bug in gmp_fprintf?

[paul zimmermann]

       Dear Torbjörn,

> Please try the following patch.  I have only fixed things that caused
> problems triggered by your test case, and things dependent on the
> variables thus type changed.

I will try that patch in CADO-NFS with batch cofactorization (which is how I
discovered the issue). It will take some time because I first need to produce
a batch file (which takes a few hours), then to use it (which will take another
few hours). I will follow up here.

Paul