Re: [AMBER] Error with na_anal.c

From: Chirag Vora <chirag740.gmail.com>
Date: Mon, 18 Jul 2011 13:07:57 +0000

Thanks DAC.
I installed serial version earlier on 32bit of same OS (ubuntu 10.04) using 32bit counterparts of these compilers. It all went perfect that time.

As of now I've turned to this 64bit system.

Though I'm working with a CPU with dual core processor. I need MPI because I intend to perform Thermodynamic Integration (TI).

Please suggest whether I'm on correct lines or not? I know performing TI requires parallel Amber 10. Is MPI required by AmberTools too?
Chirag Vora
Dept. of Pharmacoinformatics
National Institute of Pharmaceutical Education and Research (NIPER)
Sector 67, S.A.S. Nagar (Mohali) - 160062
Punjab, INDIA

Sent from my BlackBerry® smartphone from !DEA

-----Original Message-----
From: case <case.biomaps.rutgers.edu>
Date: Mon, 18 Jul 2011 08:35:22
To: AMBER Mailing List<amber.ambermd.org>
Reply-To: AMBER Mailing List <amber.ambermd.org>
Subject: Re: [AMBER] Error with na_anal.c

On Mon, Jul 18, 2011, Chirag Vora wrote:

> I am installing AMBER 10 (parallel) on ubuntu10.04. I have installed all
> necessary libraries.
>
> Following are the steps followed:
> 1) Succesfully installed openMPI-1.4.3 to ~/Documents/openmpi-1.4.3/
>
> 2) Extracted AmberTools-1.2.tar.bz2 and Amber10.tar.bz2 in ~/Documents/
>
> 3) Applied both patches (ambertools/amber10) successfully
>
> 4) After doing cd src/ and running ./configure_at I get a following copious
> output :
>
> Testing the C compiler:
> mpicc -m64 -o testp testp.c
> OK
>
> Obtaining the C++ compiler version:
> g++ -v
> The version is ../src/configure
> 4.4.3
> [: 520: -lt: unexpected operator
> OK

[Scott: we are getting *lots* of reports of configure failing to correctly
figure out the g++ version or running its tests. Can you try to figure out
what might be going on? Maybe some users don't have "cut"? Maybe it doesn't
always work correctly? I'd like to get a more robust way of getting past this
step.]

> na_anal.c: In function ‘helixanal’:
> na_anal.c:2417: error: expected ‘)’ before ‘is’
> na_anal.c:2417: error: stray ‘\’ in program
> na_anal.c:2417:41: warning: missing terminating " character
> na_anal.c:2417: error: missing terminating " character

This has come up before:

   http://archive.ambermd.org/201007/0376.html

It seems to be some unusual conflict between our code, 64-bit compilers (of
some versions), and maybe exactly what is in 1.2 code tree.

Did this work for the serial compilation? I notice that there is a "-m64"
argument to mpicc above. If you edit config.h and remove all instances to
"-m64" does it help? (Was the -m64 there in the serial build?)

(And of course, finally: are you sure you need an MPI version of NAB? It
wouldn't make much sense to try to track down an very funny bug in an old
version of parallel AmberTools if you are not planning to use it anyway....)


...dac


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jul 18 2011 - 06:30:02 PDT
Custom Search