> On Feb. 23, 2016, 2:58 p.m., Jason Lowe-Power wrote: > > src/gpu-compute/compute_unit.hh, line 305 > > <http://reviews.gem5.org/r/3338/diff/1/?file=53816#file53816line305> > > > > From what I can tell on a cursory grep there are no places where you > > call this without a req. Is this true? If so, you should remove the default > > there too.
It is called without req, up above in the same file... In any case, I would be up for getting rid of the default value here. Side note: I was a little premature in posting this... I'm spending a little more time to confirm that I did this correctly. I hope to either update the patch or this comment next week. - Marc ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/3338/#review8024 ----------------------------------------------------------- On Feb. 23, 2016, 12:40 a.m., Marc Orr wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/3338/ > ----------------------------------------------------------- > > (Updated Feb. 23, 2016, 12:40 a.m.) > > > Review request for Default. > > > Repository: gem5 > > > Description > ------- > > Changeset 11349:f14a633b3ac8 > --------------------------- > gpu-compute: fix gpu memfence > > Memory fence instructions were incorrectly annotated to indicate that they > were > initiated on a kernel launch/end. This patch correctly annotates an explicit > memory fence. > > > Diffs > ----- > > src/arch/hsail/insts/decl.hh 64e862d3758f20730c25961e0bbefb1123be4fec > src/gpu-compute/compute_unit.hh 64e862d3758f20730c25961e0bbefb1123be4fec > > Diff: http://reviews.gem5.org/r/3338/diff/ > > > Testing > ------- > > > Thanks, > > Marc Orr > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
