mem: Address mapping with fine-grained channel interleaving
authorAndreas Hansson <andreas.hansson@arm.com>
Mon, 22 Apr 2013 17:20:34 +0000 (13:20 -0400)
committerAndreas Hansson <andreas.hansson@arm.com>
Mon, 22 Apr 2013 17:20:34 +0000 (13:20 -0400)
commite23e3bea8bc332626e026078dc8b23c983fc890f
tree80ef9eff694a07b7d0907198e1ca825475552c36
parente61799aa7cfa7c9f9d4b1099ac4fd9ecbadd4b7b
mem: Address mapping with fine-grained channel interleaving

This patch adds an address mapping scheme where the channel
interleaving takes place on a cache line granularity. It is similar to
the existing RaBaChCo that interleaves on a DRAM page, but should give
higher performance when there is less locality in the address
stream.
src/mem/SimpleDRAM.py
src/mem/simple_dram.cc