Automerged bzr bundle from bug report.

Automerging bzr bundle from bug report into latest 
mysql-5.1-bugteam.

Fixing test case that fails on Windows because .dll is used.

When installing plugins, there is a missing check
for slash (/) in the path on Windows. Note that on
Windows, both / and \ can be used to separate
directories.

This patch fixes the issue by:
- Adding a FN_DIRSEP symbol for all platforms
  consisting of a string of legal directory
  separators.
- Adding a charset-aware version of strcspn().
- Adding a check_valid_path() function that uses
  my_strcspn() to check if any FN_DIRSEP character
  is in the supplied string.
- Using the check_valid_path() function in
  sql_plugin.cc and sql_udf.cc (which means
  replacing the existing test there).

include/config-netware.h:
  Adding FN_DIRSEP
  ******
  Adding FN_DIRSEP
include/config-win.h:
  Adding FN_DIRSEP
  ******
  Adding FN_DIRSEP
include/m_ctype.h:
  Adding my_strspn() and my_strcspn().
  
  ******
  Adding my_strspn() and my_strcspn().
include/my_global.h:
  Adding FN_DIRSEP
  ******
  Adding FN_DIRSEP
mysql-test/t/plugin_not_embedded.test:
  Adding test that file names containing / is
  disallowed on *all* platforms.
  ******
  Adding test that file names containing / is
  disallowed on *all* platforms.
sql/sql_plugin.cc:
  Introducing check_if_path() function for
  checking if filename is a path to include
  / on Windows.
  ******
  Introducing check_if_path() function for
  checking if filename is a path to include
  / on Windows.
sql/sql_udf.cc:
  Switching to use check_if_path() function.
  ******
  Switching to use check_if_path() function.
strings/my_strchr.c:
  Adding my_strspn() and my_strcspn().
  ******
  Adding my_strspn() and my_strcspn().

Fixing the testcase to use the database name
as connected_db instead of 'test' database.

mysql-test/r/mysql.result:
  Additional fix in the test for bug#54899.
mysql-test/t/mysql.test:
  Additional fix in the test for bug#54899.

           when generating new name.
      
If find_uniq_filename returns an error, then this error is not
being propagated upwards, and execution does not report error to
the user (although a entry in the error log is generated).
                  
Additionally, some more errors were ignored in new_file_impl:
- when writing the rotate event
- when reopening the index and binary log file
                  
This patch addresses this by propagating the error up in the
execution stack. Furthermore, when rotation of the binary log
fails, an incident event is written, because there may be a
chance that some changes for a given statement, were not properly
logged. For example, in SBR, LOAD DATA INFILE statement requires
more than one event to be logged, should rotation fail while
logging part of the LOAD DATA events, then the logged data would
become inconsistent with the data in the storage engine.

mysql-test/include/restart_mysqld.inc:
  Refactored restart_mysqld so that it is not hardcoded for
  mysqld.1, but rather for the current server.
mysql-test/suite/binlog/t/binlog_index.test:
  The error on open of index and binary log on new_file_impl 
  is now caught. Thence the user will get an error message. 
  We need to accomodate this change in the test case for the
  failing FLUSH LOGS.
mysql-test/suite/rpl/t/rpl_binlog_errors-master.opt:
  Sets max_binlog_size to 4096.
mysql-test/suite/rpl/t/rpl_binlog_errors.test:
  Added some test cases for asserting that the error is found 
  and reported.
sql/handler.cc:
  Catching error now returned by unlog (in ha_commit_trans) and 
  returning it.
sql/log.cc:
  Propagating errors from new_file_impl upwards. The errors that
  new_file_impl catches now are:
  - error on generate_new_name
  - error on writing the rotate event
  - error when opening the index or the binary log file.
sql/log.h:
  Changing declaration of:
  - rotate_and_purge
  - new_file
  - new_file_without_locking
  - new_file_impl
  - unlog
  They now return int instead of void.
sql/mysql_priv.h:
  Change signature of reload_acl_and_cache so that write_to_binlog
  is an int instead of bool.
sql/mysqld.cc:
  Redeclaring not_used var as int instead of bool.
sql/rpl_injector.cc:
  Changes to catch the return from rotate_and_purge.
sql/slave.cc:
  Changes to catch the return values for new_file and rotate_relay_log.
sql/slave.h:
  Changes to rotate_relay_log declaration (now returns int 
  instead of void).
sql/sql_load.cc:
  In SBR, some logging of LOAD DATA events goes through
  IO_CACHE_CALLBACK invocation at mf_iocache.c:_my_b_get. The
  IO_CACHE implementation is ignoring the return value for from
  these callbacks (pre_read and post_read), so we need to find out
  at the end of the execution if the error is set or not in THD.
sql/sql_parse.cc:
  Catching the rotate_relay_log and rotate_and_purge return values.
  Semantic change in reload_acl_and_cache so that we report errors
  in binlog interactions through the write_to_binlog output parameter.
  If there was any failure while rotating the binary log, we should
  then report the error to the client when handling SQLCOMM_FLUSH.

integral type. Use intptr which is designed to hold pointer values
and pass it to off_t.

mysys/stacktrace.c:
  Add a compile time assert to ensure that off_t is large enough
  to hold the pointer value.

win x86 debug_max

The windows MTR run exhibited a different test execution 
ordering (due to the fact that in these platforms MTR is invoked
with --parallel > 1). This uncovered a bug in the aforementioned
test case, which is triggered by the following conditions:

  1. server is not restarted between two different tests;
  2. the test before binlog.binlog_row_failure_mixing_engines
     issues flush logs;
  3. binlog.binlog_row_failure_mixing_engines uses binlog
     positions to limit the output of show_binlog_events;
  4. binlog.binlog_row_failure_mixing_engines does not state which
     binlog file to use, thence it uses a wrong binlog file with
     the correct position.

There are two possible fixes: 1. make sure that the test start 
from a clean slate - binlog wise; 2. in addition to the position, 
also state the binary log file before sourcing 
show_binlog_events.inc .

We go for fix #1, ie, deploy a RESET MASTER before the test is 
actually started.

The problem is that the logic which checks if a pointer is
valid relies on a poor heuristic based on the start and end
addresses of the data segment and heap.

Apart from miscalculating the heap bounds, this approach also
suffers from the fact that memory can come from places other
than the heap. See Bug#58528 for a more detailed explanation.

On Linux, the solution is to access the process's memory
through /proc/self/task/<tid>/mem, which allows for retrieving
the contents of pages within the virtual address space of
the calling process. If a address range is not mapped, a
input/output error is returned.

client/mysqltest.cc:
  Use new interface to my_safe_print_str.
include/my_stacktrace.h:
  Drop name from my_safe_print_str.
mysys/stacktrace.c:
  Access the process's memory through a file descriptor and
  dump the contents of the memory range. The file descriptor
  offset is equivalent to a offset into the address space.
  
  Do not print the name of the variable associated with the
  address. It can be better accomplished at a higher level.
sql/mysqld.cc:
  Put the variable dumping information within its own newline block.
  Use symbolic names which better convey information to the user.

the DROP statement ..."
      
Problem: When using temporary tables and closing a session, an
implicit DROP TEMPORARY TABLE IF EXISTS is written to the binary
log (while cleaning up the context of the session THD - see:
sql_class.cc:THD::cleanup which calls close_temporary_tables).
     
close_temporary_tables, first checks if the binary log is opened
and then proceeds to creating the DROP statements. Then, such
statements, are written to the binary log through
MYSQL_BIN_LOG::write(Log_event *). Inside, there is another check
if the binary log is opened and if not an error is returned. This
is where the faulty behavior is triggered. Given that the test
case replays a binary log, with temp tables statements, and right
after it issues RESET MASTER, there is a chance that is_open will
report false (when the mysql session is closed and the temporary
tables are written).
      
is_open may return false, because MYSQL_BIN_LOG::reset_logs is
not setting the correct flag (LOG_CLOSE_TO_BE_OPENED), on the
MYSQL_LOG_BIN::log_state (instead it sets just the
LOG_CLOSE_INDEX flag, leaving the log_state to
LOG_CLOSED). Thence, when writing the DROP statement as part of
the THD::cleanup, the thread could get a return value of false
for is_open - inside MYSQL_BIN_LOG::write, ultimately reporting
that it can't write the event to the binary log.
      
Fix: We fix this by adding the correct flag, missing in the
second close.

mysql-test/r/show_check.result:
  Fixed corrupted result file

metadata"
            
Improved error handling such that queries against Information_Schema.Tables won't
fail if a federated table can't make a remote connection.

mysql-test/r/merge.result:
  Updated with warnings that were previously masked.
mysql-test/r/show_check.result:
  Updated with warnings that were previously masked.
mysql-test/r/view.result:
  Updated with warnings that were previously masked.
sql/sql_show.cc:
  If get_schema_tables_record() encounters an error, push a warning,
  set the TABLE COMMENT column with the error text, and clear the
  error so that the operation can continue.

Improved error handling such that queries against Information_Schema.Tables won't
fail if a Federated table is unable to connect to remote host.



sql/sql_show.cc:
  If Handler::Info() fails, save error text in TABLE COMMENTS column, clear error.

will not display "indicated result file not found" due to wrongly named
var directory.

When using BINLOG statement to execute rows log events, session variables
foreign_key_checks and unique_checks are changed temporarily.  As each rows
log event has their own special session environment and its own
foreign_key_checks and unique_checks can be different from current session
which executing the BINLOG statement. But these variables are not restored
correctly after BINLOG statement. This problem will cause that the following
statements fail or generate unexpected data.

In this patch, code is added to backup and restore these two variables.
So BINLOG statement will not affect current session's variables again.


mysql-test/extra/binlog_tests/binlog.test:
  Add test to verify this patch.
mysql-test/suite/binlog/r/binlog_row_binlog.result:
  Add test to verify this patch.
mysql-test/suite/binlog/r/binlog_stm_binlog.result:
  Add test to verify this patch.
sql/sql_binlog.cc:
  Add code to backup and restore thd->options.

Problem: MySQL cp1251 did not support 'U+20AC EURO SIGN'
which was assigned a few years ago to 0x88.

Fix: adding mapping: 0x88 <-> U+20AC 

  @ mysql-test/include/ctype_8bit.inc
  New shared file to test 8bit character sets.

  @ mysql-test/r/ctype_cp1251.result
  @ mysql-test/t/ctype_cp1251.test
  Adding tests

  @ sql/share/charsets/cp1251.xml
  Adding mapping

  @ strings/ctype-extra.c
  Regenerating ctype-extra.c using strings/conf_to_src
  according to new cp1251.xml

After dropping and recreating the database specified along with --one-database
option at command line, mysql client keeps filtering the statements even after
the execution of a 'USE' command on the same database.

--one-database option enables the filtering of statements when the current
database is not the one specified at the command line. However, when the same
database is dropped and recreated the variable (current_db) that holds the
inital database name gets altered. This bug exploits the fact that current_db
initially gets set to null value (0) when a 'use db_name' follows the recreation
of same database db_name (speficied at the command line) and hence skip_updates
gets set to 1, which inturn triggers the further filtering of statements.

Fixed by making get_current_db() a no-op function when one_database is set,
and hence, under that condition current_db will not get altered.
Note, however the value of current_db can change when we execute 'connect'
command with a differnet database to reconnect to the server, in which case,
the behavior of --one-database will be formulated using this new database.


client/mysql.cc:
  Bug #54899 : --one-database option cannot handle DROP/CREATE DATABASE commands
  
  Added an if statement at the beginnning of get_current_db() , which makes it
  a no-op function if one-database option is specified, and hence current_db
  remains unchanged.
  
  Changed the help message for one-database option to a more appropriate message
  as specified in mysql documentation.
mysql-test/r/mysql.result:
  Added a test case for bug#54899 and some more test cases to
  check other one-database option related behaviors.
mysql-test/t/mysql.test:
  Added a test case for bug#54899 and some more test cases to
  check other one-database option related behaviors.

Fixed as suggested in the .inc file and two tests
Could not reproduce problem, but tested tests on Windows

It does work in general, the problem here was that the test name
'alter_table' matches 'main.alter_table-big' which has already been found.
Fixed by matching more explicitly (with/without suite name)

43233/55794.

mysql-test/r/change_user.result:
  Don't use -1 integer wrap around. It used to work, but now we do what's
  actually in the documentation. In tests, we now use DEFAULT or the
  numeral equivalent (as we do in the 5.6 tests).
mysql-test/r/key_cache.result:
  Can't drop default key case is an error now, not a warning, for compatibility
  with 5.6.
mysql-test/r/variables.result:
  Can't drop default key case is an error now, not a warning, for compatibility
  with 5.6.
mysql-test/t/change_user.test:
  Don't use -1 integer wrap around. It used to work, but now we do what's
  actually in the documentation. In tests, we now use DEFAULT or the
  numeral equivalent (as we do in the 5.6 tests).
mysql-test/t/key_cache.test:
  Can't drop default key case is an error now, not a warning, for compatibility
  with 5.6.
mysql-test/t/variables.test:
  Can't drop default key case is an error now, not a warning, for compatibility
  with 5.6.
sql/mysqld.cc:
  0 is a legal (albeit magic) value: "drop key cache."
sql/set_var.cc:
  bound_unsigned() can go now, it was just a kludge until things are done
  The Right Way, which they are now.
  
  Can't drop default key case is an error now, not a warning, for compatibility
  with 5.6.
tests/mysql_client_test.c:
  Don't use -1 integer wrap around. It used to work, but now we do what's
  actually in the documentation. In tests, we now use DEFAULT or the
  numeral equivalent (as we do in the 5.6 tests).

EXCEPTIONS-CLIENT from all the places.

If mysqltest dies, mtr waits to see if mysqld dies too within 100ms
But in that case, it should not care about expected crash
Fix: jump past the code that checks the expect file

Follow-up discussed with Reporter:
Avoid hard shutdown after test failure, if caused by server log warning
  AND we are running valgrind
More general pick-up of valgrind summaries, order may apparently vary
Do exit(1) if we did find valgrind summary warnings