The test actually doesn't fail on my system. However, I don't have enough time to examine why test suite fails in beaker environment, but since it works fine when running not as beaker test, I'd suspect the test to cause this.

$ scl enable mysql55 './mysql-test-run --do-test ^mysql$'
Logging: ./mysql-test-run  --do-test ^mysql$
130812  9:32:06 [Note] Plugin 'FEDERATED' is disabled.
MySQL Version 5.5.32
Checking supported features...
 - SSL connections supported
Using suites: main,sys_vars,binlog,federated,rpl,innodb,perfschema
Collecting tests...
Checking leftover processes...
Removing old var directory...
Creating var directory '/opt/rh/mysql55/root/usr/share/mysql-test/var'...
Installing system database...
Using server port 58168

==============================================================================

TEST                                      RESULT   TIME (ms) or COMMENT
--------------------------------------------------------------------------

worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 13000..13009
main.mysql                               [ pass ]   8281
--------------------------------------------------------------------------
The servers were restarted 0 times
Spent 8.281 of 32 seconds executing testcases

Completed: All 1 tests were successful.

bash-4.1$ rpm -qa|grep mysql55-mysql
mysql55-mysql-libs-5.5.32-2.el6.x86_64
mysql55-mysql-server-5.5.32-2.el6.x86_64
mysql55-mysql-5.5.32-2.el6.x86_64
mysql55-mysql-test-5.5.32-2.el6.x86_64

I can confirm it passes on ITP system

...
main.mysql-bug45236                      [ pass ]    359
main.mysql                               [ pass ]   3292
main.mysql_comments                      [ pass ]    631
...

so, as usual, it seems something in Beaker rather than the test itself

I will try to isolate the problem

separate run:
bwt --channel 'rhel-x86_64-server-6-rhscl-1' --update --collection mysql55 --arch x86_64 --enable-beaker-repos --distro RHEL6.5-20130814.2 --task /CoreOS/mysql/testsuite
=> https://beaker.engineering.redhat.com/jobs/476171

damn, I forgot to specify the erratum where to take from the packages:

bwt --channel 'rhel-x86_64-server-6-rhscl-1' --update --collection mysql55 --arch x86_64 --enable-beaker-repos --errata 14788 --distro RHEL6.5-20130814.2 --task /CoreOS/mysql/testsuite
=> https://beaker.engineering.redhat.com/jobs/476277

cool, so we see it reproduces in a separate run, now if I would only not forgot to add --reserve to be able to investigate further :-/

bwt --channel 'rhel-x86_64-server-6-rhscl-1' --update --collection mysql55 --arch x86_64 --enable-beaker-repos --errata 14788 --distro RHEL6.5-20130820.2 --task /CoreOS/mysql/testsuite --reserve
=> https://beaker.engineering.redhat.com/jobs/477915

okay, so this fails in the beaker run and passes on the same machine/install if run from commandline

the subtest which fails is this code:

#
# Bug#26780: patch to add auto vertical output option to the cli.
#
# Make this wide enough that it will wrap almost everywhere.
--exec $MYSQL test --auto-vertical-output --table -e "SELECT 1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6,7,8,9,0;"
# Too short to wrap.
--exec $MYSQL test --auto-vertical-output --table -e "SELECT 1;"


it seems that vertical output is triggered even for the second select

but it shouldn't - reading the original patch that adds the option, it defines

unsigned short terminal_width= 80;

and then the only place where it gets changes is on window_resize event:

#if defined(HAVE_TERMIOS_H)
sig_handler window_resize(int sig)
{
  struct winsize window_size;

  if (ioctl(fileno(stdin), TIOCGWINSZ, &window_size) == 0)
    terminal_width= window_size.ws_col;
}
#endif

which should not be called

then the format of the output is controlled by this condition:

  else if (vertical || (auto_vertical_output && (terminal_width < get_result_width(result))))
 	  print_table_data_vertically(result);

as vertical defaults to 0 (this subtest does not change the setting and other results are not formatted vertically), the only way it could run print_table_data_vertically() is if "terminal_width < get_result_width(result)" would be true

I see nothing suspicious in the function:

static int get_result_width(MYSQL_RES *result)
{
  unsigned int len= 0;
  MYSQL_FIELD *field;
  MYSQL_FIELD_OFFSET offset;
  
#ifndef DBUG_OFF
  offset= mysql_field_tell(result);
  DBUG_ASSERT(offset == 0);
#else
  offset= 0;
#endif

  while ((field= mysql_fetch_field(result)) != NULL)
    len+= get_field_disp_length(field) + 3; /* plus bar, space, & final space */

  (void) mysql_field_seek(result, offset);      

  return len + 1; /* plus final bar. */
}


so we are back to the question whether terminal_width gets changed, I guess ...


but what puzzles me the most is the presence of the following error:

ERROR: Can't initialize batch_readline - may be the input source is a directory or a block device.

the error message was added to MySQL in the following patch:
http://lists.mysql.com/commits/130791

the code seems completely unrelated to the vertical output problem, yet the error appears only together with this problem (and not in the first test query for the auto-vertical-output option ...)


any hints what debug output I could add to the test so we can get closer where the problem lies?

thanks to completely unrelated problem which we faced on our server yesterday evening by coincidence, I have found that mysql abuses the SIGWINCH signal for a graceful server shutdown ... yes, that's server and we are dealing with client output here, but it may (or may not!) be a trace that window size signalling is broken in mysql as a whole ...?

Just FTR, I tried to get more into it, but without any success. It seemed that window signalling worked fine outside the test-suite. Will try once more if I have some spare time, but it doesn't seem like something important to fix.

Devel CondNAK, since we don't have any patch available. It works fine during build and outside of beaker, so we don't consider this as a big issue.

Red Hat does not currently plan to provide any further changes to this collection in a Red Hat Software Collections update release.

This software collection is nearing the retirement date (October 2016) after which customers are encouraged either to upgrade to a later release or continue on as self-supported without official Red Hat Support.

Please contact Red Hat Support if you have further questions, or refer to the support lifecycle page for more information. https://access.redhat.com/support/policy/updates/rhscl/

In accordance with the Red Hat Software Collections Product Life Cycle, the support period for this collection has ended.

New bug fix, enhancement, and security errata updates, as well as technical support services will no longer be made available for this collection.

Customers are encouraged to upgrade to a later release.

Please contact Red Hat Support if you have further questions, or refer to the support lifecycle page for more information. https://access.redhat.com/support/policy/updates/rhscl/