Bug#45243: crash on win in sql thread clear_tables_to_lock() -> free()
Bug#45242: crash on win in mysql_close() -> free()
Bug#45238: rpl_slave_skip, rpl_change_master failed (lost connection) for STOP SLAVE
Bug#46030: rpl_truncate_3innodb causes server crash on windows
Bug#46014: rpl_stm_reset_slave crashes the server sporadically in pb2
When killing a user session on the server, it's necessary to
interrupt (notify) the thread associated with the session that
the connection is being killed so that the thread is woken up
if waiting for I/O. On a few platforms (Mac, Windows and HP-UX)
where the SIGNAL_WITH_VIO_CLOSE flag is defined, this interruption
procedure is to asynchronously close the underlying socket of
the connection.
In order to enable this schema, each connection serving thread
registers its VIO (I/O interface) so that other threads can
access it and close the connection. But only the owner thread of
the VIO might delete it as to guarantee that other threads won't
see freed memory (the thread unregisters the VIO before deleting
it). A side note: closing the socket introduces a harmless race
that might cause a thread attempt to read from a closed socket,
but this is deemed acceptable.
The problem is that this infrastructure was meant to only be used
by server threads, but the slave I/O thread was registering the
VIO of a mysql handle (a client API structure that represents a
connection to another server instance) as a active connection of
the thread. But under some circumstances such as network failures,
the client API might destroy the VIO associated with a handle at
will, yet the VIO wouldn't be properly unregistered. This could
lead to accesses to freed data if a thread attempted to kill a
slave I/O thread whose connection was already broken.
There was a attempt to work around this by checking whether
the socket was being interrupted, but this hack didn't work as
intended due to the aforementioned race -- attempting to read
from the socket would yield a "bad file descriptor" error.
The solution is to add a hook to the client API that is called
from the client code before the VIO of a handle is deleted.
This hook allows the slave I/O thread to detach the active vio
so it does not point to freed memory.
server-tools/instance-manager/mysql_connection.cc:
Add stub method required for linking.
sql-common/client.c:
Invoke hook.
sql/client_settings.h:
Export hook.
sql/slave.cc:
Introduce hook that clears the active VIO before it is freed
by the client API.
Bug#45243: crash on win in sql thread clear_tables_to_lock() -> free()
Bug#45242: crash on win in mysql_close() -> free()
Bug#45238: rpl_slave_skip, rpl_change_master failed (lost connection) for STOP SLAVE
Bug#46030: rpl_truncate_3innodb causes server crash on windows
Bug#46014: rpl_stm_reset_slave crashes the server sporadically in pb2
When killing a user session on the server, it's necessary to
interrupt (notify) the thread associated with the session that
the connection is being killed so that the thread is woken up
if waiting for I/O. On a few platforms (Mac, Windows and HP-UX)
where the SIGNAL_WITH_VIO_CLOSE flag is defined, this interruption
procedure is to asynchronously close the underlying socket of
the connection.
In order to enable this schema, each connection serving thread
registers its VIO (I/O interface) so that other threads can
access it and close the connection. But only the owner thread of
the VIO might delete it as to guarantee that other threads won't
see freed memory (the thread unregisters the VIO before deleting
it). A side note: closing the socket introduces a harmless race
that might cause a thread attempt to read from a closed socket,
but this is deemed acceptable.
The problem is that this infrastructure was meant to only be used
by server threads, but the slave I/O thread was registering the
VIO of a mysql handle (a client API structure that represents a
connection to another server instance) as a active connection of
the thread. But under some circumstances such as network failures,
the client API might destroy the VIO associated with a handle at
will, yet the VIO wouldn't be properly unregistered. This could
lead to accesses to freed data if a thread attempted to kill a
slave I/O thread whose connection was already broken.
There was a attempt to work around this by checking whether
the socket was being interrupted, but this hack didn't work as
intended due to the aforementioned race -- attempting to read
from the socket would yield a "bad file descriptor" error.
The solution is to add a hook to the client API that is called
from the client code before the VIO of a handle is deleted.
This hook allows the slave I/O thread to detach the active vio
so it does not point to freed memory.
The problem is that the C API function mysql_real_connect
only attempts to connect to the first IP address returned
for a hostname. This can be a problem if a hostname maps
to multiple IP address and the server is not bound to the
first one that is returned.
The solution is to augment mysql_real_connect so that it
attempts to connect to all IPv4 addresses that a domain
name maps to. The function goes over the list of address
until a successful connection is established.
No test case is provided as its not possible to test this
automatically with the current testing infrastructure.
sql-common/client.c:
The client will try to connect to each IPv4 address from
the list of addresses that hostname maps to until a successful
connection is established or there are no more address.
The problem is that the C API function mysql_real_connect
only attempts to connect to the first IP address returned
for a hostname. This can be a problem if a hostname maps
to multiple IP address and the server is not bound to the
first one that is returned.
The solution is to augment mysql_real_connect so that it
attempts to connect to all IPv4 addresses that a domain
name maps to. The function goes over the list of address
until a successful connection is established.
No test case is provided as its not possible to test this
automatically with the current testing infrastructure.
Fix sleep() synchronisation in innodb_information_schema test case.
mysql-test/t/innodb_information_schema.test:
Using sleep for synchronisation does not work!!!
Replace by looping until the required condition is met.
sql-common/client.c:
mysql_ssl_set() did not free old pointers before overwriting with new ones (happens when
mysql_ssl_set() is called twice without calling mysql_close() in-between).
This sometimes caused memory leaks in the slave depending on exact timing of
master/slave shutdown.
Fixed by freeing old pointers before installing new ones in mysql_ssl_set(), just like
mysql_options() does.
Backport to MySQL 5.0/1 fix by Vladislav Vaintroub:
In Vista and later and also in when using terminal services, when
server is started from command line, client cannot connect to it
via shared memory protocol.
This is a regression introduced when Bug#24731 was fixed. The
reason is that client is trying to attach to shared memory using
global kernel object namespace (all kernel objects are prefixed
with Global\). However, server started from the command line in
Vista and later will create shared memory and events using current
session namespace. Thus, client is unable to find the server and
connection fails.
The fix for the client is to first try to find server using "local"
names (omitting Global\ prefix) and only if server is not found,
trying global namespace.
Backport to MySQL 5.0/1 fix by Vladislav Vaintroub:
In Vista and later and also in when using terminal services, when
server is started from command line, client cannot connect to it
via shared memory protocol.
This is a regression introduced when Bug#24731 was fixed. The
reason is that client is trying to attach to shared memory using
global kernel object namespace (all kernel objects are prefixed
with Global\). However, server started from the command line in
Vista and later will create shared memory and events using current
session namespace. Thus, client is unable to find the server and
connection fails.
The fix for the client is to first try to find server using "local"
names (omitting Global\ prefix) and only if server is not found,
trying global namespace.
in load_defaults()
load_defaults(), my_search_option_files() and
my_print_default_files() utilized a global variable
containing a pointer to thread local memory. This could lead
to race conditions when those functions were called with high
concurrency.
Fixed by changing the interface of the said functions to avoid
the necessity for using a global variable.
Since we cannot change load_defaults() prototype for API
compatibility reasons, it was renamed my_load_defaults().
Now load_defaults() is a thread-unsafe wrapper around
a thread-safe version, my_load_defaults().
mysys/default.c:
1. Added a thread-safe version of load_defaults(), changed
load_defaults() with the old interface to be a thread-unsafe
wrapper around the thread-safe version.
2. Always use a private MEM_ROOT in my_print_default_files,
don't use a global variable.
sql-common/client.c:
Use a thread-safe version of load_defaults().
in load_defaults()
load_defaults(), my_search_option_files() and
my_print_default_files() utilized a global variable
containing a pointer to thread local memory. This could lead
to race conditions when those functions were called with high
concurrency.
Fixed by changing the interface of the said functions to avoid
the necessity for using a global variable.
Since we cannot change load_defaults() prototype for API
compatibility reasons, it was renamed my_load_defaults().
Now load_defaults() is a thread-unsafe wrapper around
a thread-safe version, my_load_defaults().
- Remove bothersome warning messages. This change focuses on the warnings
that are covered by the ignore file: support-files/compiler_warnings.supp.
- Strings are guaranteed to be max uint in length
- Remove bothersome warning messages. This change focuses on the warnings
that are covered by the ignore file: support-files/compiler_warnings.supp.
- Strings are guaranteed to be max uint in length
- Remove bothersome warning messages. This change focuses on the warnings
that are covered by the ignore file: support-files/compiler_warnings.supp.
- Strings are guaranteed to be max uint in length
- Remove bothersome warning messages. This change focuses on the warnings
that are covered by the ignore file: support-files/compiler_warnings.supp.
- Strings are guaranteed to be max uint in length
into quad.opbmk:/mnt/raid/alik/MySQL/devel/5.0-rt-merged
libmysql/libmysql.c:
Auto merged
sql-common/client.c:
Auto merged
tests/mysql_client_test.c:
Manually merged.
into quad.opbmk:/mnt/raid/alik/MySQL/devel/5.1-rt-merged
libmysql/libmysql.c:
Auto merged
sql-common/client.c:
Auto merged
sql/mysql_priv.h:
Auto merged
sql/mysqld.cc:
Auto merged
tests/mysql_client_test.c:
Auto merged
The problem was that the COM_STMT_SEND_LONG_DATA was sending a response
packet if the prepared statement wasn't found in the server (due to
reconnection). The commands COM_STMT_SEND_LONG_DATA and COM_STMT_CLOSE
should not send any packets, even error packets should not be sent since
they are not expected by the client API.
The solution is to clear generated during the execution of the aforementioned
commands and to skip resend of prepared statement commands. Another fix is
that if the connection breaks during the send of prepared statement command,
the command is not sent again since the prepared statement is no longer in the
server.
libmysql/libmysql.c:
The mysql handle might be reset after a reconnection.
Pass the now used stmt argument to cli_advanced_command.
sql-common/client.c:
Don't resend command if the connection broke and it's a prepared
statement command. If the session is broken, prepared statements
on the server are gone, set the error accordanly.
sql/sql_prepare.cc:
Clear any error set during the execution of the request
command.
tests/mysql_client_test.c:
Fix memory leak by freeing result associated with statement.
Remove test case for Bug 29948 because it's not reliable in
5.0 (fixed in 5.1) due to KILL queries sending two packets for
a thread that kills itself.
The problem was that the COM_STMT_SEND_LONG_DATA was sending a response
packet if the prepared statement wasn't found in the server (due to
reconnection). The commands COM_STMT_SEND_LONG_DATA and COM_STMT_CLOSE
should not send any packets, even error packets should not be sent since
they are not expected by the client API.
The solution is to clear generated during the execution of the aforementioned
commands and to skip resend of prepared statement commands. Another fix is
that if the connection breaks during the send of prepared statement command,
the command is not sent again since the prepared statement is no longer in the
server.
in some case.
ER_CON_COUNT_ERROR is defined with SQL state 08004. However, this SQL state is not always
returned.
This error can be thrown in two cases:
1. when an ordinary user (a user w/o SUPER privilege) is connecting,
and the number of active user connections is equal or greater than
max_connections.
2. when a user is connecting and the number of active user connections is
already (max_connections + 1) -- that means that no more connections will
be accepted regardless of the user credentials.
In the 1-st case, SQL state is correct.
The bug happens in the 2-nd case -- on UNIX the client gets 00000 SQL state, which is
absolutely wrong (00000 means "not error SQL state); on Windows
the client accidentally gets HY000 (which means "unknown SQL state).
The cause of the problem is that the server rejects extra connection
prior to read a packet with client capabilities. Thus, the server
does not know if the client supports SQL states or not (if the client
supports 4.1 protocol or not). So, the server supposes the worst and
does not send SQL state at all.
The difference in behavior on UNIX and Windows occurs because on Windows
CLI_MYSQL_REAL_CONNECT() invokes create_shared_memory(), which returns
an error (in default configuration, where shared memory is not configured).
Then, the client does not reset this error, so when the connection is
rejected, SQL state is HY000 (from the error from create_shared_memory()).
The bug appeared after test case for Bug#33507 -- before that, this behavior
just had not been tested.
The fix is to 1) reset the error after create_shared_memory();
2) set SQL state to 'unknown error' if it was not received from
the server.
A separate test case is not required, since the behavior is already
tested in connect.test.
Note for doc-team: the manual should be updated to say that under
some circumstances, 'Too many connections' has HY000 SQL state.
mysql-test/r/connect.result:
Update result file.
sql-common/client.c:
1. Reset an error from create_shared_memory();
2. Set SQL state to 'unknown error' if it was not received from
the server.
in some case.
ER_CON_COUNT_ERROR is defined with SQL state 08004. However, this SQL state is not always
returned.
This error can be thrown in two cases:
1. when an ordinary user (a user w/o SUPER privilege) is connecting,
and the number of active user connections is equal or greater than
max_connections.
2. when a user is connecting and the number of active user connections is
already (max_connections + 1) -- that means that no more connections will
be accepted regardless of the user credentials.
In the 1-st case, SQL state is correct.
The bug happens in the 2-nd case -- on UNIX the client gets 00000 SQL state, which is
absolutely wrong (00000 means "not error SQL state); on Windows
the client accidentally gets HY000 (which means "unknown SQL state).
The cause of the problem is that the server rejects extra connection
prior to read a packet with client capabilities. Thus, the server
does not know if the client supports SQL states or not (if the client
supports 4.1 protocol or not). So, the server supposes the worst and
does not send SQL state at all.
The difference in behavior on UNIX and Windows occurs because on Windows
CLI_MYSQL_REAL_CONNECT() invokes create_shared_memory(), which returns
an error (in default configuration, where shared memory is not configured).
Then, the client does not reset this error, so when the connection is
rejected, SQL state is HY000 (from the error from create_shared_memory()).
The bug appeared after test case for Bug#33507 -- before that, this behavior
just had not been tested.
The fix is to 1) reset the error after create_shared_memory();
2) set SQL state to 'unknown error' if it was not received from
the server.
A separate test case is not required, since the behavior is already
tested in connect.test.
Note for doc-team: the manual should be updated to say that under
some circumstances, 'Too many connections' has HY000 SQL state.
into kaamos.(none):/data/src/opt/mysql-5.1-opt
client/mysql.cc:
Auto merged
client/mysqldump.c:
Auto merged
configure.in:
Auto merged
include/my_global.h:
Auto merged
libmysql/libmysql.c:
Auto merged
libmysqld/lib_sql.cc:
Auto merged
mysql-test/include/mix1.inc:
Auto merged
mysql-test/r/create.result:
Auto merged
mysql-test/r/func_str.result:
Auto merged
mysql-test/r/innodb.result:
Auto merged
mysql-test/r/innodb_mysql.result:
Auto merged
mysql-test/r/select.result:
Auto merged
mysql-test/r/subselect.result:
Auto merged
mysql-test/t/create.test:
Auto merged
mysql-test/t/disabled.def:
Auto merged
sql/filesort.cc:
Auto merged
sql/handler.cc:
Auto merged
sql/item.cc:
Auto merged
sql/item.h:
Auto merged
sql/item_cmpfunc.cc:
Auto merged
sql/item_func.cc:
Auto merged
sql/item_func.h:
Auto merged
sql/item_sum.cc:
Auto merged
sql/log.cc:
Auto merged
sql/mysql_priv.h:
Auto merged
sql/mysqld.cc:
Auto merged
sql/set_var.cc:
Auto merged
sql/sp.cc:
Auto merged
sql/sql_acl.cc:
Auto merged
sql/sql_base.cc:
Auto merged
sql/sql_insert.cc:
Auto merged
sql/sql_plugin.cc:
Auto merged
sql/sql_select.cc:
Auto merged
sql/sql_yacc.yy:
Auto merged
storage/ndb/src/kernel/blocks/backup/Backup.hpp:
Auto merged
tests/mysql_client_test.c:
Auto merged
mysql-test/r/func_time.result:
Manual merge.
mysql-test/r/view.result:
Manual merge.
mysql-test/t/view.test:
Manual merge.
scripts/mysql_config.sh:
Manual merge.
sql-common/client.c:
Manual merge.
sql/sql_parse.cc:
Manual merge.
Rename client_last_error to last_error and client_last_errno to last_errno
to not break connectors which use the internal net structure for error handling.
include/mysql_com.h:
Rename client_last_error to last_error, client_last_errno to last_errno.
include/mysql_h.ic:
Rename client_last_error to last_error, client_last_errno to last_errno.
libmysql/libmysql.c:
Rename client_last_error to last_error, client_last_errno to last_errno.
libmysql/manager.c:
Rename client_last_error to last_error, client_last_errno to last_errno.
libmysqld/lib_sql.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
libmysqld/libmysqld.c:
Rename client_last_error to last_error, client_last_errno to last_errno.
server-tools/instance-manager/mysql_connection.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
sql/log_event.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
sql-common/client.c:
Rename client_last_error to last_error, client_last_errno to last_errno.
sql/log_event_old.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
sql/net_serv.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
sql/repl_failsafe.cc:
Rename client_last_error to last_error, client_last_errno to last_errno.
Rename client_last_error to last_error and client_last_errno to last_errno
to not break connectors which use the internal net structure for error handling.
into mysql.com:/home/hf/work/25097/my51-25097
libmysql/libmysql.c:
Auto merged
libmysqld/lib_sql.cc:
Auto merged
sql-common/client.c:
Auto merged
sql/log.cc:
Auto merged
into mysql.com:/home/hf/work/25097/my51-25097
client/mysql.cc:
Auto merged
include/sql_common.h:
Auto merged
libmysql/libmysql.c:
Auto merged
libmysqld/lib_sql.cc:
Auto merged
sql-common/client.c:
merging
sql/log.cc:
merging
There was no way to return an error from the client library
if no MYSQL connections was established.
So here i added variables to store that king of errors and
made functions like mysql_error(NULL) to return these.
client/mysql.cc:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
Show the error message on std_error
include/sql_common.h:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
cant_connect_sqlstate constant declared
libmysql/libmysql.c:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
mysql_sqlstate(NULL) returns 'unknown_sqlstate'
libmysqld/lib_sql.cc:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
EMBEDDED_SERVER version of the vprint_msg_to_log() implemented
sql-common/client.c:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
mysql_server_last_errno and mysql_server_last_error introduced
to store errors not related to particular connections.
mysql_error(NULL) and mysql_errno(NULL) now returns these
mysql_server_last_error and errno respectively
sql/log.cc:
Bug #25097 mysql_server_init fails silently if no errmsg.sys is present.
EMBEDDED_LIBRARY implementation of the vprint_msg_to_log() moved
to lib_sql.cc
There was no way to return an error from the client library
if no MYSQL connections was established.
So here i added variables to store that king of errors and
made functions like mysql_error(NULL) to return these.
--local-infile=0 checks can be bypassed by sending a FETCH LOCAL FILE response
Add a check for CLIENT_LOCAL_FILES before sending a local file.
Beware, that all binary distributions enable sending of local files and it's up
to the programs which use libmysql to disable it, if they don't use this functionality.
Otherwise they are not safe.
client/mysqltest.c:
Enable LOAD DATA LOCAL INFILE for the test suite, like some rpl and ndb test.
sql-common/client.c:
Check if the client has LOAD DATA LOCAL INFILE disabled and
don't serve such requests from the server. This is not 100% proof,
as if the client has this enabled, in all binary builds for BC,
the check won't work and the client can be tricked into sending a
local file.
tests/mysql_client_test.c:
Switch on LOCAL INFILE in client test. If one day there
is a test which uses it, then it will work out of the box.
--local-infile=0 checks can be bypassed by sending a FETCH LOCAL FILE response
Add a check for CLIENT_LOCAL_FILES before sending a local file.
Beware, that all binary distributions enable sending of local files and it's up
to the programs which use libmysql to disable it, if they don't use this functionality.
Otherwise they are not safe.
into dl145h.mysql.com:/data0/mkindahl/mysql-5.1-rpl-merge
BitKeeper/deleted/.del-show_binlog_events2.inc:
Auto merged
sql-common/client.c:
Auto merged
sql/item_cmpfunc.cc:
Auto merged
sql/log_event.cc:
Auto merged
sql/slave.cc:
Auto merged
