Bug Report GMP-5.1.1 Mini-gmp t-double.c
marc.glisse at inria.fr
Mon Feb 25 16:20:35 CET 2013
On Mon, 25 Feb 2013, Niels Möller wrote:
> Do you know if there's any better (safer or more portable) way to drop
> extended precision, or should I just stick to volatile?
volatile is good (we don't care about performance for this code). We just
need to remember to always write intermediate results to variables (like
the current code properly does, apparently).
More information about the gmp-bugs