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: RFC: MIPS clones without lwl/lwr/swl/swr


On Mon, 16 May 2005, Mark Mitchell wrote:

> > -m{,no-}patent-free is ok IMHO. For convenience there should also be a
> > -march=rtl8181 switch which implies the patent free mode.
> 
> Chiming in randomly, I actually think an option focusing on the patented-ness
> of these instructions is not the best choice.  What if there are more some
> day, or the patent gets invalidated, or the user doesn't know about the
> patent?  How about -mno-<something>-load-store?

 Don't forget the problem of system controllers required to support 
partial transfers for these instructions to work, which I've mentioned and 
which actually exists -- unfortunately there are such implementations out 
there.  Using an option called "-mpatent-free" as a workaround for that 
seems rather counter-intuitive...

  Maciej


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