Re: [Boost-users] multiprecision -- eigen matrices using dynamically allocated variable-precision mpfr_float

Subject: Re: [Boost-users] multiprecision -- eigen matrices using dynamically allocated variable-precision mpfr_float
From: John Maddock (jz.maddock_at_[hidden])
Date: 2015-04-15 13:33:26

> I have been able to pin it down to the allocate_dynamic switch; if I 
> select allocate_stack, the product computes correctly; 
> allocate_dynamic reliably generates nans. Fixed-precision numbers work 
> just fine, and expression templates on/off make no difference.  I'd 
> like to use the dynamic allocation, as I expect to use thousands of 
> digits at times.
>
> I am using clang++, with c++11, on OSX, Boost 1.57.0, Eigen 3.2.4.  
> Same behaviour in 1.56 / 3.2.3.
>
> It seems like maybe an offset issue, but I just can't get there by 
> myself.  I've also played with defining the NumTraits in Eigen various 
> ways, but nothing affects the nan behaviour. Does anyone have any ideas?
>
Well that's weird.

The only difference between dynamic and static allocation that I can 
think of at present is in the way they handle move semantics - 
dynamically allocated values have destructive moves - which is to say 
the moved-from-value is left uninitialized.  Can you try building with 
BOOST_NO_CXX11_RVALUE_REFERENCES defined and see if that fixes the 
issue?  If it does, I'll have to investigate what Eigen is doing in this 
case.

John.