Hi Moises,
Thanks for the response. I have two issues now.
1. Installation with CUDA support.
Our system runs on CentOS 7 and has CUDA 7.5.18. As given here https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1512&L=FSL&F=&S=&P=180733 , I exported the value of the necessary variables and tried to build again. Unfortunately, I couldn't build fdt and the log file is as follows:
== 2016-01-05 11:01:28,884 runpy.EB_FSL INFO Using ......../FSL/5.0.9/fsl/ as start dir
== 2016-01-05 11:01:28,884 runpy.EB_FSL INFO configuring...
== 2016-01-05 11:01:28,885 runpy.EB_FSL INFO Starting configure step
== 2016-01-05 11:01:28,886 runpy.EB_FSL INFO Running method configure_step part of step configure
== 2016-01-05 11:01:28,886 runpy.environment INFO Environment variable FSLDIR set to ..../FSL/5.0.9/fsl/
== 2016-01-05 11:01:28,978 runpy.run INFO cmd ". .../FSL/5.0.9/fsl//etc/fslconf/fslmachtype.sh" exited with exitcode 0 and output:
linux_64-gcc4.9
== 2016-01-05 11:01:29,005 runpy.EB_FSL INFO building...
== 2016-01-05 11:01:29,005 runpy.EB_FSL INFO Starting build step
== 2016-01-05 11:01:29,008 runpy.EB_FSL INFO Running method build_step part of step build
== 2016-01-05 11:20:46,749 runpy.run INFO cmd ". .../FSL/5.0.9/fsl//etc/fslconf/fsl.sh && ./build" exited with exitcode 0 and output:
Building projects - see build.log file for progress...
Finished build : end of log file shows ...
!!ERROR in BUILD!!
Could not make the following projects successfully:
fat
Could you please help me with this issue?
2. bedpostx_gpu for SLURM
According to this page --> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=fsl;5e9b51c1.1510 , I copied the patch files to the $FSLDIR/bin/ directory after the basic FSL installation without CUDA support. I made the necessary changes for SLURM in the bedpostx_gpu script.
As the same bedpostx_gpu lies in fsl/src/fdt/CUDA directory, I guess I could copy the bedpostx_gpu patch to fsl/src/fdt/CUDA directory and do the build from scratch. Am I right?
Lekshmi
|