cns 1.1 with Linux Mandrake8.1

Stanley, Matthew D. bcmatt at
Tue Feb 5 10:23:31 EST 2002


There are several methods of fixing this problem:

1. Compile with a different version of g77.  I have had success with v2.95.x without making any changes to the Makefile.

2. Use the default method of compiling CNS on linux by installing fort77 and f2c.  The problem is you need to compile f2c from scratch and not use the rpm's that are provided by Mandrake.

3. Edit the Makefile and modify the line F77OPT by changing the optimization to -O1 from -O3.

4. Edit the Makefile and modify the line F77OPT by adding the option -fno-automatic to the end after -ffast-math.

I have used each of these methods to compile CNS and have found option 4 provides the closest results even with the newer g77 compilers.  I compiled these options from the CNS web site and from emails that Kevin Cowtan sent out regarding CCP4 (many thanks Kevin!).  If you want to start the whole CNS install from scratch you can find the Makefile.header in:

Best of luck,

Matthew Stanley
Structural Biology Core
University of Missouri - Columbia

-----Original Message-----
From: Laurent Maveyraud [mailto:maveyrau at]
Sent: Tuesday, February 05, 2002 4:31 AM
To: x-plor at
Subject: cns 1.1 with Linux Mandrake8.1

Dear all,

I just installed CNS 1.1 on a Linus machine using g77, compilation went
fine. I then ran the test procedures, and all input files seem to
produce normal results, except rotman.inp which produced a coredump... 

is there anything known about this ?

thanks for your help


Laurent Maveyraud               
Groupe de Biophysique Structurale  -  IPBS-CNRS

205, route de Narbonne       	 Tel (+33) 05 61 17 54 35
31077 TOULOUSE CEDEX        	 Fax (+33) 05 61 17 54 48
France                       	 mailto:laurent.maveyraud at


More information about the X-plor mailing list