Memory was used after it has been alreay freed.
This led to server crash on FreeBSD.
sql/examples/ha_archive.cc:
Memory was used after it has been alreay freed.
This led to server crash on FreeBSD.
Don't wait for more SCANTAB_CONF when received SCAN_TABREF
ndb/src/ndbapi/NdbScanOperation.cpp:
Don't wait for more SCANTAB_CONF when received SCAN_TABREF
More debug
- Add sequence to the initial trans id so that create/drop ndb object
don't generate same transid
ndb/src/ndbapi/NdbConnection.cpp:
Add sequence to the initial trans id so that create/drop ndb object
don't generate same transid
ndb/src/ndbapi/Ndbif.cpp:
Add sequence to the initial trans id so that create/drop ndb object
don't generate same transid
ndb/src/ndbapi/TransporterFacade.cpp:
Add sequence to the initial trans id so that create/drop ndb object
don't generate same transid
ndb/src/ndbapi/TransporterFacade.hpp:
Add sequence to the initial trans id so that create/drop ndb object
don't generate same transid
Fix assertion failure where ha_myisam::index_next() is called
without first calling ha_myisam::ha_index_init()
BitKeeper/etc/logging_ok:
Logging to logging@openlogging.org accepted
various fixes (for mysql-test-run at least):
* accept --no-defaults
* 4.0 compatible password handling (only it no other argumensts, only if doesn't start with --)
ps_3innodb.test:
only run ps_3innodb.test if innodb is present
mysql-test-run.sh:
it's --verbose, not --verbose=1
mysql-test/mysql-test-run.sh:
it's --verbose, not --verbose=1
mysql-test/t/ps_3innodb.test:
only run ps_3innodb.test if innodb is present
scripts/mysql_fix_privilege_tables.sh:
various fixes (for mysql-test-run at least):
* accept --no-defaults
* 4.0 compatible password handling (only it no other argumensts, only if doesn't start with --)
Bug #4338 mysql-test-run fails if compiled with non-latin1 character set.
client/mysqltest.c:
Bug #4338 mysql-test-run fails if compiled with non-latin1 character set.
into mysql.com:/home/jonas/src/wl1873
ndb/include/ndbapi/NdbScanOperation.hpp:
Auto merged
ndb/src/ndbapi/NdbResultSet.cpp:
Auto merged
ndb/src/ndbapi/NdbScanOperation.cpp:
Auto merged
ndb/test/ndbapi/testScan.cpp:
Auto merged
An addition to 1.2036 04/07/02: armscii8_bin was referenced to as armscii_bin in a mistake.
mysql-test/t/rpl_charset.test:
An addition to 1.2036 04/07/02: armscii8_bin was referenced to as armscii_bin in a mistake.
mysql-test/r/rpl_charset.result:
An addition to 1.2036 04/07/02: armscii8_bin was referenced to as armscii_bin in a mistake.
Enabled ordered scan in handler
ndb/include/ndbapi/NdbIndexScanOperation.hpp:
Moved saveBound to NdbIndexScanOperation
ndb/include/ndbapi/NdbScanOperation.hpp:
Moved saveBound to NdbIndexScanOperation
ndb/src/ndbapi/NdbDictionaryImpl.cpp:
Introduced map for
index attributes (keys) -> real attr id (and back)
ndb/src/ndbapi/NdbDictionaryImpl.hpp:
Introduced map for
index attributes (keys) -> real attr id (and back)
ndb/src/ndbapi/NdbOperationDefine.cpp:
Moved saveBound to NdbIndexScanOperation
ndb/src/ndbapi/NdbOperationInt.cpp:
Moved saveBound to NdbIndexScanOperation
ndb/src/ndbapi/NdbScanOperation.cpp:
Moved saveBound to NdbIndexScanOperation
Fixed bugs in handling of setBounds w.r.t getValues and index keys
(use new reverse map)
Fixed bugs in next_result_ordered
sql/ha_ndbcluster.cc:
Use sorted scan when requested
Fix for BUG#4375 "Windows specific directories are copied during
replication": in the Windows version of my_dir(), do not show hidden
or system files which Windows sometimes creates in the root directory
of drive (like RECYCLER and SYSTEM VOLUME INFORMATION directories) (so
this problem showed up only when the datadir was the root dir of a
drive).
This will make SHOW DATABASES display better, and will do no harm to
MySQL. For example, DROP DATABASE will not miss some files, as MySQL
creates no hidden or system files.
mysys/my_lib.c:
In my_dir(), do not show hidden or system files
which Windows sometimes creates.
ndb/src/common/util/ConfigValues.cpp:
BUG#4384
Compute pointers manually instead of using ptr[-x]
Changed return type from "*&" to "**"
BitKeeper/etc/logging_ok:
Logging to logging@openlogging.org accepted
"test: argument expected" error when trying to run the NDB test
suite on Solaris
mysql-test/ndb/ndbcluster.sh:
- fixed typo that caused a "test: argument expected" error when trying
to run the NDB test suite on Solaris
mysql-test/r/ps_2myisam.result:
changes in results of select without order by
mysql-test/r/ps_3innodb.result:
changes in results of select without order by
mysql-test/r/ps_5merge.result:
changes in results of select without order by
mysql-test/r/ps_6bdb.result:
changes in results of select without order by
sql/item_cmpfunc.cc:
do not assign values of left expression of IN/ANN/ANY subquery in case of PS preparation
restoring current senect pointer before PS rexecution (backport from 5.0)
removed spaces at lines ends
mysql-test/r/subselect.result:
Aggregate function comparation with ALL/ANY/SOME subselect test
mysql-test/t/subselect.test:
Aggregate function comparation with ALL/ANY/SOME subselect test
sql/item_subselect.cc:
removed spaces at lines ends
aggregate functions check during substitution made only for single row subselects
sql/item_subselect.h:
removed spaces at lines ends
sql/sql_prepare.cc:
restoring current senect pointer before PS rexecution (backport from 5.0)
mysql-test/include/ps_create.inc:
Added comments
mysql-test/include/ps_modify.inc:
added comments
mysql-test/include/ps_modify1.inc:
Added comments
mysql-test/include/ps_query.inc:
Added and modified comments
mysql-test/t/ps_1general.test:
Description of the test structure and rules how to extend the tests added and one comment modified .
mysql-test/t/ps_2myisam.test:
Added comments
mysql-test/t/ps_3innodb.test:
added comments
mysql-test/t/ps_4heap.test:
added and altered comments
mysql-test/t/ps_5merge.test:
added comments
mysql-test/t/ps_6bdb.test:
added comments
if a command has a comment at the end of line, like:
error 2002 ; # this is error 2002
then the parsing of comment should not make mysqltest
forget about the value of expected error.
Reason it forgot it (so the next query caused the test to fail)
is that internally the above line is 2 queries.
client/mysqltest.c:
if a command has a comment at the end of line, like:
error 2002 ; # this is error 2002
then the parsing of comment should not make mysqltest
forget about the value of expected error.
