Previous patch didn't work: moving variable settings inside critical region.

into  romeo.(none):/home/bk/b20821-mysql-5.1-new-rpl

in pushbuild on Replication/Backup team tree.

into  romeo.(none):/home/bkroot/mysql-5.1-new-rpl

into  eucla.lemis.com:/home/MySQL/5.1-Bug-20850

  BUG#20850: Assert during slave shutdown in many rpl_* tests.
  This was caused by a race condition at the end of handle_slave_io
  which under some circumstances allowed the cleanup to proceed before
  the thread had completed.

  Tidy up white space.

into  mysql.com:/home/bk/b20821-mysql-5.1-new-rpl

Reverting to old behaviour of writing the query before all rows
have been written.

into  mysql.com:/users/lthalmann/bk/MERGE/mysql-5.1-merge

into  dator5.(none):/home/pappa/bug17138

into  dator5.(none):/home/pappa/bug17138

into  dator5.(none):/home/pappa/bug20583

into  dator5.(none):/home/pappa/bug17138

Last round of review fixes

into moonbone.local:/work/merge-5.1

into  mysql.com:/data0/knielsen/tmp-5.1

into  mysql.com:/usr/local/mysql/tmp-5.1

binary disctribution for Falcon.

into moonbone.local:/work/merge-5.1

go from 623 seconds to 11 seconds on AMD64.
This is because we have no native atomic implementation on AMD64, 
so the rwlock-based implementation is used, which is  a bit slow.  
It will be probably be optimized for AMD64 quite soon, but for now 
this test case will do fewer iterations.

into  mysql.com:/home/cps/mysql/trees/5.1-team

into  mysql.com:/home/my/mysql-5.1

(It failed even if test was marked as skipped)

into  poseidon.ndb.mysql.com:/home/tomas/mysql-5.1-new-ndb

into  poseidon.ndb.mysql.com:/home/tomas/mysql-5.1-new-ndb

into  poseidon.ndb.mysql.com:/home/tomas/mysql-5.0

into  mysql.com:/home/cps/mysql/trees/5.1-team