----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/2170/#review4926 -----------------------------------------------------------
Ship it! Ship It! - Andreas Hansson On Feb. 22, 2014, 3:34 p.m., Anthony Gutierrez wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/2170/ > ----------------------------------------------------------- > > (Updated Feb. 22, 2014, 3:34 p.m.) > > > Review request for Default. > > > Repository: gem5 > > > Description > ------- > > Changeset 10087:1c31282d70a0 > --------------------------- > arm: make the bi-mode predictor the default for O3_ARM_v7a_BP > > the branch predictor used in the Cortex-A15 is a bi-mode style predictor, > see: > > http://arm.com/files/pdf/at-exploring_the_design_of_the_cortex-a15.pdf > and > http://nvidia.com/docs/IO/116757/NVIDIA_Quad_a15_whitepaper_FINALv2.pdf > > this patch makes the bi-mode predictor the default for the ARM O3 CPU. > > > Diffs > ----- > > configs/common/O3_ARM_v7a.py 2beea2a439b43bf95bc040198322d4869b07f4f3 > > Diff: http://reviews.gem5.org/r/2170/diff/ > > > Testing > ------- > > Ran several of the SPEC CPU2006 benchmarks with the default tournament > predictor and this bi-mode predictor. With this patch, the branch miss > statistics are more accurate when compared to a VExpress board. > > > Thanks, > > Anthony Gutierrez > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
