1. 31 Jul, 2007 1 commit
    • unknown's avatar
      Bug #29419 "Specifying a join_buffer > 4GB on 64 bit machines not possible." · 0d7461d6
      unknown authored
        
      Use size_t instead of uint when calculating join buffer size, because uint can be overflown on 64-bit platforms and join_buffer_size > 4 GB.
        
      The test case for this bug is a part of the test suite for bug #5731.
      
      
      sql/sql_select.cc:
        Use size_t instead of uint when calculating join buffer size, because uint can be overflown on 64-bit platforms and join_buffer_size > 4G.
      0d7461d6
  2. 30 Jul, 2007 1 commit
    • unknown's avatar
      mysqld_safe.sh: · 2b255602
      unknown authored
        Post-review fix, if 'logger' can't be found, and --syslog is requested, exit with error message instead of fall back to logging to error file.
      
      
      scripts/mysqld_safe.sh:
        Post-review fix, if 'logger' can't be found, and --syslog is requested, exit with error message instead of fall back to logging to error file.
      2b255602
  3. 28 Jul, 2007 1 commit
  4. 27 Jul, 2007 9 commits
    • unknown's avatar
      Merge ramayana.hindu.god:/home/tsmith/m/bk/maint/51-b29992 · 4cdc0566
      unknown authored
      into  ramayana.hindu.god:/home/tsmith/m/bk/maint/51
      
      4cdc0566
    • unknown's avatar
      Bug #29992: syslog error logging does not flush · 1c528e94
      unknown authored
      Don't use syslog by default; user will have to request it explicitly with the --syslog option.
      
      Use "sed -u" to get unbuffered output from sed, if it's supported.
      
      Otherwise, don't use sed at all - don't strip the timestamp from mysqld messages.
      
      Also, add new --syslog-tag=FOO option, which adds "-FOO" to the tag used when logging messages to syslog (i.e., mysqld-FOO or mysqld_safe-FOO)
      
      Also, explicitly mention where log messages are going, so user can more easily find them.
      
      Also, check if 'logger' is in the PATH, and log to the error log file if it can't be found.
      
      1c528e94
    • unknown's avatar
      Merge bk-internal.mysql.com:/home/bk/mysql-5.1-maint · 855159e1
      unknown authored
      into  alf.(none):/src/bug30103/my51-bug30103
      
      855159e1
    • unknown's avatar
      Bug#30103 CMake build solution does not remove auto-generated resources · 63a346f1
      unknown authored
      - Add CMake rule to remove files.
      - Manual merge.
      
      
      sql/CMakeLists.txt:
        Bug#30103 CMake build solution does not remove auto-generated resources
        - Add CMake rule to remove files.
      63a346f1
    • unknown's avatar
      Merge alf.(none):/src/bug30103/my50-bug30103 · 497b9e45
      unknown authored
      into  alf.(none):/src/bug30103/my51-bug30103
      
      
      BitKeeper/deleted/.del-CMakeLists.txt~3:
        Auto merged
      497b9e45
    • unknown's avatar
      Merge damien-katzs-computer.local:/Users/dkatz/50_hpux · 1c477824
      unknown authored
      into  damien-katzs-computer.local:/Users/dkatz/mysql51
      
      
      sql/mysqld.cc:
        Auto merged
      BitKeeper/deleted/.del-yassl.m4~e55e55c1e863abaf:
        Auto merged
      1c477824
    • unknown's avatar
      Merge rkalimullin@bk-internal.mysql.com:/home/bk/mysql-5.1-maint · 135e1926
      unknown authored
      into  mysql.com:/home/ram/work/b29980/b29980.5.1
      
      
      sql/sql_table.cc:
        Auto merged
      135e1926
    • unknown's avatar
      Bug#30103 CMake build solution does not remove auto-generated resources · 5485f65b
      unknown authored
      - Add CMake rule to remove files.
      
      5485f65b
    • unknown's avatar
      Bug #25061 Build failure on HP/UX similar to BUG#20507 · 5db069c4
      unknown authored
      Added libCsup as a mysql library dependency on hpux11.
      
      This means any projects statically linking to our libraries using the HPUX11 compiler will need to add the -lCSup option to their build link options. If they use the mysql_config tool this will happen automatically.
      
      Projects that dynamically link to libmysqlclient will continue to work without any modifications.
      
      Library libCsup is a C++ runtime library needed for yassl support when built with the hpux11 C/C++ compiler. The first attempts to fix this bug were to remove the yassl dependencies on libCsup. We tried removing all pure virtual functions and other hacks, but the dependency remained. The only other options left involve extensive restructuring of the yassl library.
          
      
      
      config/ac-macros/yassl.m4:
        Added libCsup as a mysql library dependency on hpux11 when building yassl.
      5db069c4
  5. 26 Jul, 2007 1 commit
    • unknown's avatar
      Fix for bug #29980: 5.1.20 ate my table · 330399b6
      unknown authored
      Problem: trying to repair an old (e.g. with "old" varstring fields) corrupted 
      table with use_frm option we don't actually repair the table, just altering it
      which may couse data loss.
      
      Fix: if use_frm repair option is set, do repair instead of altering even 
      if the table needs upgrade.
      
      
      sql/sql_table.cc:
        Fix for bug #29980: 5.1.20 ate my table
      330399b6
  6. 25 Jul, 2007 10 commits
    • unknown's avatar
      tmp_table_size is not about user-created temporary tables, only · 3190cf37
      unknown authored
      internal ones (like those of GROUP BY): fixing the --help text.
      
      
      sql/mysqld.cc:
        tmp_table_size is not about user-created temporary tables, only
        internal ones (like those of GROUP BY)
      3190cf37
    • unknown's avatar
      lock0iter.c: · f33238ff
      unknown authored
        remove unused variable "bit_no"
      
      
      storage/innobase/lock/lock0iter.c:
        remove unused variable "bit_no"
      f33238ff
    • unknown's avatar
      Merge tsmith@bk-internal.mysql.com:/home/bk/mysql-5.1-maint · 57a7aeeb
      unknown authored
      into  ramayana.hindu.god:/home/tsmith/m/bk/inno/jul24/51
      
      57a7aeeb
    • unknown's avatar
      Apply snapshot innodb-51-ss1644: Part #2. · a0047baa
      unknown authored
      I forgot to check in some new files: lock0priv.h lock0priv.ic lock0iter.c lock0iter.h
      
      
      storage/innobase/include/lock0iter.h:
        BitKeeper file /home/tsmith/m/bk/inno/jul24/51/storage/innobase/include/lock0iter.h
      storage/innobase/include/lock0priv.h:
        BitKeeper file /home/tsmith/m/bk/inno/jul24/51/storage/innobase/include/lock0priv.h
      storage/innobase/include/lock0priv.ic:
        BitKeeper file /home/tsmith/m/bk/inno/jul24/51/storage/innobase/include/lock0priv.ic
      storage/innobase/lock/lock0iter.c:
        BitKeeper file /home/tsmith/m/bk/inno/jul24/51/storage/innobase/lock/lock0iter.c
      a0047baa
    • unknown's avatar
      Merge bk-internal.mysql.com:/home/bk/mysql-5.1-maint · 7f18fe8d
      unknown authored
      into  amd64.(none):/src/bug24732/my51-bug24732
      
      7f18fe8d
    • unknown's avatar
      Bug#24732 Executables do not include Vista manifests · 76f1a4fd
      unknown authored
      - Required manual merge.
      
      
      CMakeLists.txt:
        Bug#24732 Executables do not include Vista manifests
        - Search for additional tools necessary to embed, catalog and sign
        targets.
      win/README:
        Bug#24732 Executables do not include Vista manifests
        - Add internal only note to EMBED_MANIFESTS option.
      76f1a4fd
    • unknown's avatar
      Merge amd64.(none):/src/bug24732/my50-bug24732 · 5a7ec278
      unknown authored
      into  amd64.(none):/src/bug24732/my51-bug24732
      
      
      BitKeeper/etc/ignore:
        auto-union
      BitKeeper/deleted/.del-CMakeLists.txt~1:
        Auto merged
      BitKeeper/deleted/.del-README~1:
        Auto merged
      5a7ec278
    • unknown's avatar
      Bug#24732 Executables do not include Vista manifests · c2f41696
      unknown authored
      - Sign executables with MySQL AB security certificate.
      
      
      BitKeeper/etc/ignore:
        Bug#24732 Executables do not include Vista manifests
        - Ignore security catalog descriptions
      CMakeLists.txt:
        Bug#24732 Executables do not include Vista manifests
        - Search for additional tools necessary to embed, catalog and sign
        targets.
      win/README:
        Bug#24732 Executables do not include Vista manifests
        - Add internal only note to EMBED_MANIFESTS option.
      win/create_manifest.js:
        Bug#24732 Executables do not include Vista manifests
        - Added publicKeyToken attribute to manifest.
      win/mysql_manifest.cmake:
        Bug#24732 Executables do not include Vista manifests
        - Add additional commands to create security catalog and sign 
        targets.
        - Add parameters to add appropiate hash attribute to manifest
        and create security content description of the security catalog.
      c2f41696
    • unknown's avatar
      Apply snapshot innodb-51-ss1644 · 2d5ecc40
      unknown authored
      Fixes:
      - Bug #23710: crash_commit_before fails if innodb_file_per_table=1
      - Bug #28254: innodb crash if shutdown during innodb_table_monitor is running
      - Bug #28604: innodb_force_recovery restricts data dump
      - Bug #29097: fsp_get_available_space_in_free_extents() is capped at 4TB
      - Bug #29155: Innodb "Parallel recovery" is not prevented
      
      
      storage/innobase/Makefile.am:
        Apply snapshot innodb-51-ss1644
        
        Revision r1632:
        Add include/lock0priv.h to noinst_HEADERS in Makefile.am.
        
        
        
        Revision r1636:
        Move lock_get_type() from lock/lock0lock.c to include/lock0priv.ic:
        * lock0lock.c: remove lock_get_type() and include include/lock0priv.ic
        * lock0priv.h: include lock0priv.ic and add lock_get_type() prototype
        * Makefile.am: add lock0priv.ic to noinst_HEADERS
        * lock0priv.ic: introduce this new file containing the body of
          lock_get_type()
        
        This move is necessary in order to use lock_get_type() from other lock/
        source files (it's going to be used in lock/lock0iter.c).
        
        Approved by:	Heikki
        
        
        Revision r1638:
        Introduce a lock queue iterator for easy (and opaque) traversing of lock
        queues. Supports table and record lock queues via the same interface.
        
        There is only "get previous" method because currently there is no need
        for "get next" - it would be unused. Feel free to add one if needed.
        
        Approved by:	Heikki
      storage/innobase/buf/buf0buf.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1624:
        Fix change missed as part of Bug 15815. Use a function to check if a block
        needs to be made younger.
        
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/buf/buf0lru.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/fsp/fsp0fsp.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1605:
        Fix Bug#29097 "fsp_get_available_space_in_free_extents() is capped at 4TB"
        by typecasting the variables before multiplying them, so that the result of
        the multiplication is of type "unsigned long long".
        
        I verified this fix by creating a sparse file of 6TB and forcing InnoDB to
        use it without overwriting it with zeroes (by commenting the code that
        overwrites :newraw files).
        
        New type ullint is introduced with the sole purpose of shortening
        "unsigned long long", please do not define it to something else than
        "unsigned long long".
        
        Approved by:	Heikki
      storage/innobase/handler/ha_innodb.cc:
        Apply snapshot innodb-51-ss1644
        
        Revision r1605:
        Fix Bug#29097 "fsp_get_available_space_in_free_extents() is capped at 4TB"
        by typecasting the variables before multiplying them, so that the result of
        the multiplication is of type "unsigned long long".
        
        I verified this fix by creating a sparse file of 6TB and forcing InnoDB to
        use it without overwriting it with zeroes (by commenting the code that
        overwrites :newraw files).
        
        New type ullint is introduced with the sole purpose of shortening
        "unsigned long long", please do not define it to something else than
        "unsigned long long".
        
        Approved by:	Heikki
        
        
        Revision r1573:
        create_table_def(): Eliminate the inline function call to dict_table_is_comp()
        that was introduced in r1571.  Inlining is disabled in ha_innodb.cc.
        
        
        Revision r1574:
        innodb_check_for_record_too_big_error(): Divide the return value of
        page_get_free_space_of_empty_noninline() by 2.  Until r1571, that function
        did not return the same value as page_get_free_space_of_empty().
        
        
        Revision r1571:
        Fix a severe bug that was introduced in r1422 when fixing Bug 21101.
        When creating an index containing a too long record, InnoDB would
        dereference a NULL pointer when trying to determine the maximum row length.
        
        innodb_check_for_record_too_big_error(): Replace the dict_table_t*
        parameter with a Boolean flag.  There is not always a dict_table_t object
        when this function is called.
        
        page_get_free_space_of_empty_noninline(): Move the definition and
        declaration from row0mysql (!) to page0page.  Make the signature
        identical with page_get_free_space_of_empty().
        
        create_clustered_index_when_no_primary(): Add the parameter "comp".
        Remove unnecessary casts.
      storage/innobase/include/buf0buf.ic:
        Apply snapshot innodb-51-ss1644
        
        Revision r1624:
        Fix change missed as part of Bug 15815. Use a function to check if a block
        needs to be made younger.
      storage/innobase/include/fsp0fsp.h:
        Apply snapshot innodb-51-ss1644
        
        Revision r1605:
        Fix Bug#29097 "fsp_get_available_space_in_free_extents() is capped at 4TB"
        by typecasting the variables before multiplying them, so that the result of
        the multiplication is of type "unsigned long long".
        
        I verified this fix by creating a sparse file of 6TB and forcing InnoDB to
        use it without overwriting it with zeroes (by commenting the code that
        overwrites :newraw files).
        
        New type ullint is introduced with the sole purpose of shortening
        "unsigned long long", please do not define it to something else than
        "unsigned long long".
        
        Approved by:	Heikki
      storage/innobase/include/lock0lock.h:
        Apply snapshot innodb-51-ss1644
        
        Revision r1623:
        Fix typo in comment.
        
        
        Revision r1628:
        lock_has_to_wait() is needed in the INFORMATION_SCHEMA implementation
        in order to determine which lock is blocking which. Make it non-static
        and put its definition in include/lock0lock.h.
        
        Approved by:	Heikki (via IM)
      storage/innobase/include/mem0mem.ic:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/include/page0page.h:
        Apply snapshot innodb-51-ss1644
        
        Revision r1571:
        Fix a severe bug that was introduced in r1422 when fixing Bug 21101.
        When creating an index containing a too long record, InnoDB would
        dereference a NULL pointer when trying to determine the maximum row length.
        
        innodb_check_for_record_too_big_error(): Replace the dict_table_t*
        parameter with a Boolean flag.  There is not always a dict_table_t object
        when this function is called.
        
        page_get_free_space_of_empty_noninline(): Move the definition and
        declaration from row0mysql (!) to page0page.  Make the signature
        identical with page_get_free_space_of_empty().
        
        create_clustered_index_when_no_primary(): Add the parameter "comp".
        Remove unnecessary casts.
      storage/innobase/include/row0mysql.h:
        Apply snapshot innodb-51-ss1644
        
        Revision r1571:
        Fix a severe bug that was introduced in r1422 when fixing Bug 21101.
        When creating an index containing a too long record, InnoDB would
        dereference a NULL pointer when trying to determine the maximum row length.
        
        innodb_check_for_record_too_big_error(): Replace the dict_table_t*
        parameter with a Boolean flag.  There is not always a dict_table_t object
        when this function is called.
        
        page_get_free_space_of_empty_noninline(): Move the definition and
        declaration from row0mysql (!) to page0page.  Make the signature
        identical with page_get_free_space_of_empty().
        
        create_clustered_index_when_no_primary(): Add the parameter "comp".
        Remove unnecessary casts.
      storage/innobase/include/univ.i:
        Apply snapshot innodb-51-ss1644
        
        Revision r1605:
        Fix Bug#29097 "fsp_get_available_space_in_free_extents() is capped at 4TB"
        by typecasting the variables before multiplying them, so that the result of
        the multiplication is of type "unsigned long long".
        
        I verified this fix by creating a sparse file of 6TB and forcing InnoDB to
        use it without overwriting it with zeroes (by commenting the code that
        overwrites :newraw files).
        
        New type ullint is introduced with the sole purpose of shortening
        "unsigned long long", please do not define it to something else than
        "unsigned long long".
        
        Approved by:	Heikki
        
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/lock/lock0lock.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1631:
        Move lock_rec_find_set_bit() and lock_rec_get_prev() from
        lock/lock0lock.c to include/lock0priv.h and make them non-static.
        They will be used in lock/lock0iter.c.
        
        Approved by:	Heikki
        
        
        Revision r1636:
        Move lock_get_type() from lock/lock0lock.c to include/lock0priv.ic:
        * lock0lock.c: remove lock_get_type() and include include/lock0priv.ic
        * lock0priv.h: include lock0priv.ic and add lock_get_type() prototype
        * Makefile.am: add lock0priv.ic to noinst_HEADERS
        * lock0priv.ic: introduce this new file containing the body of
          lock_get_type()
        
        This move is necessary in order to use lock_get_type() from other lock/
        source files (it's going to be used in lock/lock0iter.c).
        
        Approved by:	Heikki
        
        
        Revision r1628:
        lock_has_to_wait() is needed in the INFORMATION_SCHEMA implementation
        in order to determine which lock is blocking which. Make it non-static
        and put its definition in include/lock0lock.h.
        
        Approved by:	Heikki (via IM)
        
        
        Revision r1629:
        Add "const" qualifiers to lock_get_type() and lock_get_mode().
        
        Approved by:	Sunny
        
        
        Revision r1626:
        Move lock_*struct structures from lock/lock0lock.c to include/lock0priv.h.
        
        This is needed in order to add more code to lock/ that uses members of
        these structures (internal to the lock module) but in a separate file,
        rather than lock0lock.c. lock0lock.c is a way too big already.
        
        Approved by:	Sunny
      storage/innobase/log/log0recv.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1607:
        Bug#23710
        
        At InnoDB startup consider the case where log scan went beyond checkpoint_lsn as a crash and initiate crash recovery code path.
        
        reviewed by: Heikki
      storage/innobase/mem/mem0mem.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/mem/mem0pool.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/os/os0file.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1613:
        Fix Bug#29155 by enabling file locking on FreeBSD.
        It has been disabled because InnoDB has refused to start on
        FreeBSD & LinuxThreads, but now it starts just fine.
        
        Approved by:	Heikki
      storage/innobase/page/page0page.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1571:
        Fix a severe bug that was introduced in r1422 when fixing Bug 21101.
        When creating an index containing a too long record, InnoDB would
        dereference a NULL pointer when trying to determine the maximum row length.
        
        innodb_check_for_record_too_big_error(): Replace the dict_table_t*
        parameter with a Boolean flag.  There is not always a dict_table_t object
        when this function is called.
        
        page_get_free_space_of_empty_noninline(): Move the definition and
        declaration from row0mysql (!) to page0page.  Make the signature
        identical with page_get_free_space_of_empty().
        
        create_clustered_index_when_no_primary(): Add the parameter "comp".
        Remove unnecessary casts.
      storage/innobase/rem/rem0rec.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/row/row0mysql.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1571:
        Fix a severe bug that was introduced in r1422 when fixing Bug 21101.
        When creating an index containing a too long record, InnoDB would
        dereference a NULL pointer when trying to determine the maximum row length.
        
        innodb_check_for_record_too_big_error(): Replace the dict_table_t*
        parameter with a Boolean flag.  There is not always a dict_table_t object
        when this function is called.
        
        page_get_free_space_of_empty_noninline(): Move the definition and
        declaration from row0mysql (!) to page0page.  Make the signature
        identical with page_get_free_space_of_empty().
        
        create_clustered_index_when_no_primary(): Add the parameter "comp".
        Remove unnecessary casts.
      storage/innobase/sync/sync0rw.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1598:
        Add some comments.
        
        Approved by:	Heikki (via IM)
      storage/innobase/sync/sync0sync.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1598:
        Add some comments.
        
        Approved by:	Heikki (via IM)
      storage/innobase/trx/trx0sys.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      storage/innobase/trx/trx0trx.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1595:
        trx_commit_for_mysql(): Avoid acquiring and releasing kernel_mutex when
        trx->sess or trx_dummy_sess is non-NULL.
      storage/innobase/ut/ut0mem.c:
        Apply snapshot innodb-51-ss1644
        
        Revision r1581:
        Port extra Valgrind instrumentation (UNIV_DEBUG_VALGRIND) from branches/zip.
      2d5ecc40
    • unknown's avatar
      Merge ramayana.hindu.god:/home/tsmith/m/bk/maint/jul24/50 · 1288616e
      unknown authored
      into  ramayana.hindu.god:/home/tsmith/m/bk/maint/jul24/51
      
      
      sql/sql_table.cc:
        Auto merged
      mysql-test/include/mix1.inc:
        Auto merged
      mysql-test/r/innodb_mysql.result:
        Use local (manual merge)
      1288616e
  7. 24 Jul, 2007 2 commits
  8. 21 Jul, 2007 1 commit
    • unknown's avatar
      Merge trift2.:/MySQL/M50/push-5.0 · 9e9f196b
      unknown authored
      into  trift2.:/MySQL/M51/push-5.1
      
      
      libmysql/libmysql.c:
        Auto merged
      scripts/make_binary_distribution.sh:
        Auto merged
      sql/field.cc:
        Auto merged
      9e9f196b
  9. 20 Jul, 2007 14 commits