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: Fix prefetch address generation for XScale


On Thu, 2001-12-20 at 10:52, Nick Clifton wrote:
> Hi Aldy,
> 
> >  >   The builtin-prefetch?.c tests have exposed a problem with the ARM
> >  >   port's implementation of the prefetch instruction.  The tests could
> >  >   generate an instruction of the form:
> > 
> >  >         (prefetch (plus:SI (reg:SI) (const_int 0)))
> > 
> > why wasn't this folded?
> 
> Umm, what do you mean by folded ?
> 
> Do you mean why wasn't it like this ?
> 
>      (prefetch (plus:SI (reg:SI)
>                         (const_int 0)))

sorry, i am speaking purely out of ignorance, but shouldn't that be
folded to:

	(prefetch (reg SI))


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