Re: bignum branch

[Andy Moreton <andrewjmoreton@gmail.com>]

On Tue 07 Aug 2018, Tom Tromey wrote:

>>>>>> "Andy" == Andy Moreton <andrewjmoreton@gmail.com> writes:
>
> Andy> The mpz_import patch [1] results in appears to cause a problem in the CCL
> Andy> tests. I'll report back when I have time to look into it further.
>
> Thanks.  This isn't a major issue, but the current code does cause a
> compiler warning in alloc.c on my machine -- the code in question won't
> ever be run here, but gcc complains about an overlong shift.  The
> mpz_import approach would avoid this.

I think the problem was pilot error. I have now run the CCL tests with
the mpz_import patch without problems on the following Windows builds:
1) mingw64 x86_64 (MSYS2)
2) mingw64 i686   (MSYS2)
3) mingw32 i686 (mingw.org)
   This works after also applying commit 024d20f81e from master, needed
   to fix build problems (i.e. the problem will be fixed by merging into
   master, or rebasing the bignum branch).

I have not tried a 32bit wide int build, but I would expect that they
should work given that (1) works.

    AndyM