Remove unused libm-test expected-failure mechanism

[Joseph S. Myers]

On Sun, 5 May 2013, Carlos O'Donell wrote:

> It is my opinion that such an expected failure mechanism should be
> handled by a higher level test framework, and that libm-test.inc
> should simply report the list of all passed or failed tests given
> that each test has a unique name.

 ... or that a test was unsupported / unresolved (e.g. because setting the 
rounding mode failed).

But first we need to be able to track pass / fail information for whole 
tests, before tracking "unresolved" etc. and individual assertions.  What 
happened with Tomas Dohnalek's patch from last September / October?  It 
looks like the discussion tailed off with 
<http://sourceware.org/ml/libc-alpha/2012-10/msg00278.html>, but getting 
PASS/FAIL lines for each overall test run from the makefile ought to be an 
incremental piece of progress towards having the other statuses and more 
precise information from within individual tests; someone probably needs 
to pick up that patch or else decide that a different approach is needed.

-- 
Joseph S. Myers
joseph@codesourcery.com