into mysql.com:/home/mysql_src/mysql-4.0

into narttu.mysql.fi:/my/mysql-4.0

Fixed problem with localtime -> gmt where some times resulted in
different (but correct) timestamps. Now MySQL should use the smallest
possible timestamp value in this case.  (Bug 316)

soon.

into narttu.mysql.fi:/my/mysql-3.23

into ua126d19.elisa.omakaista.fi:/home/my/bk/mysql-4.0

into narttu.mysql.fi:/my/mysql-4.0

Added read_only variable

into mashka.mysql.fi:/home/my/mysql-3.23

- A few more mutex locks and unlocks of rli.log_space_lock for doing clean reads of
rli.ignore_log_space_limit
- Broadcast after unlock, not before (small speed optimisation).

  If FOREIGN_KEY_CHECKS=0, wrap in binlog SQL statements inside SET FOREIGN_...=0; ... ; SET FOREIGN_...=1

into mysql.com:/home/mysql_src/mysql-4.0

(this bug was already fixed in 4.0, I just copied and pasted two lines).

   now have a thread safe realpath(3) implementation - added check to only
   define -DHAVE_BROKEN_REALPATH where required
   (thanks to Martin Blapp from the FreeBSD team for the patch)

  Disable UNIQU KEY error reporting in SHOW INNODB STATUS until we know if it slows down REPLACE significantly

into eagle.mysql.r18.ru:/home/vva/work/LOADDATA/mysql-3.23

into eagle.mysql.r18.ru:/home/vva/work/LOADDATA/mysql-4.0

 - Updated version number in configure.in to 4.0.14 now that 4.0.13 has
   been tagged 

into narttu.mysql.fi:/my/mysql-4.0

  Better instructions for the user if adding of new ibdata files fails

  When calculating the buf pool dirty pages ratio, add also free pages to the LRU list length: no need to active flushing if there are lots of free pages in the buffer pool

  Search first only 10 % of the LRU list for a replaceable block before doing an LRU flush; enable again flushing of close pages also in a flush list (checkpointing) flush
trx0trx.c:
  Add forgotten return value (it was not used anywhere, fortunately)
ha_innodb.h, mysql_priv.h:
  Move declaration of srv_buf_pool_max_modified_pct to ha_innodb.h and enclose it to denote it is a C variable, not C++

  - QNX 6.2.1 actually defines HAVE_RINT
  - fixed rint(A) #define as it was causing compile errors on QNX