develooper Front page | perl.perl5.porters | Postings from September 2014

Re: new long doubles option: gcc libquadmath

Thread Previous | Thread Next
Josh Juran
September 20, 2014 02:25
Re: new long doubles option: gcc libquadmath
Message ID:
On Sep 19, 2014, at 1:12 PM, Jarkko Hietaniemi <> wrote:

> I think the "12 bytes" for the x86 extended precision is an "urban
> legend" of misunderstanding: the number of bits is always 80 (64 for mantissa, 15 for exponent, 1 for sign, no "implicit bit").  The 12
> comes from the ILP32 way of storing the 10-byte thing in three 4-byte words.  In LP64 (like x86_64) it's stored in two 8-byte words, hence 16
> bytes.  The unused bytes (2 or 6) are unused garbage.  They might be
> zero bits, or they might contain your credit card number.

I know you're being facetious, and crypto uses integer math so we're not talking about leaking private key material, but as a general principle, transporting garbage should be avoided.  It was bad when Microsoft Word bumped a document file's logical EOF to match its physical EOF (thus carrying around a partial block of who-knows-what prior data, which very well could have included your credit card number), and it's bad now.  Any unused bits in the in-memory format should be zeroed before transmission or storage.


Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About