- moved helper functions from Bootstrap script to logger.pm
Build-tools/Bootstrap:
- renamed logfile -> LOGFILE
- moved functions abort() and run_command() to logger.pm for better
reusability
Build-tools/Do-pkg:
- heavily reworked: now make use of the logger.pm helper functions
- added new options: --dry-run, --log, --mail, --verbose
Build-tools/logger.pm:
- added functions run_command() and abort() from Bootstrap script for
better code reusability
support-files/MacOSX/postinstall.sh:
- made script more robust
2. Test for corrupted .frm and the SHOW TABLE STATUS
sql/sql_show.cc:
Fix for the crash in debug mode when the .frm is corrupted and SHOW TABLE STATUS is executed
This also fixes to return the error message to comment field in SHOW TABLE STATUS when the .frm is corrupted
tests/client_test.c:
test for corrupted .frm and the SHOW TABLE STATUS
Print correct OS error number also on Windows if we run out of memory
innobase/ut/ut0mem.c:
Print correct OS error number also on Windows if we run out of memory
Correction for min/max optimization of queries with outer joins
sql/opt_sum.cc:
Correction for min/max optimization of queries with outer joins
BitKeeper/etc/logging_ok:
Logging to logging@openlogging.org accepted
fixed error handling in HANDLER ... READ FIRST
sql/sql_handler.cc:
fix error handling in HANDLER ... READ FIRST
sql/sql_table.cc:
ALTER TABLE closes all open HANDLER's for the table in question
Increased some tests' robustness a bit :
- Added RESET MASTER to all tests who do SHOW BINLOG EVENTS, to make them independent from previous tests.
- Added save_master_pos to these tests to wait for the slave to start before purging its binary logs.
mysql-test/r/rpl_log.result:
Updated SHOW BINLOG EVENTS's positions as Monty lengthened std_data/words.dat by 500 bytes.
mysql-test/r/rpl_temporary.result:
Added RESET MASTER to make positions shown by SHOW BINLOG EVENTS independent of any previous test.
mysql-test/r/rpl_user_variables.result:
Added RESET MASTER to make positions shown by SHOW BINLOG EVENTS independent of any previous test.
mysql-test/t/rpl_log.test:
Added save_master_pos to wait for the slave to have started before purging its binary logs.
Otherwise, we could purge before the slave finishes starting, and then the slave could log events implicitly provoked by previous tests.
mysql-test/t/rpl_temporary.test:
Added RESET MASTER to make positions shown by SHOW BINLOG EVENTS independent of any previous test.
mysql-test/t/rpl_user_variables.test:
Added RESET MASTER to make positions shown by SHOW BINLOG EVENTS independent of any previous test.
BitKeeper/etc/logging_ok:
Logging to logging@openlogging.org accepted
Add diagnostic prints if insert buffer merge is tried to a page whose type is not an index page, try to recover from the situation by discarding the insert buffer records
innobase/ibuf/ibuf0ibuf.c:
Add diagnostic prints if insert buffer merge is tried to a page whose type is not an index page, try to recover from the situation by discarding the insert buffer records
configure.in - there are 32bit installations of HPUX11 out there
and compiling for these by manually adding +DAportable to CFLAGS
did not work. Better add +DD64 manually, if you need it.
strings/strto.c:
Windows won't compile for a wrong argument passing, expects a charset pointer
sql/filesort.cc:
Remove unused variable
sql/lex.h:
Change POINT->POINT_SYM due to redefination error from windef.h
sql/mysqld.cc:
Fix Embedded Server related code
sql/protocol.cc:
Remove unused variables from Embedded Server
sql/sql_class.h:
Remove silly usage
sql/sql_derived.cc:
Windows won't let you compile for this kind of assignments
sql/sql_help.cc:
Need a cast to make it compile
sql/sql_parse.cc:
Remove unused variables from Embedded server
sql/sql_yacc.yy:
Change POINT->POINT_SYM due to redefination error from windef.h
mysys/mf_keycache.c:
Add a correct cast and remove the unused variables
libmysqld/lib_sql.cc:
Remove unused variable
libmysqld/libmysqld.c:
Port the netware change to remove mysql_once_init to libmysqld
SHOW CHARACTER SET now displays "1" in mb_maxlen even if a charset is not loaded yet
mysys/charset.c:
SHOW CHARACTER SET now displays "1" in mb_maxlen even if a charset is not loaded yet
