[mpich2-dev] [mpich2-maint] #262: MPICH2 use of AC_C_CONST

[mpich2]

--------------------------------------------------------+-------------------
 Reporter:  David Gingold <david.gingold at sicortex.com>  |        Owner:  chan        
     Type:  bug                                         |       Status:  new         
 Priority:  major                                       |    Milestone:  mpich2-1.1a2
Component:  mpich2                                      |   Resolution:              
 Keywords:                                              |  
--------------------------------------------------------+-------------------

Changes (by thakur):

 * cc: mpich2-dev at mcs.anl.gov (added)
  * owner:  thakur => chan

Comment:

 We are planning to fix this by defining a new macro PAC_C_CONST that
 checks if const is available and calls AC_MSG_ERROR if not.

 Rajeev


 Replying to [ticket:262 David Gingold <david.gingold at sicortex.com>]:
 >
 >
 > {{{
 >
 > Last night I chased down an ugly problem that showed up when I tried
 > to build MPICH2 with -Wall -Werror (telling gcc to turn on more
 > warnings, and to consider warnings to be errors).
 >
 > The upshot is that -Wall -Werror makes autoconf believe that the
 > compiler doesn't support "const", and then things go downhill from
 > there.  The root cause is described here:
 >
 >         http://www.nabble.com/fix-AC_C_CONST-to-work-with-CFLAGS--O2--
 Wall--
 > Werror-td17164627.html
 >
 > ...where there is some discussion about whether the problem ought to
 > be fixed in autoconf.
 >
 > Does MPICH2 in fact build on pre-C89 compilers?  If not, perhaps it
 > would make sense to drop the use of AC_C_CONST in configure.in (as a
 > poster above suggests) and to enforce that the compiler be at least C89.
 >
 > -dg
 >
 > --
 > David Gingold
 > Principal Software Engineer
 > SiCortex
 > Three Clock Tower Place, Suite 210
 > Maynard MA 01754
 > (978) 897-0214 x224
 >
 >
 >
 > }}}

-- 
Ticket URL: <https://trac.mcs.anl.gov/projects/mpich2/ticket/262#comment:3>