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: [PATCH] Making double precision constants work with -fsingle-precision-constant


Richard Henderson wrote:
> On Thu, Dec 08, 2005 at 11:48:54AM -0500, Carlos O'Donell wrote:
> 
>>Would it be useful to define __DBL_HAS_DENORM__, it is already a
>>property of the format (gcc/gcc/real.h)?
>  
> I didn't think it was useful before, but now it would appear
> to be required.  Supposing we keep this option.

I guess I think we should fix the bug.  So, Carlos, would you add
__DBL_HAS_DENORM__ and then revise the libstdc++ changes, since we need
__DBL_HAS_DENORM__ to make the libstdc++ changes correct?

Then, we could also deprecate the option in 4.2, if there's consensus
that it should go away.

We've advised our customer that there's not a lot of love of the option.
and that it might go away.  They feel it's a useful option for embedded
users who are reusing workstation code in an embedded environment, where
doubles are expensive.  (I wonder what happens -- or should happen, even
-- when you use this option with a double-precision constant.)  I've
advised them that if they feel strongly about keeping this option in the
long term, they should make their case here.

-- 
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304


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