Do a functional access to levels above on a read as a temporary solution for L2's in FS

Fix a small writeback bug when missing in the L2 in atomic mode

src/mem/bus.cc:
    Fix a comment to make sense
src/mem/cache/cache_impl.hh:
    Do a functional access to levels above on a read as a temporary solution for L2's in FS
    Also fix a small writeback miss in L2 issue
src/mem/cache/coherence/simple_coherence.hh:
src/mem/cache/coherence/uni_coherence.cc:
src/mem/cache/coherence/uni_coherence.hh:
    Do a functional access to levels above on a read as a temporary solution for L2's in FS
tests/quick/00.hello/ref/alpha/linux/o3-timing/m5stats.txt:
tests/quick/00.hello/ref/alpha/linux/simple-timing/m5stats.txt:
tests/quick/01.hello-2T-smt/ref/alpha/linux/o3-timing/m5stats.txt:
    Update ref's for writeback changes

--HG--
extra : convert_revision : 937febd577b16b7fd97a5a68acaf53541828a251

8 files changed, 119 insertions, 77 deletions

diff --git a/src/mem/bus.cc b/src/mem/bus.cc
index 6b5b63f50..e9a870b80 100644
--- a/src/mem/bus.cc
+++ b/src/mem/bus.cc
@@ -296,7 +296,7 @@ Bus::findPort(Addr addr, int id)
 
 
     // we shouldn't be sending this back to where it came from
-    // only on a functional access and then we should terminate
+    // do the snoop access and then we should terminate
     // the cyclical call.
     if (dest_id == id)
         return 0;
diff --git a/src/mem/cache/cache_impl.hh b/src/mem/cache/cache_impl.hh
index df59b0a4f..3a681bc52 100644
--- a/src/mem/cache/cache_impl.hh
+++ b/src/mem/cache/cache_impl.hh
@@ -391,7 +391,13 @@ Cache<TagStore,Buffering,Coherence>::snoop(PacketPtr &pkt)
     }
 
     //Send a timing (true) invalidate up if the protocol calls for it
-    coherence->propogateInvalidate(pkt, true);
+    if (coherence->propogateInvalidate(pkt, true)) {
+        //Temp hack, we had a functional read hit in the L1, mark as success
+        pkt->flags |= SATISFIED;
+        pkt->result = Packet::Success;
+        respondToSnoop(pkt, curTick + hitLatency);
+        return;
+    }
 
     Addr blk_addr = pkt->getAddr() & ~(Addr(blkSize-1));
     BlkType *blk = tags->findBlock(pkt);
@@ -562,6 +568,7 @@ Cache<TagStore,Buffering,Coherence>::probe(PacketPtr &pkt, bool update,
 
     PacketList writebacks;
     int lat;
+
     BlkType *blk = tags->handleAccess(pkt, lat, writebacks, update);
 
     DPRINTF(Cache, "%s %x %s\n", pkt->cmdString(),
@@ -615,7 +622,8 @@ Cache<TagStore,Buffering,Coherence>::probe(PacketPtr &pkt, bool update,
             // Can't handle it, return request unsatisfied.
             panic("Atomic access ran into outstanding MSHR's or WB's!");
         }
-        if (!pkt->req->isUncacheable()) {
+        if (!pkt->req->isUncacheable() /*Uncacheables just go through*/
+            && (pkt->cmd != Packet::Writeback)/*Writebacks on miss fall through*/) {
                 // Fetch the cache block to fill
             BlkType *blk = tags->findBlock(pkt);
             Packet::Command temp_cmd = coherence->getBusCmd(pkt->cmd,
@@ -691,7 +699,12 @@ Tick
 Cache<TagStore,Buffering,Coherence>::snoopProbe(PacketPtr &pkt)
 {
     //Send a atomic (false) invalidate up if the protocol calls for it
-    coherence->propogateInvalidate(pkt, false);
+    if (coherence->propogateInvalidate(pkt, false)) {
+        //Temp hack, we had a functional read hit in the L1, mark as success
+        pkt->flags |= SATISFIED;
+        pkt->result = Packet::Success;
+        return hitLatency;
+    }
 
     Addr blk_addr = pkt->getAddr() & ~(Addr(blkSize-1));
     BlkType *blk = tags->findBlock(pkt);
diff --git a/src/mem/cache/coherence/simple_coherence.hh b/src/mem/cache/coherence/simple_coherence.hh
index 5316e64b9..a1fd33080 100644
--- a/src/mem/cache/coherence/simple_coherence.hh
+++ b/src/mem/cache/coherence/simple_coherence.hh
@@ -161,10 +161,10 @@ class SimpleCoherence
 
     bool hasProtocol() { return true; }
 
-    void propogateInvalidate(PacketPtr pkt, bool isTiming)
+    bool propogateInvalidate(PacketPtr pkt, bool isTiming)
     {
         //For now we do nothing, asssumes simple coherence is top level of cache
-        return;
+        return false;
     }
 };
 
diff --git a/src/mem/cache/coherence/uni_coherence.cc b/src/mem/cache/coherence/uni_coherence.cc
index 19230e35b..5813a0281 100644
--- a/src/mem/cache/coherence/uni_coherence.cc
+++ b/src/mem/cache/coherence/uni_coherence.cc
@@ -54,6 +54,7 @@ UniCoherence::sendResult(PacketPtr &pkt, MSHR* cshr, bool success)
     {
         bool unblock = cshrs.isFull();
 //        cshrs.markInService(cshr);
+        delete pkt->req;
         cshrs.deallocate(cshr);
         if (!cshrs.havePending()) {
             cache->clearSlaveRequest(Request_Coherence);
@@ -81,17 +82,28 @@ UniCoherence::handleBusRequest(PacketPtr &pkt, CacheBlk *blk, MSHR *mshr,
     }
     else if (blk) {
         new_state = blk->status;
+        if (pkt->isRead()) {
+            DPRINTF(Cache, "Uni-coherence snoops a read that hit in itself"
+                    ". Should satisfy the packet\n");
+            return true; //Satisfy Reads if we can
+        }
     }
     return false;
 }
 
-void
+bool
 UniCoherence::propogateInvalidate(PacketPtr pkt, bool isTiming)
 {
+    //Make sure we don't snoop a write
+    //we are expecting writeInvalidates on the snoop port of a uni-coherent cache
+    assert(!(!pkt->isInvalidate() && pkt->isWrite()));
+
     if (pkt->isInvalidate()) {
+/*  Temp Fix for now, forward all invalidates up as functional accesses */
         if (isTiming) {
             // Forward to other caches
-            PacketPtr tmp = new Packet(pkt->req, Packet::InvalidateReq, -1);
+            Request* req = new Request(pkt->req->getPaddr(), pkt->getSize(), 0);
+            PacketPtr tmp = new Packet(req, Packet::InvalidateReq, -1);
             cshrs.allocate(tmp);
             cache->setSlaveRequest(Request_Coherence, curTick);
             if (cshrs.isFull())
@@ -102,5 +114,26 @@ UniCoherence::propogateInvalidate(PacketPtr pkt, bool isTiming)
             cache->cpuSidePort->sendAtomic(tmp);
             delete tmp;
         }
+/**/
+/*            PacketPtr tmp = new Packet(pkt->req, Packet::InvalidateReq, -1);
+            cache->cpuSidePort->sendFunctional(tmp);
+            delete tmp;
+*/
+    }
+    if (pkt->isRead()) {
+        /*For now we will see if someone above us has the data by
+          doing a functional access on reads.  Fix this later */
+            PacketPtr tmp = new Packet(pkt->req, Packet::ReadReq, -1);
+            tmp->allocate();
+            cache->cpuSidePort->sendFunctional(tmp);
+            bool hit = (tmp->result == Packet::Success);
+            if (hit) {
+                memcpy(pkt->getPtr<uint8_t>(), tmp->getPtr<uint8_t>(),
+                       pkt->getSize());
+                DPRINTF(Cache, "Uni-coherence snoops a read that hit in L1\n");
+            }
+            delete tmp;
+            return hit;
     }
+    return false;
 }
diff --git a/src/mem/cache/coherence/uni_coherence.hh b/src/mem/cache/coherence/uni_coherence.hh
index 44c752088..9a4aacdec 100644
--- a/src/mem/cache/coherence/uni_coherence.hh
+++ b/src/mem/cache/coherence/uni_coherence.hh
@@ -140,7 +140,7 @@ class UniCoherence
 
     bool hasProtocol() { return false; }
 
-    void propogateInvalidate(PacketPtr pkt, bool isTiming);
+    bool propogateInvalidate(PacketPtr pkt, bool isTiming);
 };
 
 #endif //__UNI_COHERENCE_HH__
diff --git a/tests/quick/00.hello/ref/alpha/linux/o3-timing/m5stats.txt b/tests/quick/00.hello/ref/alpha/linux/o3-timing/m5stats.txt
index d34c19255..8b8a25405 100644
--- a/tests/quick/00.hello/ref/alpha/linux/o3-timing/m5stats.txt
+++ b/tests/quick/00.hello/ref/alpha/linux/o3-timing/m5stats.txt
@@ -8,10 +8,10 @@ global.BPredUnit.condIncorrect                    437                       # Nu
 global.BPredUnit.condPredicted                   1563                       # Number of conditional branches predicted
 global.BPredUnit.lookups                         5229                       # Number of BP lookups
 global.BPredUnit.usedRAS                         2821                       # Number of times the RAS was used to get a target.
-host_inst_rate                                  15039                       # Simulator instruction rate (inst/s)
-host_mem_usage                                 180156                       # Number of bytes of host memory used
-host_seconds                                     0.37                       # Real time elapsed on the host
-host_tick_rate                                3741816                       # Simulator tick rate (ticks/s)
+host_inst_rate                                  15743                       # Simulator instruction rate (inst/s)
+host_mem_usage                                 180184                       # Number of bytes of host memory used
+host_seconds                                     0.36                       # Real time elapsed on the host
+host_tick_rate                                3916768                       # Simulator tick rate (ticks/s)
 memdepunit.memDep.conflictingLoads                 23                       # Number of conflicting loads.
 memdepunit.memDep.conflictingStores               117                       # Number of conflicting stores.
 memdepunit.memDep.insertedLoads                  3775                       # Number of loads inserted to the mem dependence unit.
@@ -73,7 +73,7 @@ system.cpu.dcache.WriteReq_mshr_hits              181                       # nu
 system.cpu.dcache.WriteReq_mshr_miss_latency       375299                       # number of WriteReq MSHR miss cycles
 system.cpu.dcache.WriteReq_mshr_miss_rate     0.089901                       # mshr miss rate for WriteReq accesses
 system.cpu.dcache.WriteReq_mshr_misses             73                       # number of WriteReq MSHR misses
-system.cpu.dcache.avg_blocked_cycles_no_mshrs <err: div-0>                       # average number of cycles each access was blocked
+system.cpu.dcache.avg_blocked_cycles_no_mshrs     no value                       # average number of cycles each access was blocked
 system.cpu.dcache.avg_blocked_cycles_no_targets  3366.651163                       # average number of cycles each access was blocked
 system.cpu.dcache.avg_refs                  11.587209                       # Average number of references to valid blocks.
 system.cpu.dcache.blocked_no_mshrs                  0                       # number of cycles access was blocked
@@ -325,49 +325,48 @@ system.cpu.iq.iqSquashedInstsExamined           13381                       # Nu
 system.cpu.iq.iqSquashedInstsIssued                72                       # Number of squashed instructions issued
 system.cpu.iq.iqSquashedNonSpecRemoved              7                       # Number of squashed non-spec instructions that were removed
 system.cpu.iq.iqSquashedOperandsExamined         9575                       # Number of squashed operands that are examined and possibly removed from graph
-system.cpu.l2cache.ReadReq_accesses               482                       # number of ReadReq accesses(hits+misses)
+system.cpu.l2cache.ReadReq_accesses               480                       # number of ReadReq accesses(hits+misses)
 system.cpu.l2cache.ReadReq_avg_miss_latency  4520.691667                       # average ReadReq miss latency
 system.cpu.l2cache.ReadReq_avg_mshr_miss_latency  2303.372917                       # average ReadReq mshr miss latency
-system.cpu.l2cache.ReadReq_hits                     2                       # number of ReadReq hits
 system.cpu.l2cache.ReadReq_miss_latency       2169932                       # number of ReadReq miss cycles
-system.cpu.l2cache.ReadReq_miss_rate         0.995851                       # miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_miss_rate                1                       # miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_misses                 480                       # number of ReadReq misses
 system.cpu.l2cache.ReadReq_mshr_miss_latency      1105619                       # number of ReadReq MSHR miss cycles
-system.cpu.l2cache.ReadReq_mshr_miss_rate     0.995851                       # mshr miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_mshr_miss_rate            1                       # mshr miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_mshr_misses            480                       # number of ReadReq MSHR misses
 system.cpu.l2cache.avg_blocked_cycles_no_mshrs <err: div-0>                       # average number of cycles each access was blocked
 system.cpu.l2cache.avg_blocked_cycles_no_targets <err: div-0>                       # average number of cycles each access was blocked
-system.cpu.l2cache.avg_refs                  0.004167                       # Average number of references to valid blocks.
+system.cpu.l2cache.avg_refs                         0                       # Average number of references to valid blocks.
 system.cpu.l2cache.blocked_no_mshrs                 0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_no_targets               0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_mshrs            0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_targets            0                       # number of cycles access was blocked
 system.cpu.l2cache.cache_copies                     0                       # number of cache copies performed
-system.cpu.l2cache.demand_accesses                482                       # number of demand (read+write) accesses
+system.cpu.l2cache.demand_accesses                480                       # number of demand (read+write) accesses
 system.cpu.l2cache.demand_avg_miss_latency  4520.691667                       # average overall miss latency
 system.cpu.l2cache.demand_avg_mshr_miss_latency  2303.372917                       # average overall mshr miss latency
-system.cpu.l2cache.demand_hits                      2                       # number of demand (read+write) hits
+system.cpu.l2cache.demand_hits                      0                       # number of demand (read+write) hits
 system.cpu.l2cache.demand_miss_latency        2169932                       # number of demand (read+write) miss cycles
-system.cpu.l2cache.demand_miss_rate          0.995851                       # miss rate for demand accesses
+system.cpu.l2cache.demand_miss_rate                 1                       # miss rate for demand accesses
 system.cpu.l2cache.demand_misses                  480                       # number of demand (read+write) misses
 system.cpu.l2cache.demand_mshr_hits                 0                       # number of demand (read+write) MSHR hits
 system.cpu.l2cache.demand_mshr_miss_latency      1105619                       # number of demand (read+write) MSHR miss cycles
-system.cpu.l2cache.demand_mshr_miss_rate     0.995851                       # mshr miss rate for demand accesses
+system.cpu.l2cache.demand_mshr_miss_rate            1                       # mshr miss rate for demand accesses
 system.cpu.l2cache.demand_mshr_misses             480                       # number of demand (read+write) MSHR misses
 system.cpu.l2cache.fast_writes                      0                       # number of fast writes performed
 system.cpu.l2cache.mshr_cap_events                  0                       # number of times MSHR cap was activated
 system.cpu.l2cache.no_allocate_misses               0                       # Number of misses that were no-allocate
-system.cpu.l2cache.overall_accesses               482                       # number of overall (read+write) accesses
+system.cpu.l2cache.overall_accesses               480                       # number of overall (read+write) accesses
 system.cpu.l2cache.overall_avg_miss_latency  4520.691667                       # average overall miss latency
 system.cpu.l2cache.overall_avg_mshr_miss_latency  2303.372917                       # average overall mshr miss latency
 system.cpu.l2cache.overall_avg_mshr_uncacheable_latency <err: div-0>                       # average overall mshr uncacheable latency
-system.cpu.l2cache.overall_hits                     2                       # number of overall hits
+system.cpu.l2cache.overall_hits                     0                       # number of overall hits
 system.cpu.l2cache.overall_miss_latency       2169932                       # number of overall miss cycles
-system.cpu.l2cache.overall_miss_rate         0.995851                       # miss rate for overall accesses
+system.cpu.l2cache.overall_miss_rate                1                       # miss rate for overall accesses
 system.cpu.l2cache.overall_misses                 480                       # number of overall misses
 system.cpu.l2cache.overall_mshr_hits                0                       # number of overall MSHR hits
 system.cpu.l2cache.overall_mshr_miss_latency      1105619                       # number of overall MSHR miss cycles
-system.cpu.l2cache.overall_mshr_miss_rate     0.995851                       # mshr miss rate for overall accesses
+system.cpu.l2cache.overall_mshr_miss_rate            1                       # mshr miss rate for overall accesses
 system.cpu.l2cache.overall_mshr_misses            480                       # number of overall MSHR misses
 system.cpu.l2cache.overall_mshr_uncacheable_latency            0                       # number of overall MSHR uncacheable cycles
 system.cpu.l2cache.overall_mshr_uncacheable_misses            0                       # number of overall MSHR uncacheable misses
@@ -384,7 +383,7 @@ system.cpu.l2cache.replacements                     0                       # nu
 system.cpu.l2cache.sampled_refs                   480                       # Sample count of references to valid blocks.
 system.cpu.l2cache.soft_prefetch_mshr_full            0                       # number of mshr full events for SW prefetching instrutions
 system.cpu.l2cache.tagsinuse               248.469634                       # Cycle average of tags in use
-system.cpu.l2cache.total_refs                       2                       # Total number of references to valid blocks.
+system.cpu.l2cache.total_refs                       0                       # Total number of references to valid blocks.
 system.cpu.l2cache.warmup_cycle                     0                       # Cycle when the warmup percentage was hit.
 system.cpu.l2cache.writebacks                       0                       # number of writebacks
 system.cpu.numCycles                            53773                       # number of cpu cycles simulated
diff --git a/tests/quick/00.hello/ref/alpha/linux/simple-timing/m5stats.txt b/tests/quick/00.hello/ref/alpha/linux/simple-timing/m5stats.txt
index 27822f334..68f6bcca4 100644
--- a/tests/quick/00.hello/ref/alpha/linux/simple-timing/m5stats.txt
+++ b/tests/quick/00.hello/ref/alpha/linux/simple-timing/m5stats.txt
@@ -1,9 +1,9 @@
 
 ---------- Begin Simulation Statistics ----------
-host_inst_rate                                 179790                       # Simulator instruction rate (inst/s)
-host_mem_usage                                 179436                       # Number of bytes of host memory used
+host_inst_rate                                 167195                       # Simulator instruction rate (inst/s)
+host_mem_usage                                 179768                       # Number of bytes of host memory used
 host_seconds                                     0.03                       # Real time elapsed on the host
-host_tick_rate                               55533187                       # Simulator tick rate (ticks/s)
+host_tick_rate                               51710933                       # Simulator tick rate (ticks/s)
 sim_freq                                 1000000000000                       # Frequency of simulated ticks
 sim_insts                                        5642                       # Number of instructions simulated
 sim_seconds                                  0.000002                       # Number of seconds simulated
@@ -143,49 +143,48 @@ system.cpu.icache.total_refs                     5366                       # To
 system.cpu.icache.warmup_cycle                      0                       # Cycle when the warmup percentage was hit.
 system.cpu.icache.writebacks                        0                       # number of writebacks
 system.cpu.idle_fraction                            0                       # Percentage of idle cycles
-system.cpu.l2cache.ReadReq_accesses               442                       # number of ReadReq accesses(hits+misses)
+system.cpu.l2cache.ReadReq_accesses               441                       # number of ReadReq accesses(hits+misses)
 system.cpu.l2cache.ReadReq_avg_miss_latency  2984.340136                       # average ReadReq miss latency
 system.cpu.l2cache.ReadReq_avg_mshr_miss_latency  1983.340136                       # average ReadReq mshr miss latency
-system.cpu.l2cache.ReadReq_hits                     1                       # number of ReadReq hits
 system.cpu.l2cache.ReadReq_miss_latency       1316094                       # number of ReadReq miss cycles
-system.cpu.l2cache.ReadReq_miss_rate         0.997738                       # miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_miss_rate                1                       # miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_misses                 441                       # number of ReadReq misses
 system.cpu.l2cache.ReadReq_mshr_miss_latency       874653                       # number of ReadReq MSHR miss cycles
-system.cpu.l2cache.ReadReq_mshr_miss_rate     0.997738                       # mshr miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_mshr_miss_rate            1                       # mshr miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_mshr_misses            441                       # number of ReadReq MSHR misses
 system.cpu.l2cache.avg_blocked_cycles_no_mshrs <err: div-0>                       # average number of cycles each access was blocked
 system.cpu.l2cache.avg_blocked_cycles_no_targets <err: div-0>                       # average number of cycles each access was blocked
-system.cpu.l2cache.avg_refs                  0.002268                       # Average number of references to valid blocks.
+system.cpu.l2cache.avg_refs                         0                       # Average number of references to valid blocks.
 system.cpu.l2cache.blocked_no_mshrs                 0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_no_targets               0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_mshrs            0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_targets            0                       # number of cycles access was blocked
 system.cpu.l2cache.cache_copies                     0                       # number of cache copies performed
-system.cpu.l2cache.demand_accesses                442                       # number of demand (read+write) accesses
+system.cpu.l2cache.demand_accesses                441                       # number of demand (read+write) accesses
 system.cpu.l2cache.demand_avg_miss_latency  2984.340136                       # average overall miss latency
 system.cpu.l2cache.demand_avg_mshr_miss_latency  1983.340136                       # average overall mshr miss latency
-system.cpu.l2cache.demand_hits                      1                       # number of demand (read+write) hits
+system.cpu.l2cache.demand_hits                      0                       # number of demand (read+write) hits
 system.cpu.l2cache.demand_miss_latency        1316094                       # number of demand (read+write) miss cycles
-system.cpu.l2cache.demand_miss_rate          0.997738                       # miss rate for demand accesses
+system.cpu.l2cache.demand_miss_rate                 1                       # miss rate for demand accesses
 system.cpu.l2cache.demand_misses                  441                       # number of demand (read+write) misses
 system.cpu.l2cache.demand_mshr_hits                 0                       # number of demand (read+write) MSHR hits
 system.cpu.l2cache.demand_mshr_miss_latency       874653                       # number of demand (read+write) MSHR miss cycles
-system.cpu.l2cache.demand_mshr_miss_rate     0.997738                       # mshr miss rate for demand accesses
+system.cpu.l2cache.demand_mshr_miss_rate            1                       # mshr miss rate for demand accesses
 system.cpu.l2cache.demand_mshr_misses             441                       # number of demand (read+write) MSHR misses
 system.cpu.l2cache.fast_writes                      0                       # number of fast writes performed
 system.cpu.l2cache.mshr_cap_events                  0                       # number of times MSHR cap was activated
 system.cpu.l2cache.no_allocate_misses               0                       # Number of misses that were no-allocate
-system.cpu.l2cache.overall_accesses               442                       # number of overall (read+write) accesses
+system.cpu.l2cache.overall_accesses               441                       # number of overall (read+write) accesses
 system.cpu.l2cache.overall_avg_miss_latency  2984.340136                       # average overall miss latency
 system.cpu.l2cache.overall_avg_mshr_miss_latency  1983.340136                       # average overall mshr miss latency
-system.cpu.l2cache.overall_avg_mshr_uncacheable_latency <err: div-0>                       # average overall mshr uncacheable latency
-system.cpu.l2cache.overall_hits                     1                       # number of overall hits
+system.cpu.l2cache.overall_avg_mshr_uncacheable_latency     no value                       # average overall mshr uncacheable latency
+system.cpu.l2cache.overall_hits                     0                       # number of overall hits
 system.cpu.l2cache.overall_miss_latency       1316094                       # number of overall miss cycles
-system.cpu.l2cache.overall_miss_rate         0.997738                       # miss rate for overall accesses
+system.cpu.l2cache.overall_miss_rate                1                       # miss rate for overall accesses
 system.cpu.l2cache.overall_misses                 441                       # number of overall misses
 system.cpu.l2cache.overall_mshr_hits                0                       # number of overall MSHR hits
 system.cpu.l2cache.overall_mshr_miss_latency       874653                       # number of overall MSHR miss cycles
-system.cpu.l2cache.overall_mshr_miss_rate     0.997738                       # mshr miss rate for overall accesses
+system.cpu.l2cache.overall_mshr_miss_rate            1                       # mshr miss rate for overall accesses
 system.cpu.l2cache.overall_mshr_misses            441                       # number of overall MSHR misses
 system.cpu.l2cache.overall_mshr_uncacheable_latency            0                       # number of overall MSHR uncacheable cycles
 system.cpu.l2cache.overall_mshr_uncacheable_misses            0                       # number of overall MSHR uncacheable misses
@@ -202,7 +201,7 @@ system.cpu.l2cache.replacements                     0                       # nu
 system.cpu.l2cache.sampled_refs                   441                       # Sample count of references to valid blocks.
 system.cpu.l2cache.soft_prefetch_mshr_full            0                       # number of mshr full events for SW prefetching instrutions
 system.cpu.l2cache.tagsinuse               220.802916                       # Cycle average of tags in use
-system.cpu.l2cache.total_refs                       1                       # Total number of references to valid blocks.
+system.cpu.l2cache.total_refs                       0                       # Total number of references to valid blocks.
 system.cpu.l2cache.warmup_cycle                     0                       # Cycle when the warmup percentage was hit.
 system.cpu.l2cache.writebacks                       0                       # number of writebacks
 system.cpu.not_idle_fraction                        1                       # Percentage of non-idle cycles
diff --git a/tests/quick/01.hello-2T-smt/ref/alpha/linux/o3-timing/m5stats.txt b/tests/quick/01.hello-2T-smt/ref/alpha/linux/o3-timing/m5stats.txt
index aa4877339..365f828db 100644
--- a/tests/quick/01.hello-2T-smt/ref/alpha/linux/o3-timing/m5stats.txt
+++ b/tests/quick/01.hello-2T-smt/ref/alpha/linux/o3-timing/m5stats.txt
@@ -8,10 +8,10 @@ global.BPredUnit.condIncorrect                   1201                       # Nu
 global.BPredUnit.condPredicted                   4031                       # Number of conditional branches predicted
 global.BPredUnit.lookups                        12370                       # Number of BP lookups
 global.BPredUnit.usedRAS                         6337                       # Number of times the RAS was used to get a target.
-host_inst_rate                                   9282                       # Simulator instruction rate (inst/s)
-host_mem_usage                                 181628                       # Number of bytes of host memory used
-host_seconds                                     1.21                       # Real time elapsed on the host
-host_tick_rate                                1845830                       # Simulator tick rate (ticks/s)
+host_inst_rate                                   9475                       # Simulator instruction rate (inst/s)
+host_mem_usage                                 181200                       # Number of bytes of host memory used
+host_seconds                                     1.19                       # Real time elapsed on the host
+host_tick_rate                                1884343                       # Simulator tick rate (ticks/s)
 memdepunit.memDep.conflictingLoads                 27                       # Number of conflicting loads.
 memdepunit.memDep.conflictingLoads                 20                       # Number of conflicting loads.
 memdepunit.memDep.conflictingStores                97                       # Number of conflicting stores.
@@ -571,36 +571,34 @@ system.cpu.iq.iqSquashedInstsExamined           27426                       # Nu
 system.cpu.iq.iqSquashedInstsIssued               185                       # Number of squashed instructions issued
 system.cpu.iq.iqSquashedNonSpecRemoved              8                       # Number of squashed non-spec instructions that were removed
 system.cpu.iq.iqSquashedOperandsExamined        20011                       # Number of squashed operands that are examined and possibly removed from graph
-system.cpu.l2cache.ReadReq_accesses               975                       # number of ReadReq accesses(hits+misses)
-system.cpu.l2cache.ReadReq_accesses_0             975                       # number of ReadReq accesses(hits+misses)
+system.cpu.l2cache.ReadReq_accesses               973                       # number of ReadReq accesses(hits+misses)
+system.cpu.l2cache.ReadReq_accesses_0             973                       # number of ReadReq accesses(hits+misses)
 system.cpu.l2cache.ReadReq_avg_miss_latency  6750.932169                       # average ReadReq miss latency
 system.cpu.l2cache.ReadReq_avg_miss_latency_0  6750.932169                       # average ReadReq miss latency
 system.cpu.l2cache.ReadReq_avg_mshr_miss_latency  3603.773895                       # average ReadReq mshr miss latency
 system.cpu.l2cache.ReadReq_avg_mshr_miss_latency_0  3603.773895                       # average ReadReq mshr miss latency
-system.cpu.l2cache.ReadReq_hits                     2                       # number of ReadReq hits
-system.cpu.l2cache.ReadReq_hits_0                   2                       # number of ReadReq hits
 system.cpu.l2cache.ReadReq_miss_latency       6568657                       # number of ReadReq miss cycles
 system.cpu.l2cache.ReadReq_miss_latency_0      6568657                       # number of ReadReq miss cycles
-system.cpu.l2cache.ReadReq_miss_rate         0.997949                       # miss rate for ReadReq accesses
-system.cpu.l2cache.ReadReq_miss_rate_0       0.997949                       # miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_miss_rate                1                       # miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_miss_rate_0              1                       # miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_misses                 973                       # number of ReadReq misses
 system.cpu.l2cache.ReadReq_misses_0               973                       # number of ReadReq misses
 system.cpu.l2cache.ReadReq_mshr_miss_latency      3506472                       # number of ReadReq MSHR miss cycles
 system.cpu.l2cache.ReadReq_mshr_miss_latency_0      3506472                       # number of ReadReq MSHR miss cycles
-system.cpu.l2cache.ReadReq_mshr_miss_rate     0.997949                       # mshr miss rate for ReadReq accesses
-system.cpu.l2cache.ReadReq_mshr_miss_rate_0     0.997949                       # mshr miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_mshr_miss_rate            1                       # mshr miss rate for ReadReq accesses
+system.cpu.l2cache.ReadReq_mshr_miss_rate_0            1                       # mshr miss rate for ReadReq accesses
 system.cpu.l2cache.ReadReq_mshr_misses            973                       # number of ReadReq MSHR misses
 system.cpu.l2cache.ReadReq_mshr_misses_0          973                       # number of ReadReq MSHR misses
 system.cpu.l2cache.avg_blocked_cycles_no_mshrs <err: div-0>                       # average number of cycles each access was blocked
 system.cpu.l2cache.avg_blocked_cycles_no_targets <err: div-0>                       # average number of cycles each access was blocked
-system.cpu.l2cache.avg_refs                  0.002055                       # Average number of references to valid blocks.
+system.cpu.l2cache.avg_refs                         0                       # Average number of references to valid blocks.
 system.cpu.l2cache.blocked_no_mshrs                 0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_no_targets               0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_mshrs            0                       # number of cycles access was blocked
 system.cpu.l2cache.blocked_cycles_no_targets            0                       # number of cycles access was blocked
 system.cpu.l2cache.cache_copies                     0                       # number of cache copies performed
-system.cpu.l2cache.demand_accesses                975                       # number of demand (read+write) accesses
-system.cpu.l2cache.demand_accesses_0              975                       # number of demand (read+write) accesses
+system.cpu.l2cache.demand_accesses                973                       # number of demand (read+write) accesses
+system.cpu.l2cache.demand_accesses_0              973                       # number of demand (read+write) accesses
 system.cpu.l2cache.demand_accesses_1                0                       # number of demand (read+write) accesses
 system.cpu.l2cache.demand_avg_miss_latency  6750.932169                       # average overall miss latency
 system.cpu.l2cache.demand_avg_miss_latency_0  6750.932169                       # average overall miss latency
@@ -608,14 +606,14 @@ system.cpu.l2cache.demand_avg_miss_latency_1 <err: div-0>
 system.cpu.l2cache.demand_avg_mshr_miss_latency  3603.773895                       # average overall mshr miss latency
 system.cpu.l2cache.demand_avg_mshr_miss_latency_0  3603.773895                       # average overall mshr miss latency
 system.cpu.l2cache.demand_avg_mshr_miss_latency_1 <err: div-0>                       # average overall mshr miss latency
-system.cpu.l2cache.demand_hits                      2                       # number of demand (read+write) hits
-system.cpu.l2cache.demand_hits_0                    2                       # number of demand (read+write) hits
+system.cpu.l2cache.demand_hits                      0                       # number of demand (read+write) hits
+system.cpu.l2cache.demand_hits_0                    0                       # number of demand (read+write) hits
 system.cpu.l2cache.demand_hits_1                    0                       # number of demand (read+write) hits
 system.cpu.l2cache.demand_miss_latency        6568657                       # number of demand (read+write) miss cycles
 system.cpu.l2cache.demand_miss_latency_0      6568657                       # number of demand (read+write) miss cycles
 system.cpu.l2cache.demand_miss_latency_1            0                       # number of demand (read+write) miss cycles
-system.cpu.l2cache.demand_miss_rate          0.997949                       # miss rate for demand accesses
-system.cpu.l2cache.demand_miss_rate_0        0.997949                       # miss rate for demand accesses
+system.cpu.l2cache.demand_miss_rate                 1                       # miss rate for demand accesses
+system.cpu.l2cache.demand_miss_rate_0               1                       # miss rate for demand accesses
 system.cpu.l2cache.demand_miss_rate_1    <err: div-0>                       # miss rate for demand accesses
 system.cpu.l2cache.demand_misses                  973                       # number of demand (read+write) misses
 system.cpu.l2cache.demand_misses_0                973                       # number of demand (read+write) misses
@@ -626,8 +624,8 @@ system.cpu.l2cache.demand_mshr_hits_1               0                       # nu
 system.cpu.l2cache.demand_mshr_miss_latency      3506472                       # number of demand (read+write) MSHR miss cycles
 system.cpu.l2cache.demand_mshr_miss_latency_0      3506472                       # number of demand (read+write) MSHR miss cycles
 system.cpu.l2cache.demand_mshr_miss_latency_1            0                       # number of demand (read+write) MSHR miss cycles
-system.cpu.l2cache.demand_mshr_miss_rate     0.997949                       # mshr miss rate for demand accesses
-system.cpu.l2cache.demand_mshr_miss_rate_0     0.997949                       # mshr miss rate for demand accesses
+system.cpu.l2cache.demand_mshr_miss_rate            1                       # mshr miss rate for demand accesses
+system.cpu.l2cache.demand_mshr_miss_rate_0            1                       # mshr miss rate for demand accesses
 system.cpu.l2cache.demand_mshr_miss_rate_1 <err: div-0>                       # mshr miss rate for demand accesses
 system.cpu.l2cache.demand_mshr_misses             973                       # number of demand (read+write) MSHR misses
 system.cpu.l2cache.demand_mshr_misses_0           973                       # number of demand (read+write) MSHR misses
@@ -637,8 +635,8 @@ system.cpu.l2cache.mshr_cap_events                  0                       # nu
 system.cpu.l2cache.mshr_cap_events_0                0                       # number of times MSHR cap was activated
 system.cpu.l2cache.mshr_cap_events_1                0                       # number of times MSHR cap was activated
 system.cpu.l2cache.no_allocate_misses               0                       # Number of misses that were no-allocate
-system.cpu.l2cache.overall_accesses               975                       # number of overall (read+write) accesses
-system.cpu.l2cache.overall_accesses_0             975                       # number of overall (read+write) accesses
+system.cpu.l2cache.overall_accesses               973                       # number of overall (read+write) accesses
+system.cpu.l2cache.overall_accesses_0             973                       # number of overall (read+write) accesses
 system.cpu.l2cache.overall_accesses_1               0                       # number of overall (read+write) accesses
 system.cpu.l2cache.overall_avg_miss_latency  6750.932169                       # average overall miss latency
 system.cpu.l2cache.overall_avg_miss_latency_0  6750.932169                       # average overall miss latency
@@ -649,15 +647,15 @@ system.cpu.l2cache.overall_avg_mshr_miss_latency_1 <err: div-0>
 system.cpu.l2cache.overall_avg_mshr_uncacheable_latency <err: div-0>                       # average overall mshr uncacheable latency
 system.cpu.l2cache.overall_avg_mshr_uncacheable_latency_0 <err: div-0>                       # average overall mshr uncacheable latency
 system.cpu.l2cache.overall_avg_mshr_uncacheable_latency_1 <err: div-0>                       # average overall mshr uncacheable latency
-system.cpu.l2cache.overall_hits                     2                       # number of overall hits
-system.cpu.l2cache.overall_hits_0                   2                       # number of overall hits
+system.cpu.l2cache.overall_hits                     0                       # number of overall hits
+system.cpu.l2cache.overall_hits_0                   0                       # number of overall hits
 system.cpu.l2cache.overall_hits_1                   0                       # number of overall hits
 system.cpu.l2cache.overall_miss_latency       6568657                       # number of overall miss cycles
 system.cpu.l2cache.overall_miss_latency_0      6568657                       # number of overall miss cycles
 system.cpu.l2cache.overall_miss_latency_1            0                       # number of overall miss cycles
-system.cpu.l2cache.overall_miss_rate         0.997949                       # miss rate for overall accesses
-system.cpu.l2cache.overall_miss_rate_0       0.997949                       # miss rate for overall accesses
-system.cpu.l2cache.overall_miss_rate_1       no value                       # miss rate for overall accesses
+system.cpu.l2cache.overall_miss_rate                1                       # miss rate for overall accesses
+system.cpu.l2cache.overall_miss_rate_0              1                       # miss rate for overall accesses
+system.cpu.l2cache.overall_miss_rate_1   <err: div-0>                       # miss rate for overall accesses
 system.cpu.l2cache.overall_misses                 973                       # number of overall misses
 system.cpu.l2cache.overall_misses_0               973                       # number of overall misses
 system.cpu.l2cache.overall_misses_1                 0                       # number of overall misses
@@ -667,8 +665,8 @@ system.cpu.l2cache.overall_mshr_hits_1              0                       # nu
 system.cpu.l2cache.overall_mshr_miss_latency      3506472                       # number of overall MSHR miss cycles
 system.cpu.l2cache.overall_mshr_miss_latency_0      3506472                       # number of overall MSHR miss cycles
 system.cpu.l2cache.overall_mshr_miss_latency_1            0                       # number of overall MSHR miss cycles
-system.cpu.l2cache.overall_mshr_miss_rate     0.997949                       # mshr miss rate for overall accesses
-system.cpu.l2cache.overall_mshr_miss_rate_0     0.997949                       # mshr miss rate for overall accesses
+system.cpu.l2cache.overall_mshr_miss_rate            1                       # mshr miss rate for overall accesses
+system.cpu.l2cache.overall_mshr_miss_rate_0            1                       # mshr miss rate for overall accesses
 system.cpu.l2cache.overall_mshr_miss_rate_1 <err: div-0>                       # mshr miss rate for overall accesses
 system.cpu.l2cache.overall_mshr_misses            973                       # number of overall MSHR misses
 system.cpu.l2cache.overall_mshr_misses_0          973                       # number of overall MSHR misses
@@ -696,7 +694,7 @@ system.cpu.l2cache.soft_prefetch_mshr_full            0                       #
 system.cpu.l2cache.soft_prefetch_mshr_full_0            0                       # number of mshr full events for SW prefetching instrutions
 system.cpu.l2cache.soft_prefetch_mshr_full_1            0                       # number of mshr full events for SW prefetching instrutions
 system.cpu.l2cache.tagsinuse               489.113488                       # Cycle average of tags in use
-system.cpu.l2cache.total_refs                       2                       # Total number of references to valid blocks.
+system.cpu.l2cache.total_refs                       0                       # Total number of references to valid blocks.
 system.cpu.l2cache.warmup_cycle                     0                       # Cycle when the warmup percentage was hit.
 system.cpu.l2cache.writebacks                       0                       # number of writebacks
 system.cpu.l2cache.writebacks_0                     0                       # number of writebacks