Hi,
This was fixed as of 10-17 (see commit text below). You're right
Scott, the '<' operator needed an explicit const (I guess intel/gnu
compilers promoted it automatically or something?).
-Dan
commit 1bc44eca0cfd480bfc2a2f83bbcfecd9768b2f68
Author: Daniel R. Roe <Daniel.R.Roe.gmail.com>
Date: Wed Oct 17 13:14:41 2012 -0600
DRR - Cpptraj: Fix to ClusterNode for PGI compilers. Add PGI
option to standalone configure.
On Wed, Nov 7, 2012 at 11:20 AM, Scott Brozell <sbrozell.rci.rutgers.edu> wrote:
> Hi,
>
> I started to investigate this last month but got distracted...
> Dan have you investigated this ?
> See below for where i left it.
>
> On Thu, Oct 11, 2012 at 07:50:56AM -0400, David A Case wrote:
>> On Wed, Oct 10, 2012, Sajeewa Pemasinghe wrote:
>> >
>> > 1 error detected in the compilation of "ClusterList.cpp".
>> > make[3]: *** [ClusterList.o] Error 2
>> > make[3]: Leaving directory
>>
>> FWIW, I'm having the same problems with pgCC and cpptraj, latest version.
>> You should consider changing compilers, or skipping cpptraj compilation.
>> C++ compiler error messages can be tremendously opaque.
>
> But as with most compiler error messages, start from the first error emission:
>
> "/wsu/usr/pgi/linux86-64_2.6/11.5/include/CC/stl/_function_base.h", line 76: error:
> no operator "<" matches these operands
> operand types are: const ClusterNode < const ClusterNode
> bool operator()(const _Tp& __x, const _Tp& __y) const { return __x < __y; }
> ^
> detected during:
> ...
>
> So operator< that takes two const ClusterNode's is missing.
> Surprise, surprise, the compiler seems to be right on that account as
> shown by grep 'operator<'
> ./ClusterNode.cpp:41:bool ClusterNode::operator<(const ClusterNode& rhs) {
> ./ClusterNode.h:12: bool operator<(const ClusterNode&);
> ./DataSet.h:133: bool operator<(const DataSet& rhs) {
> ./Action_ClusterDihedral.h:21: bool operator<(const DCnode& rhs) const { return (count_ > rhs.count_); }
>
> And it looks like merely a missing const on the operator< member function.
> But i never tested that to nail it down...
>
> scott
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
--
-------------------------
Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 201
Salt Lake City, UT 84112-5820
http://home.chpc.utah.edu/~cheatham/
(801) 587-9652
(801) 585-9119 (Fax)
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed Nov 07 2012 - 11:00:03 PST
