Why all the new methods on FloatRegFile? Are they used only by the in-order CPU? I'm just surprised they weren't there already.
Also, is this change below really necessary? diff -r 0555121b5c5f -r d4ff65e87a0b src/cpu/inorder/resources/tlb_unit.cc > --- a/src/cpu/inorder/resources/tlb_unit.cc Wed Apr 08 22:21:30 2009 > -0700 > +++ b/src/cpu/inorder/resources/tlb_unit.cc Fri Apr 10 01:04:35 2009 > -0400 > @@ -82,7 +83,7 @@ TLBUnit::execute(int slot_idx) > // After this is working, change this to a reinterpret cast > // for performance considerations > TLBUnitRequest* tlb_req = > dynamic_cast<TLBUnitRequest*>(reqMap[slot_idx]); > - assert(tlb_req); > + assert(tlb_req != 0x0); > > DynInstPtr inst = tlb_req->inst; > int tid, seq_num, stage_num; >
_______________________________________________ m5-dev mailing list m5-dev@m5sim.org http://m5sim.org/mailman/listinfo/m5-dev
