|Subject:||Re: [Tinycc-devel] __builtin_expect|
|Date:||Thu, 31 May 2012 22:58:55 +0200|
|User-agent:||KMail/1.13.7 (Linux/3.2.0-2-amd64; KDE/4.7.4; x86_64; ; )|
Le jeudi 31 mai 2012 22:49:19, grischka a écrit :
> Thomas Preud'homme wrote:
> > Le jeudi 31 mai 2012 22:31:55, Milutin Jovanović a écrit :
> > > How about asking them (glibc people)? Maybe they have a reason for
> > > this.
> > >
> > > Maybe they will give ETA. If you get silence, then we could ask if tcc
> > >
> > > changes can address this.
> > It's already done. See
> > http://sourceware.org/bugzilla/show_bug.cgi?id=14188
> If that works may I suggest also to try to get rid of the define
> __REDIRECT() uglinesses in TinyCC? ;)
I believe it's a quite different problem. According to the commit message I did:
"Add support for __REDIRECT_NTH as eglibc makes use of this macro to redirect long double functions to long functions on arch not supporting long double."
So if tcc is ported once to these architecture we'll need it. Also, I don't think it would be convenient for them to do differently. And finally, it's just a small line in tcc. The real stuff is asm aliasing but we probably want to keep this. Or maybe not?
> --- grischka
Description: This is a digitally signed message part.
|[Prev in Thread]||Current Thread||[Next in Thread]|