RE: [Patch ping] was: Option to compile for both AMD64 and EM64T?

["Justin Palmer" <Justin dot E dot Palmer at saic dot com>]

Hi Kelley,

Any clue whether this patch has been backported?  In particular, I'm curious as to whether Red Hat included the patch since Red Hat has released Enterprise Linux Update 3 for both AMD64 and EM64T.  Will GCC compile an application under that OS that will execute on both AMD and Intel processors?

Thanks for any help,
Justin


> -----Original Message-----
> From: Kelley Cook [mailto:kcook@gcc.gnu.org]
> Sent: Tuesday, August 31, 2004 7:09 PM
> To: Justin Palmer
> Subject: [Patch ping] was: Option to compile for both AMD64 and EM64T?
> 
> 
> Justin Palmer wrote:
> > Hi,
> > 
> > Since the differences between AMD64 and EM64T are minor, 
> will GCC have an option to compile applications for both?
> > 
> > Linux-Kernel Archive reference:
> > http://www.ussg.iu.edu/hypermail/linux/kernel/0402.3/0276.html
> > 
> > Currently, the -mtune, -march, and -mcpu values must 
> specify one of: nocona, k8, opteron, athlon64, or athlon-fx.  
> My software group has been using i686 to build one executable 
> for both Intel and AMD 32-bit processors and would like to 
> build one executable for their 64-bit processors.
> > 
> > Also, do the docs need updated for -m64 or is that option 
> really for AMD64 only?
> 
> With the experimental GCC 3.5, -march=x86-64 does what you want.  You 
> probably could backport the patch to GCC 3.4 without much problem.
> 
> And the reminds of a patch that I submitted a while ago which 
> resets the 
> default x86 complier to the generic version for same reason 
> you mentioned.
> 
> Could someone review 
> http://gcc.gnu.org/ml/gcc-patches/2004-05/msg00352.html?
> 
>