This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: mips SNaN/QNaN is swapped

On Thu, Mar 27, 2003 at 12:28:22AM -0300, Alexandre Oliva wrote:
> Thanks, but it doesn't solve the problem of the canonical
> representation.

No, but it solves everything else.  It should go in as is
for the nonce.  Frankly I don't consider the canonical QNaN
problem to be critical.  You'll get correct runtime behaviour
either way.

> To fix that, I only see two alternatives: make
> get_canonical_[qs]nan() format-specific entry points, or introduce
> strings to be used to create the canonical values.

Strings are ugly; I'd like to avoid that.  What's the form
of the MIPS canonical QNaN?


r~
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]