[with patch, spkg] update flint in sage again to r1047

[was]

From Bill Hart:

Hi William,

I've just dug another pile of bugs out of FLINT. All functions in
fmpz_poly normalise correctly now (or at least they should).

I ran a much more comprehensive set of tests, checking for unusual
corner cases. I found a handful of them and removed them.

I've also added a make all to the makefile.

Functions still may not deal with length zero polynomials correctly
(though most functions do) and with aliased inputs, so there will
probably be more changes over the next few days. But for now, FLINT is
much more bulletproof than it was.

Moreover, the svn revision to get it: 1045

[was]

Actually, update to 1047:

As of revision 1047 all functions now deal correctly with length zero
functions. I have also added some managed functions for scalar
multiplication of a polynomial by a scalar. I've also hardened some of
the memory management functions so that stupid users who try to
allocate -1 limbs won't run into unexpected trouble.

[robertwb]

New flint spkg and bundle (to apply to sage-main) up at

 http://sage.math.washington.edu/home/robertwb/flint/

[mabshoff]

Ok, this should go into 2.8.13. I will ask for somebody to review this.

Cheers,

Michael

[mabshoff]

This looks good to me and should be in 2.8.13.alpha0.

Cheers,

Michael

[mabshoff]

Merged in 2.8.13.alpha0.

[mabshoff]

I fixed a ntl linking issue, so now there is a flint-0.9.p0.spkg

Cheers,

Michael