build_libsmm generate small1

Alfio Lazzaro alaz... at cray.com
Fri Sep 13 09:45:48 CEST 2013


Hi,
is the tiny1 phase generating the correct output files:

tiny_gen_optimal_dnn_linux.gnu.out

? You should see a list of entries like:

1 1 1    5   1   1   1    2.490000       0.080
1 1 2    4   1   1   1    3.210003       0.125

Then, the tiny1 phase doesn't link to BLAS, while small1 does.
I wonder if you are linking to the correct mkl libraries.
Could you take a look inside the directory

run_small_*

and see if the source files are produced? Then you can compile by hand one 
of this file and see if it works when executed.
For instance, you can do something like:

cd <small directory>
gfortran -O2 -funroll-loops -ffast-math -ftree-vectorize -march=native 
-fno-inline-functions small_find_1_1_1.f90 -o 
output_linux.gnu/small_find_1_1_1.x -L/mkl/lib/intel64 -Wl,--start-group 
-lmkl_gf_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group 
cd <output_linux.gnu>
./small_find_1_1_1.x

Is /mkl/lib/intel64 correct? 
I think the problem is somewhere in the compilation of the files...

Alfio



Il giorno venerdì 13 settembre 2013 09:26:54 UTC+2, Joe Greenstone ha 
scritto:
>
> Hi,
>  
> After successful tiny1:
>  
> module load intel/11.10.319
> module load gnu/4.8.0
> ./generate -c config/linux.gnu -j 0 -t 8 small1
>  
> gives me:
>  
> gfortran -O2 -funroll-loops -ffast-math -ftree-vectorize -march=native 
> -fno-inline-functions small_find_32_32_26.f90 -o 
> output_linux.gnu/small_find_32_32_26.x -L/mkl/lib/intel64 -Wl,--start-group 
> -lmkl_gf_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group 
> gfortran -O2 -funroll-loops -ffast-math -ftree-vectorize -march=native 
> -fno-inline-functions small_find_32_32_32.f90 -o 
> output_linux.gnu/small_find_32_32_32.x -L/mkl/lib/intel64 -Wl,--start-group 
> -lmkl_gf_lp64 -lmkl_sequential -lmkl_core -Wl,--end-group 
> ./output_linux.gnu/small_find_1_1_1.x > 
> output_linux.gnu/small_find_1_1_1.out
> ./output_linux.gnu/small_find_1_1_4.x > 
> output_linux.gnu/small_find_1_1_4.out
> ./output_linux.gnu/small_find_1_1_5.x > 
> output_linux.gnu/small_find_1_1_5.out
> ./output_linux.gnu/small_find_1_1_6.x > 
> output_linux.gnu/small_find_1_1_6.out
> ./output_linux.gnu/small_find_1_1_8.x > 
> output_linux.gnu/small_find_1_1_8.out
> ./output_linux.gnu/small_find_1_1_9.x > 
> output_linux.gnu/small_find_1_1_9.out
> ./output_linux.gnu/small_find_1_1_13.x > 
> output_linux.gnu/small_find_1_1_13.out
> ./output_linux.gnu/small_find_1_1_16.x > 
> output_linux.gnu/small_find_1_1_16.out
> Program received signal SIGSEGV: Segmentation fault - invalid memory 
> reference.
> Backtrace for this error:
> #0  0x7ff304a2a1b2
> #1  0x7ff304a2a8ee
> #2  0x3e12a3291f
> #3  0x40235a
> #4  0x40258b
> #5  0x4010e9
> #6  0x401efa
> #7  0x402cfc
> #8  0x3e12a1ecdc
> #9  0x400978
> /bin/sh: line 1: 69121 Segmentation fault      (core dumped) 
> ./output_linux.gnu/small_find_1_1_5.x > 
> output_linux.gnu/small_find_1_1_5.out
> make: *** [output_linux.gnu/small_find_1_1_5.out] Error 139
> make: *** Waiting for unfinished jobs....
>  
>  
> config/linux.intel also segfaults.
> When I run batch (j=1 and w=), the Makefile is not found though it is 
> where it should be.
> Stuck for a way to proceed.
>  
> -Joe 
>  
>  
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20130913/0ebc4269/attachment.html>


More information about the CP2K-user mailing list