[tlbuild] armhf-linux

[Peter Breitenlohner]

On Wed, 22 Apr 2015, Karl Berry wrote:

>    === configuring in mpfr (/path/to/src/luatex/build/libs/mpfr)
>    configure: WARNING: oops, unrecognised float format: IEEE double, little endian
>
> Please ask Luigi (he isn't on this list).  If it indicates that some
> float support really won't be available in the binary, it should
> probably be fatal, not a warning.

I just had another look at the sources. Analyzing the format of 'long
double' for armhf-linux yields 'IEEE double, little endian' which is fine. 
The 'unrecognised format' refers to the fact that there is no special code
to detect NANs, so the generic code is used.  Perhaps the warning should be
phrased differently or even be omitted, but I wanted to keep the warning as
used by the original build system distributed with mpfr-3.1.2.

Regards
Peter