cpu: Implement per-thread GHRs
authorMitch Hayenga <mitch.hayenga@arm.com>
Tue, 5 Apr 2016 17:20:19 +0000 (12:20 -0500)
committerMitch Hayenga <mitch.hayenga@arm.com>
Tue, 5 Apr 2016 17:20:19 +0000 (12:20 -0500)
commitd99deff8ea296fd28b48da08aba577a1e7dfc01b
treebb8645f94f33b45b040776e5e0dd0a0d8edb306f
parent0fd4bb7f12d8a633f3ff0abe61d4f3a78bca6f84
cpu: Implement per-thread GHRs

Branch predictors that use GHRs should index them on a
per-thread basis.  This makes that so.

This is a re-spin of fb51231 after the revert (bd1c6789).
src/cpu/pred/2bit_local.cc
src/cpu/pred/2bit_local.hh
src/cpu/pred/bi_mode.cc
src/cpu/pred/bi_mode.hh
src/cpu/pred/bpred_unit.cc
src/cpu/pred/bpred_unit.hh
src/cpu/pred/tournament.cc
src/cpu/pred/tournament.hh