> On 2010-11-21 00:21:28, Gabe Black wrote: > > src/arch/arm/isa/insts/swap.isa, line 68 > > <http://reviews.m5sim.org/r/319/diff/1/?file=5378#file5378line68> > > > > Is this line too long now? Also, I'm not sure what's going on with the > > new 'IsStoreConditional' flag. Doesn't there need to be a load locked > > somewhere? Also, can't those fail? Would the x86 version of locking be more > > appropriate since it can't? Why is any flag necessary since the swap is (I > > think) done atomically through the memory request itself? Or are we moving > > away from that?
fixed > On 2010-11-21 00:21:28, Gabe Black wrote: > > src/cpu/o3/iew_impl.hh, line 1255 > > <http://reviews.m5sim.org/r/319/diff/1/?file=5380#file5380line1255> > > > > The only difference between this else if and the if above it is that > > activityThisCycle is called for the if but not the else if. I'm not sure > > that's correct, though admittedly the use of the activity stuff in O3 is a > > little mysterious to me, but in any case I don't think it would hurt > > anything to call it in both cases and merge those two blocks. assuming it doesn't break something, I'll make the change. - Ali ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.m5sim.org/r/319/#review498 ----------------------------------------------------------- On 2010-11-19 16:12:35, Ali Saidi wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.m5sim.org/r/319/ > ----------------------------------------------------------- > > (Updated 2010-11-19 16:12:35) > > > Review request for Default. > > > Summary > ------- > > O3: Support SWAP and predicated loads/store in ARM. > > > Diffs > ----- > > src/arch/arm/isa/insts/swap.isa 6286bb50127e > src/cpu/o3/iew.hh 6286bb50127e > src/cpu/o3/iew_impl.hh 6286bb50127e > src/cpu/o3/lsq_unit_impl.hh 6286bb50127e > > Diff: http://reviews.m5sim.org/r/319/diff > > > Testing > ------- > > > Thanks, > > Ali > >
_______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
