Abort trap

[Markus Rexroth]

Dear all,

 

Justin Bird (in cc) tried to use LENSTOOL in inverse mode, but during burn-in, it aborts with the error

>>>Abort trap: 6.000159    20     8460.397     -0.909   0/174  -174chi2/s

 

Do you know what goes wrong there?

 

Best,

Markus

[Eric Jullo]

Dear Markus,

This is hard to say with just this error message. 

I recommend you check the input files. Maybe you can start with smaller input catalogs of strong or weak lensing. You can also check with a simpler parameter file, with fewer potentials. 

If you really don’t find the error, and it happens each time you run this job, eventually you can recompile lenstool in debug mode by changing the option -O2 by -O0 -g in the src/Makefile. 

Then, you can run lenstool with gdb, and wait until the error happens. You will then know which part of the code is producing the error.

Best

Eric

-- 
-- 
_____________________________________________________________________
LENSTOOL:  http://projets.lam.fr/projects/lenstool/wiki
To post to this group, send email to lens...@googlegroups.com
To unsubscribe from this group, send email to  lenstool-u...@googlegroups.com
For more options, visit this group at  http://groups.google.com/group/lenstool?
--- 
You received this message because you are subscribed to the Google Groups "lenstool" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lenstool+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Eric Jullo]

Dear Justin,

This look like related to the openmp library. Can you try with just one thread? Just type this command, before launching lenstool.

export OMP_NUM_THREADS=1

Also, which compiler and version do you use?

Eric

On 29 Oct 2015, at 16:56, Justin Bird <justin...@gmail.com> wrote:

Hello,

Thanks for the reply. Unfortunately, the errors I am receiving are occurring while trying to run the inverse example files. The forward processes seem to work fine. In the gdb debug mode, while running examples_with_images, the terminal aborts after a few burn in cycles:

Starting program: /Users/jbird/LENSTOOL/src/lenstool test_image.par -n
Reading symbols for shared libraries +++.............................. done
You are running openMP version of lenstool with 2 threads
You can change number of threads by set environment variable OMP_NUM_THREADS
Burn-in :   0.000042     5      148.211     -0.371   0/62  -62chi2/s           
Program received signal SIGABRT, Aborted.
0x00007fff8743c122 in __psynch_mutexwait ()

When running examples_with_galaxies, the terminal outputs this information for the aborted optimization:

OPTIMISATION :
Bayesys Rate : 0.100000
Number of contraints : 10
Number of free parameters : 6
Burn-in :   0.000151    13     2106.417     -3.414   0/460  460chi2/s          
Program received signal SIGABRT, Aborted.
0x00007fff8743c0fa in __psynch_cvwait ()

I get a similar error in example_with_arcletstat:

Starting program: /Users/jbird/LENSTOOL/src/lenstool nfw.par -n
Reading symbols for shared libraries +++.............................. done
You are running openMP version of lenstool with 2 threads
You can change number of threads by set environment variable OMP_NUM_THREADS
INFO: prepare arclet 3471/3471
Burn-in :   0.000009     3    18131.735     -0.070   0/38  38chi2/s            
Program received signal SIGABRT, Aborted.
0x00007fff8743c0fa in __psynch_cvwait ()

Do you know what could be causing these psynch errors?

Sincerely,

Justin

--

Justin Bird, Ph.D. Student

Physics Department, Drexel University

3141 Chestnut St., Philadelphia, PA 19104

Disque Bldg. 808

jb3...@drexel.edu

[Michael Hansen]

Dear Markus (and Justin)

Can you try to compile the code using something like

./configure <what extensions you use> 2>&1 | tee c.txt
make 2>&1 | tee m.txt
make install 2>&1 | tee mi.txt (Only if you actually do 'make install')
 
And then append the results from the c, m and mi.txt files here.

I am definitely no expert in neither Lenstool code nor OpenMP, but from my perspective it seems like a problem that occurred during compiling the code?

Cheers,
Michael

[Justin Bird]

Hi Eric and Michael,

It seems as if the multiple threads are causing the troubles. After exporting OMP_NUM_THREADS=1, the examples run fine. I'm using gcc to compile the v6.8 lenstool on my osx mountain lion.

Here are the results of the configure output (I haven't used a make install). I'm appending the m.txt as it is rather large:

c.txt:

checking for a BSD-compatible install... /sw/bin/ginstall -c
checking whether build environment is sane... yes
checking for a thread-safe mkdir -p... /sw/bin/gmkdir -p
checking for gawk... no
checking for mawk... no
checking for nawk... no
checking for awk... awk
checking whether make sets $(MAKE)... yes
checking whether make supports nested variables... yes
checking for icc... no
checking for gcc... gcc
checking whether the C compiler works... yes
checking for C compiler default output file name... a.out
checking for suffix of executables...
checking whether we are cross compiling... no
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
checking whether gcc understands -c and -o together... yes
checking for style of include used by make... GNU
checking dependency style of gcc... gcc3
checking for ranlib... ranlib
checking for gcc option to support OpenMP... -fopenmp
checking for sin in -lm... yes
checking for /Users/jbird/AstroLibs/cfitsio/include/fitsio.h... yes
checking for ffdkopn in -lcfitsio... yes
checking for /Users/jbird/AstroLibs/wcstools/libwcs/wcs.h... yes
checking for pix2wcs in -lwcs... yes
configure: WARNING: PGPLOT disabled
checking for /Users/jbird/AstroLibs/gsl/include/gsl/gsl_cblas.h... yes
checking for cblas_dgemm in -lgslcblas... yes
checking for gsl_blas_dgemm in -lgsl... yes
checking that generated files are newer than configure... done
configure: creating ./config.status
config.status: creating Makefile
config.status: creating src/Makefile
config.status: creating liblt/Makefile
config.status: creating table_src/Makefile
config.status: creating utils/Makefile
config.status: executing depfiles commands

Best,

Justin

[Michael Hansen]

Hi Justin,

Well, both the configure and make log seems to be in order, but the compilation usually also aborts in case of errors.
So the question is whether the problem lies within the OpenMP compilation. You could try to recompile the OpenMP library.
Another way of testing this would be for you to send the necessary files to run your model and then I can test it on my machine.

What versions of WCSTools, CFITSIO and GSL are you using?

Cheers,
Michael