If all the other ISA implementations won't do anything, it probably shouldn't be part of the ISA interface. What do those functions do, and can it be distilled down to primitives the other ISAs can use? If not, can it be hidden behind some other ISA interface? I'd rather not have a whole class of functions that only do anything useful in MIPS (or x86, or SPARC, etc.).
Gabe Korey Sewell wrote: > I'm a little unsure of what you are saying here Nate. > > I am trying to add a mt.hh for ALPHA (so it can use the InOrder model) > but through some fault of mines the alpha version of the mt.hh file > got omitted. > > Eventually, any ISA wanting to use InOrder would need to at the very > least have some "bare" MT functions since there is no "base" or > "generic" ISA subdirectory that contains bare-bone implementations of > ISA functions needed by the CPUs. > > On Sat, Feb 21, 2009 at 11:22 PM, nathan binkert <[email protected]> wrote: > >>> 1) not sure why the mt.hh file didnt get added to the patch, but it >>> should be there >>> >> I think the real problem is that you're adding an mt.hh, but only for >> mips. That doesn't work so well with the switching directory code. >> _______________________________________________ >> m5-dev mailing list >> [email protected] >> http://m5sim.org/mailman/listinfo/m5-dev >> >> > > > > _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
