> On 2011-03-11 19:21:23, Ali Saidi wrote: > > src/arch/arm/predecoder.hh, line 154 > > <http://reviews.m5sim.org/r/570/diff/1/?file=10841#file10841line154> > > > > i'll look, but I'm not sure that it does. I'm pretty sure it still > > causes a mispredict without it.
Ok, yeah, I looked at it more carefully just now. I think all you need to do is check whether the size is set on the pc already, and if so don't update the npc or the size. In those cases the pc will have already been set up by the predecoder with a valid npc. If it comes back that way, the npc may have an updated value and shouldn't be overwritten. - Gabe ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.m5sim.org/r/570/#review949 ----------------------------------------------------------- On 2011-03-11 15:21:28, Ali Saidi wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.m5sim.org/r/570/ > ----------------------------------------------------------- > > (Updated 2011-03-11 15:21:28) > > > Review request for Default, Ali Saidi, Gabe Black, Steve Reinhardt, and > Nathan Binkert. > > > Summary > ------- > > ARM: Identify branches as conditional or unconditional and direct or > indirect. > > > Diffs > ----- > > src/arch/arm/insts/branch.hh 5138d1e453f1 > src/arch/arm/isa/insts/branch.isa 5138d1e453f1 > src/arch/arm/isa/templates/branch.isa 5138d1e453f1 > src/arch/arm/predecoder.hh 5138d1e453f1 > src/arch/arm/types.hh 5138d1e453f1 > src/cpu/o3/bpred_unit_impl.hh 5138d1e453f1 > > Diff: http://reviews.m5sim.org/r/570/diff > > > Testing > ------- > > > Thanks, > > Ali > > _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
