[ET Trac] [Einstein Toolkit] #1294: Don't use STRICT_ANSI on stampede

[Einstein Toolkit]

#1294: Don't use STRICT_ANSI on stampede
------------------------+---------------------------------------------------
 Reporter:  knarf       |       Owner:  eschnett  
     Type:  defect      |      Status:  new       
 Priority:  minor       |   Milestone:  ET_2013_05
Component:  SimFactory  |     Version:  ET_2012_11
 Keywords:  backport    |  
------------------------+---------------------------------------------------
 The attached patch undefines U__STRICT_ANSI__ on stampede, similar to
 other machines. One effect of this is that M_PI is defined when math.h is
 included. By default, the lib otherwise follows the strict standard that
 math.h is not to pollute the namespace and doesn't define M_PI because it
 is not in the ANSI standard (but e.g. in POSIX). The ET depends on M_PI
 being defined in several thorns and a discussion on this topic took place
 on the mailing list some time ago. The consensus there was to try to make
 sure with compiler flags that POSIX is available, and with that M_PI.

 I committed this change to /trunk of simfactory already (in r2008), but
 want to have a positive review for a backport to Ørsted.

-- 
Ticket URL: <https://trac.einsteintoolkit.org/ticket/1294>
Einstein Toolkit <http://einsteintoolkit.org>
The Einstein Toolkit