This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH PING]: middle-end: provide generic "finite" transformation
On 7/9/07, Kaveh R. GHAZI <email@example.com> wrote:
On Sun, 8 Jul 2007, Richard Guenther wrote:
> On 7/8/07, Kaveh R. GHAZI <firstname.lastname@example.org> wrote:
> > PING.
> > http://gcc.gnu.org/ml/gcc-patches/2007-06/msg01211.html
> > This one also generates correct (but slightly suboptimal) code for floats.
> > It also will be fixed with my type-generic builtins patch. But this one
> > can be installed now.
> isfinite should be DEF_C99_C90RES_BUILTIN (or DEF_C99_BUILTIN,
> I don't have C90 handy to check if it is a reserved keyword there). Also
> "I'll resubmit with documenation for the new
> builtin. (I wasn't sure which doc section to put it in right now.)"
> Builtins are documented in extend.texi.
> Same concerns on optimize_size and soft-fp targets.
> Please re-post with the changes above (after checking C90).
By "doc section" back then I meant whether it should go in the C99 or
gcc-internal section. This was determined by whether I used DEF_GCC or
DEF_C99. Since Joseph later clarified DEF_GCC, that's what I went with.
This version adds documentation in extend.texi, and updates the attributes
in builtins.def to use "type generic".
Ok for mainline?