Frequently Asked Questions | The Grigorieff Lab

Forums

Frealign

Q. I get a "Segmentation Fault" when executing an older version of Frealign. What should I do?
A. You may need to change NN1, NNBIG etc. and recompile.

Q. How do I compile Frealign?
A. Use the appropriate Makefile and the make command. For example:
make -f Makefile_linux

Trace/BPT trap

Denis Mon, 2012-03-19 14:05

Edit added: Sorry for posting here, I was thinking creating a new subject. Could not find a way to cancel this message ...

When running the helical refinment example on MacPro workstations (10.7.3 or 10.6.8), I get a Trace/BTP trap error:

pr086203:helical_refinement TIPS2$ tcsh refine.com 

 SET_POLARITY v1.01, 02.07.2010

 PSI angle variation?
       5.000

 Flip polarity if ave. PRES above?
      90.000

 Input parameter file?
/Applications/frealign_v8.09/examples/helical_refinement/TMV_1.par

 Output parameter file?
../scratch/temp.par

 No. of header lines read =         53
 No. of footer lines read =          0
 No. of particles read    =        179
 No. of filaments         =         10

 Best PSI angle for filament       162 =   180 deg,  ave. PRES =   68.11, score =   1.00
 Best PSI angle for filament       163 =   213 deg,  ave. PRES =   73.40, score =   1.00
 Best PSI angle for filament       164 =   210 deg,  ave. PRES =   71.37, score =   1.00
 Best PSI angle for filament       165 =   200 deg,  ave. PRES =   71.27, score =   1.00
 Best PSI angle for filament       166 =   195 deg,  ave. PRES =   71.15, score =   1.00
 Best PSI angle for filament       167 =   194 deg,  ave. PRES =   71.12, score =   1.00
 Best PSI angle for filament       168 =     0 deg,  ave. PRES =   68.18, score =   1.00
 Best PSI angle for filament       169 =   179 deg,  ave. PRES =   68.47, score =   1.00
 Best PSI angle for filament       170 =    29 deg,  ave. PRES =   71.55, score =   1.00
 Best PSI angle for filament       171 =   252 deg,  ave. PRES =   70.64, score =   1.00
Trace/BPT trap
0.000u 0.000s 0:01.93 0.0%	0+0k 0+0io 0pf+0w
mv: rename ../scratch/TMV_2.par_1_179 to /Applications/frealign_v8.09/examples/helical_refinement/TMV_2.par: No such file or directory
Cycle 2 finished....  Lun 19 mar 2012 18:45:23 CET
Lun 19 mar 2012 18:45:23 CET

Yet, the program works fine on my (quite old) MacBook Pro with 10.6.8. Any idea ?

I have not seen this error before. It looks like it is processing all filaments and then crashes when it is trying to write out the new file temp.par. What system are you running this on? Which Makefile did you use to build the binary?

I get this error both on Mac

Denis Tue, 2012-03-20 13:44

I get this error both on Mac Os X Snow Leopard (10.6.8) and Mac Os X Lion (10.7.3), using Mac Pro Intel workstations. I'm unsure, but it may have to do with the compiler. I've installed Intel compilers (ifort and icc) and used Makefile_OSX_86-64_intel. On the Mac under Snow Leopard, I could use the frealign_v8.exe file that I generated on my MacBook Pro (also under snow Leopard) and it worked fine, but on the one under Lion this did'nt work.

ifort on OSX

Alexis Tue, 2012-03-20 14:08

Hi Denis,

Which version of ifort & Xcode are you using?

I don't use OSX myself, so may be of limited use, but:

Have a look at the ifort forum for Linux/OSX, in particular look at the sticky threads at the top, which concern OSX, Xcode etc: http://software.intel.com/en-us/forums/intel-fortran-compiler-for-linux…
to get more information about the error, you could try to rebuild frealign with debug, traceback etc. Modify the FFLAGS variable in your Makefile to include, e.g., -debug -O0 -traceback

ifort on OSX

Denis Wed, 2012-03-21 16:38

Indeed, in this forum they state not using Xcode 4.3 but stick to 4.2 with Intel compilers. But the one installed on the Powermac running Lion is 4.2.1, so that should work... I'll explore the other solution that you propose. If ever I can solve this problem, I'll keep you aware.
Thanks a lot for your help ;-)
Denis