diff options
| author | Brad Beckmann <Brad.Beckmann@amd.com> | 2010-08-20 11:46:13 -0700 |
|---|---|---|
| committer | Brad Beckmann <Brad.Beckmann@amd.com> | 2010-08-20 11:46:13 -0700 |
| commit | 984adf198ae60eb0d32d0818fc6233d4475b6eb1 (patch) | |
| tree | 498d146f16206751db2239eda5da1c4764079b49 /src/mem/protocol/MOESI_hammer-msg.sm | |
| parent | 29c45ccd2322470d0d6cef0ae20600c8c68f97e9 (diff) | |
| download | gem5-984adf198ae60eb0d32d0818fc6233d4475b6eb1.tar.xz | |
ruby: Updated MOESI_hammer L2 latency behavior
Previously, the MOESI_hammer protocol calculated the same latency for L1 and
L2 hits. This was because the protocol was written using the old ruby
assumption that L1 hits used the sequencer fast path. Since ruby no longer
uses the fast-path, the protocol delays L2 hits by placing them on the
trigger queue.
Diffstat (limited to 'src/mem/protocol/MOESI_hammer-msg.sm')
| -rw-r--r-- | src/mem/protocol/MOESI_hammer-msg.sm | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/src/mem/protocol/MOESI_hammer-msg.sm b/src/mem/protocol/MOESI_hammer-msg.sm index 5d8226eb6..4856178a1 100644 --- a/src/mem/protocol/MOESI_hammer-msg.sm +++ b/src/mem/protocol/MOESI_hammer-msg.sm @@ -55,6 +55,7 @@ enumeration(CoherenceResponseType, desc="...") { // TriggerType enumeration(TriggerType, desc="...") { + L2_to_L1, desc="L2 to L1 transfer"; ALL_ACKS, desc="See corresponding event"; ALL_ACKS_NO_SHARERS, desc="See corresponding event"; } |