yade-dev team mailing list archive
-
yade-dev team
-
Mailing list archive
-
Message #03487
Re: wm3-eigen glue
Now we have just 2 problem places in tt.py:
assertSeqAlmostEqual(q1*x,y)
assertSeqAlmostEqual(q1*q1*x,-x)
r58
______________________________
Anton Gladkyy
2010/2/21 Anton Gladky <gladky.anton@xxxxxxxxx>
> Hi, Vaclav!
>
> I have some progress on tt.py, but I have not resolved the problem
> completely.
> I have found some "problem" places in those script and marked them. It is
> there just 5 ones.
> tt.py is in rev.56 now. Could you not check it? What can you advise?
>
> Thank you
> ______________________________
>
> Anton Gladkyy
>
>
> 2010/2/19 Václav Šmilauer <eudoxos@xxxxxxxx>
>
>
>> > I tried to disable vectorization by both ways, indicated on the
>> > webpage. But it didn't help. I still have crash.
>> > But I think it is not the same, like you had, right?
>> Yes, this one is real crash, whereas I had assertion failuer (in case of
>> the tt.py script I sent lastly). This looks as if vectorization weren't
>> disabled, though:
>> > #4 <signal handler called>
>> > #5 0x00007f1bec85586b in _mm_store_pd (x=0x7fff72cdb030)
>> > at /usr/lib/gcc/x86_64-linux-gnu/4.4.1/include/emmintrin.h:148
>> > #6 ei_pstore<double, double __vector__> (x=0x7fff72cdb030)
>> > at /usr/include/eigen2/Eigen/src/Core/arch/SSE/PacketMath.h:110
>> emmintrin.h is sse2-specific header (judging by comments at the
>> beginning of that file), and it shouldn't be used at all if you disable
>> vectorization.
>>
>> Sorry, I don't have the answer here and I wish you good luck finding the
>> problem. I might try monday, I have to write an exam paper deadlines
>> sunday evening.
>>
>> Cheers, Vaclav
>>
>>
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~yade-dev<https://launchpad.net/%7Eyade-dev>
>> Post to : yade-dev@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~yade-dev<https://launchpad.net/%7Eyade-dev>
>> More help : https://help.launchpad.net/ListHelp
>>
>
>
Follow ups
References