_perfschema::~ha_perfschema() = default;
|
|
|
|
int ha_perfschema::open(const char *name, int mode, uint test_if_locked)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::open");
|
|
|
|
m_table_share= find_table_share(table_share->db.str,
|
|
table_share->table_name.str);
|
|
if (! m_table_share)
|
|
DBUG_RETURN(HA_ERR_NO_SUCH_TABLE);
|
|
|
|
thr_lock_data_init(m_table_share->m_thr_lock_ptr, &m_thr_lock, NULL);
|
|
ref_length= m_table_share->m_ref_length;
|
|
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
int ha_perfschema::close(void)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::close");
|
|
m_table_share= NULL;
|
|
delete m_table;
|
|
m_table= NULL;
|
|
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
int ha_perfschema::write_row(const uchar *buf)
|
|
{
|
|
int result;
|
|
|
|
DBUG_ENTER("ha_perfschema::write_row");
|
|
if (!PFS_ENABLED())
|
|
DBUG_RETURN(HA_ERR_WRONG_COMMAND);
|
|
|
|
DBUG_ASSERT(m_table_share);
|
|
result= m_table_share->write_row(table, buf, table->field);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
void ha_perfschema::use_hidden_primary_key(void)
|
|
{
|
|
/*
|
|
This is also called in case of row based replication,
|
|
see TABLE::mark_columns_needed_for_update().
|
|
Add all columns to the read set, but do not touch the write set,
|
|
as some columns in the SETUP_ tables are not writable.
|
|
*/
|
|
table->column_bitmaps_set_no_signal(&table->s->all_set, table->write_set);
|
|
}
|
|
|
|
int ha_perfschema::update_row(const uchar *old_data, const uchar *new_data)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::update_row");
|
|
if (!PFS_ENABLED())
|
|
DBUG_RETURN(HA_ERR_WRONG_COMMAND);
|
|
|
|
if (is_executed_by_slave())
|
|
DBUG_RETURN(0);
|
|
|
|
DBUG_ASSERT(m_table);
|
|
int result= m_table->update_row(table, old_data, new_data, table->field);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
int ha_perfschema::delete_row(const uchar *buf)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::delete_row");
|
|
if (!PFS_ENABLED())
|
|
DBUG_RETURN(HA_ERR_WRONG_COMMAND);
|
|
|
|
DBUG_ASSERT(m_table);
|
|
int result= m_table->delete_row(table, buf, table->field);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
int ha_perfschema::rnd_init(bool scan)
|
|
{
|
|
int result;
|
|
DBUG_ENTER("ha_perfschema::rnd_init");
|
|
|
|
assert(m_table_share);
|
|
assert(m_table_share->m_open_table != NULL);
|
|
|
|
stats.records= 0;
|
|
if (m_table == NULL)
|
|
m_table= m_table_share->m_open_table();
|
|
else
|
|
m_table->reset_position();
|
|
|
|
if (m_table != NULL)
|
|
m_table->rnd_init(scan);
|
|
|
|
result= m_table ? 0 : HA_ERR_OUT_OF_MEM;
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
int ha_perfschema::rnd_end(void)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::rnd_end");
|
|
assert(m_table);
|
|
delete m_table;
|
|
m_table= NULL;
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
int ha_perfschema::rnd_next(uchar *buf)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::rnd_next");
|
|
if (!PFS_ENABLED())
|
|
{
|
|
table->status= STATUS_NOT_FOUND;
|
|
DBUG_RETURN(HA_ERR_END_OF_FILE);
|
|
}
|
|
|
|
DBUG_ASSERT(m_table);
|
|
|
|
int result= m_table->rnd_next();
|
|
if (result == 0)
|
|
{
|
|
result= m_table->read_row(table, buf, table->field);
|
|
if (result == 0)
|
|
stats.records++;
|
|
}
|
|
table->status= (result ? STATUS_NOT_FOUND : 0);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
void ha_perfschema::position(const uchar *record)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::position");
|
|
|
|
assert(m_table);
|
|
m_table->get_position(ref);
|
|
DBUG_VOID_RETURN;
|
|
}
|
|
|
|
int ha_perfschema::rnd_pos(uchar *buf, uchar *pos)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::rnd_pos");
|
|
if (!PFS_ENABLED())
|
|
{
|
|
table->status= STATUS_NOT_FOUND;
|
|
DBUG_RETURN(HA_ERR_END_OF_FILE);
|
|
}
|
|
|
|
DBUG_ASSERT(m_table);
|
|
int result= m_table->rnd_pos(pos);
|
|
if (result == 0)
|
|
result= m_table->read_row(table, buf, table->field);
|
|
table->status= (result ? STATUS_NOT_FOUND : 0);
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
int ha_perfschema::info(uint flag)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::info");
|
|
assert(m_table_share);
|
|
if (flag & HA_STATUS_VARIABLE)
|
|
stats.records= m_table_share->get_row_count();
|
|
if (flag & HA_STATUS_CONST)
|
|
ref_length= m_table_share->m_ref_length;
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
int ha_perfschema::delete_all_rows(void)
|
|
{
|
|
int result;
|
|
|
|
DBUG_ENTER("ha_perfschema::delete_all_rows");
|
|
if (!PFS_ENABLED())
|
|
DBUG_RETURN(0);
|
|
|
|
if (is_executed_by_slave())
|
|
DBUG_RETURN(0);
|
|
|
|
assert(m_table_share);
|
|
if (m_table_share->m_delete_all_rows)
|
|
result= m_table_share->m_delete_all_rows();
|
|
else
|
|
{
|
|
result= HA_ERR_WRONG_COMMAND;
|
|
}
|
|
DBUG_RETURN(result);
|
|
}
|
|
|
|
int ha_perfschema::truncate()
|
|
{
|
|
return delete_all_rows();
|
|
}
|
|
|
|
THR_LOCK_DATA **ha_perfschema::store_lock(THD *thd,
|
|
THR_LOCK_DATA **to,
|
|
enum thr_lock_type lock_type)
|
|
{
|
|
if (lock_type != TL_IGNORE && m_thr_lock.type == TL_UNLOCK)
|
|
m_thr_lock.type= lock_type;
|
|
*to++= &m_thr_lock;
|
|
m_thr_lock.m_psi= m_psi;
|
|
return to;
|
|
}
|
|
|
|
int ha_perfschema::delete_table(const char *name)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::delete_table");
|
|
|
|
/*
|
|
The name string looks like:
|
|
"./performance_schema/processlist"
|
|
|
|
Make a copy of it, parse the '/' to
|
|
isolate the schema and table name.
|
|
*/
|
|
|
|
char table_path[FN_REFLEN+1];
|
|
strncpy(table_path, name, sizeof(table_path));
|
|
table_path[FN_REFLEN]='\0';
|
|
|
|
char *ptr;
|
|
char *table_name;
|
|
char *db_name;
|
|
const PFS_engine_table_share *share;
|
|
|
|
/* Start scan from the end. */
|
|
ptr = strend(table_path) - 1;
|
|
|
|
/* Find path separator */
|
|
while ((ptr >= table_path) && (*ptr != '\\') && (*ptr != '/')) {
|
|
ptr--;
|
|
}
|
|
|
|
table_name = ptr + 1;
|
|
*ptr = '\0';
|
|
|
|
/* Find path separator */
|
|
while ((ptr >= table_path) && (*ptr != '\\') && (*ptr != '/')) {
|
|
ptr--;
|
|
}
|
|
|
|
db_name = ptr + 1;
|
|
|
|
share = find_table_share(db_name, table_name);
|
|
if (share != NULL) {
|
|
if (share->m_optional) {
|
|
/*
|
|
An optional table is deleted,
|
|
disarm the checked flag so we don't trust it any more.
|
|
*/
|
|
share->m_state->m_checked = false;
|
|
}
|
|
}
|
|
|
|
DBUG_RETURN(0);
|
|
}
|
|
|
|
int ha_perfschema::rename_table(const char * from, const char * to)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::rename_table ");
|
|
DBUG_RETURN(HA_ERR_WRONG_COMMAND);
|
|
}
|
|
|
|
int ha_perfschema::create(const char *name, TABLE *table_arg,
|
|
HA_CREATE_INFO *create_info)
|
|
{
|
|
DBUG_ENTER("ha_perfschema::create");
|
|
/*
|
|
This is not a general purpose engine.
|
|
Failure to CREATE TABLE is the expected result.
|
|
*/
|
|
DBUG_RETURN(HA_ERR_WRONG_COMMAND);
|
|
}
|
|
|
|
void ha_perfschema::print_error(int error, myf errflag)
|
|
{
|
|
switch (error)
|
|
{
|
|
case HA_ERR_TABLE_NEEDS_UPGRADE:
|
|
/*
|
|
The error message for ER_TABLE_NEEDS_UPGRADE refers to REPAIR table,
|
|
which does not apply to performance schema tables.
|
|
*/
|
|
my_error(ER_WRONG_NATIVE_TABLE_STRUCTURE, MYF(0),
|
|
table_share->db.str, table_share->table_name.str);
|
|
break;
|
|
case HA_ERR_WRONG_COMMAND:
|
|
/*
|
|
The performance schema is not a general purpose storage engine,
|
|
some operations are not supported, by design.
|
|
We do not want to print "Command not supported",
|
|
which gives the impression that a command implementation is missing,
|
|
and that the failure should be considered a bug.
|
|
We print "Invalid performance_schema usage." instead,
|
|
to emphasise that the operation attempted is not meant to be legal,
|
|
and that the failure returned is indeed the expected result.
|
|
*/
|
|
my_error(ER_WRONG_PERFSCHEMA_USAGE, MYF(0));
|
|
break;
|
|
default:
|
|
handler::print_error(error, errflag);
|
|
break;
|
|
}
|
|
}
|
|
|