ralna/spral

Errors in ssmfet_c

amontoison opened this issue · 7 comments

amontoison commented

In some cases generalized_shift_double could fail.

test:         ssmfet_c
start time:   03:55:12
duration:     1.80s
result:       exit status 7
command:      MALLOC_PERTURB_=212 OMP_PROC_BIND=true OMP_CANCELLATION=true LD_LIBRARY_PATH=/home/alexis/Applications/spral/builddir/ /home/alexis/Applications/spral/builddir/ssmfet_c
----------------------------------- stdout -----------------------------------
testing ssmfe_core...
testing standard_double...ok
testing generalized_double...ok
testing largest_double...ok
testing standard_double_complex...ok
testing generalized_double_complex...ok
testing largest_double_complex...ok
0 errors
testing ssmfe_expert...
testing standard_double...ok
testing generalized_double...ok
testing standard_shift_double...ok
testing generalized_shift_double...ok
testing buckling_double...ok
testing standard_double_complex...ok
testing generalized_double_complex...ok
testing standard_shift_double_complex...ok
testing generalized_shift_double_complex...ok
testing buckling_double_complex...ok
0 errors
testing ssmfe...
testing standard_double...ok
testing generalized_double...ok
testing standard_shift_double...ok
testing generalized_shift_double...7 errors
testing buckling_double...ok
testing standard_double_complex...ok
testing generalized_double_complex...ok
testing standard_shift_double_complex...ok
testing generalized_shift_double_complex...ok
testing buckling_double_complex...ok
jfowkes commented

Is this an intermittent random failure? Does it happen on all platforms?

amontoison commented

It seems to be an intermittent random failure. I only got it on my computer (Linux x64).

jfowkes commented

@amontoison I am unable to reproduce, please post the failing testlog.txt

amontoison commented

It's my first message of this Issue Jari.
I don't have more details than that.

jfowkes commented

@amontoison when you run the meson tests for SPRAL they always generate a testlog.txt in the meson logs.

amontoison commented

Yes, and the message at the top of this issue is what I got from the testlog.txt.
Do you want the full testlog.txt?

jfowkes commented

Ah I see thanks, unfortunately that's not the most informative test log.

  • 😕1