Check for big-test, ndb-extra, require_manager and binlog format row/stmt 

Change 'mtr_kill_process' to only take three arguments. The last indicating max number of seconds to wait for process to be killed.
Leave it to the implementation of 'mtr_kill_process' how long to wait between each attempt to kill/detect if process has been killed.

Add comment why test is skipped

Fix merge problem

Cacluate nodeid and ndb_fs path when initializing ndbds, to be used when defining log files and when removing the ndb_fs after failed test case

Use function 'stop_all_servers' to stop the running servers before restoring snapshot and resuming tests

into  neptunus.(none):/home/msvensson/mysql/mysql-5.1-new-maint

into  mysql.com:/home/mydev/mysql-5.1-amerge

into  rurik.mysql.com:/home/igor/mysql-5.1

into  rurik.mysql.com:/home/igor/dev/mysql-5.1-0

into mysql.com:/home/psergey/mysql-5.1-bug19684

into  mysql.com:/home/mydev/mysql-5.1-amerge

into  mysql.com:/home/mydev/mysql-5.1-amerge

into  mysql.com:/home/mydev/mysql-5.1-amerge

into  rurik.mysql.com:/home/igor/dev/mysql-5.1-0

into  rurik.mysql.com:/home/igor/dev/mysql-5.1-0

mysql-test/mysql-test-run.pl  :  Output the usage error if one is given. (Found when fixing bug#20303)

mysql-test/mysql-test-run.pl  :    A fix for bug#20303 "mysql-test-run.pl: Does not recognize -- argument"

into  neptunus.(none):/home/msvensson/mysql/mysql-5.1-new-maint

Changed error to be more descriptive when you are refering to a not existing key
Fixed core dump in view test and changed to better error message

BUG#19544 mysqldump does not backup TS and LG information correctly.
  
Problem was the type of the column was varchar(8) not an integer type. So mysqld
was outputting a string representation of the number in scientific notation,
which wasn't exact. Now we do it as a bigint and output the number in bytes.

Fix assertion when loading plugins due to handler changes

RPL Test cleanup of options that are not needed when using ndbcluster with binlog. The binlog is forced to row just by connecting to a cluster

Add define YASSL_PREFIX when compiling yassl

Import patch from yaSSL
- avoid allocating memory for each call to 'EVP_md5' and 
  'EVP_des_ede3_cbc' which were not released until server was stopped
- Those functions are used from the SQL function 'des_encrypt' and
  'des_decrypt'.

Add new define YASSL_PREFIX beforee including ssl.h to activate inclusion of prefix_*.h files

Bug#20022 mysql-test-run can't be run with secure connections turned on for all testcases
- Part 1, fixes rpl- and federated-tests where connection is made to 127.0.0.1

- Include prefix files that renames all public functions in yaSSLs
  OpenSSL API to ya<function_name>. They will otherwise conflict
  with OpenSSL functions if loaded by an application that uses OpenSSL
  as well as libmysqlclient with yaSSL support.

Bug#18235: assertion/crash when windows mysqld is ended with ctrl-c
  
Two threads both try a shutdown sequence which creates a race to the
de-init/free of certain resources.
  
This exists in similar form in the client as 17926: "mysql.exe crashes
when ctrl-c is pressed in windows."

Update after merge to 5.0

BUG#18669: Session COM_STATISTICS breaks mysqladmin status.
Changed COM_STATISTICS to display the global status, instead of thead status, for slow queries and table opens.

- In function 'handle_grant_struct' when searching the memory structures for an 
  entry to modify, convert all entries here host.hostname is NULL to "" and compare that 
  with the host passed in argument "user_from".
- A user created with hostname "" is stored in "mysql.user" table as host="" but when loaded into 
  memory it'll be stored as host.hostname NULL. Specifiying "" as hostname means
  that "any host" can connect. Thus is's correct to turn on allow_all_hosts
  when such a user is found. 
- Review and fix other places where host.hostname may be NULL.

BUG#19394 OPT_INNODB_THREAD_CONCURRENCY duplicated
Removed duplication (not a user-visible change)

Add definition of YASSL_PREFIX to Cmake file
Change "and" to "&&"
Remove the "static" directive SunCC can't use the function

(one testcase, one memory leak and some accesses to not initialized memory)

In select_describe(), make the String object that holds the value of 
"partitions" column to "own" the value buffer, so the buffer isn't
prematurely freed.
[this is the second attempt with review fixes]