> On May 28, 2012, 7:11 a.m., Ali Saidi wrote: > > > > Ali Saidi wrote: > did you happen to see if this improved performance? I imagine it would.
For twolf on simple atomic original: host_seconds 80.03 host_seconds 79.97 host_seconds 80.16 new: host_seconds 73.04 host_seconds 72.95 host_seconds 73.47 > On May 28, 2012, 7:11 a.m., Ali Saidi wrote: > > src/arch/arm/remote_gdb.cc, line 296 > > <http://reviews.gem5.org/r/1231/diff/1/?file=26972#file26972line296> > > > > space You're right that there should be a space, but it's a pre-existing condition that matches the rest of the file. It should be cleaned up in a later change, but for now it's better for it to be consistently wrong. - Gabe ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/1231/#review2828 ----------------------------------------------------------- On May 28, 2012, 2:56 a.m., Gabe Black wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/1231/ > ----------------------------------------------------------- > > (Updated May 28, 2012, 2:56 a.m.) > > > Review request for Default. > > > Description > ------- > > Changeset 9032:a43d6b23beec > --------------------------- > ARM: Keep a copy of the fpscr len and stride fields in the decoder. > > Avoid reading them every instruction, and also eliminate the last use of the > thread context in the decoders. > > > Diffs > ----- > > src/arch/arm/decoder.hh 1f2568933bc5 > src/arch/arm/decoder.cc 1f2568933bc5 > src/arch/arm/isa.cc 1f2568933bc5 > src/arch/arm/remote_gdb.cc 1f2568933bc5 > > Diff: http://reviews.gem5.org/r/1231/diff/ > > > Testing > ------- > > > Thanks, > > Gabe Black > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
