mem-cache: Move access latency calculation to Cache
authorDaniel R. Carvalho <odanrc@yahoo.com.br>
Thu, 18 Oct 2018 13:31:51 +0000 (15:31 +0200)
committerDaniel Carvalho <odanrc@yahoo.com.br>
Wed, 14 Nov 2018 21:02:08 +0000 (21:02 +0000)
commitc6e0d8f54f1ce90933f95a7a3a875fed53b8ee3e
tree60bb086bf5b96e5c694eb662f0e34abeae26e6ad
parentea0f654923d13a2f81fe7c22eeca4c8cffae947a
mem-cache: Move access latency calculation to Cache

Access latency was not being calculated properly, as it was
always assuming that for hits reads take as long as writes,
and that parallel accesses would produce the same latency
for read and write misses.

By moving the calculation to the Cache we can use the write/
read information, reduce latency variables duplication and
remove Cache dependency from Tags.

The tag lookup latency is still calculated by the Tags.

Change-Id: I71bc68fb5c3515b372c3bf002d61b6f048a45540
Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Reviewed-on: https://gem5-review.googlesource.com/c/13697
Reviewed-by: Nikos Nikoleris <nikos.nikoleris@arm.com>
Maintainer: Nikos Nikoleris <nikos.nikoleris@arm.com>
src/mem/cache/base.cc
src/mem/cache/base.hh
src/mem/cache/tags/Tags.py
src/mem/cache/tags/base.cc
src/mem/cache/tags/base.hh
src/mem/cache/tags/base_set_assoc.hh
src/mem/cache/tags/fa_lru.cc
src/mem/cache/tags/fa_lru.hh
src/mem/cache/tags/sector_tags.cc
src/mem/cache/tags/sector_tags.hh