[CP2K:9863] Re: Build failure on 32bit

Michael Banck mba... at gmx.net
Mon Jan 15 14:29:38 UTC 2018


On Mon, Jan 15, 2018 at 02:51:54AM -0800, Alfio Lazzaro wrote:
> is there a log that I can see? 

I've uploaded it now, so the log is here (without a useful backtrace):


> We test CP2K against libint-1.1.6, I'm not sure, what's wrong there...

I don't think there's a huge difference code-wise between 1.1.6 and

> There are also people reporting problem with libint is dynamically
> linked 
> (see https://groups.google.com/d/msg/cp2k/qvJQpxU3hzc/GXVyaHhyCQAJ )

It looks like it might be the same crash indeed, but again, that
backtrace is not very informative.

One additional issue is that I'm building libint twice now, once using
autotools and once using cmake, as Psi4 requires the cmake files to
build against a system libint.  It could be that there is some skew
there; but I will also see how a static-only build goes.



More information about the CP2K-user mailing list