Update for metric calculator to support event options. Consequently, …

…updated perf. groups with event options. Options are not needed anymore in formulas, counter name is sufficient
RRZE-HPC · Nov 30, 2017 · 88dedc3 · 88dedc3
1 parent 89bdb44
commit 88dedc3
Show file tree

Hide file tree

Showing 29 changed files with 143 additions and 112 deletions.
diff --git a/groups/haswell/UOPS_EXEC.txt b/groups/haswell/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/haswell/UOPS_ISSUE.txt b/groups/haswell/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/haswell/UOPS_RETIRE.txt b/groups/haswell/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/haswellEP/CBOX.txt b/groups/haswellEP/CBOX.txt
@@ -47,7 +47,7 @@ Runtime unhalted [s] FIXC1*inverseClock
 Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 LLC misses per instruction (CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0+CBOX15C0+CBOX16C0+CBOX17C0)/FIXC0
-LL2 data written to MEM [MBytes] 1E-6*(CBOX0C1:STATE=0x1+CBOX1C1:STATE=0x1+CBOX2C1:STATE=0x1+CBOX3C1:STATE=0x1+CBOX4C1:STATE=0x1+CBOX5C1:STATE=0x1+CBOX6C1:STATE=0x1+CBOX7C1:STATE=0x1+CBOX8C1:STATE=0x1+CBOX9C1:STATE=0x1+CBOX10C1:STATE=0x1+CBOX11C1:STATE=0x1+CBOX12C1:STATE=0x1+CBOX13C1:STATE=0x1+CBOX14C1:STATE=0x1+CBOX15C1:STATE=0x1+CBOX16C1:STATE=0x1+CBOX17C1:STATE=0x1)*64
+LL2 data written to MEM [MBytes] 1E-6*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1+CBOX15C1+CBOX16C1+CBOX17C1)*64
 
 
 LONG

diff --git a/groups/haswellEP/UOPS_EXEC.txt b/groups/haswellEP/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/haswellEP/UOPS_ISSUE.txt b/groups/haswellEP/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/haswellEP/UOPS_RETIRE.txt b/groups/haswellEP/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridge/UOPS_EXEC.txt b/groups/ivybridge/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridge/UOPS_ISSUE.txt b/groups/ivybridge/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridge/UOPS_RETIRE.txt b/groups/ivybridge/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridgeEP/CACHES.txt b/groups/ivybridgeEP/CACHES.txt
@@ -74,12 +74,12 @@ L2 to L3 evict bandwidth [MBytes/s]  1.0E-06*PMC3*64.0/time
 L2 to L3 evict data volume [GBytes]  1.0E-09*PMC3*64.0
 L2 to/from L3 bandwidth [MBytes/s] 1.0E-06*(PMC2+PMC3)*64.0/time
 L2 to/from L3 data volume [GBytes] 1.0E-09*(PMC2+PMC3)*64.0
-System to L3 bandwidth [MBytes/s] 1.0E-06*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F+CBOX8C0:STATE=0x3F+CBOX9C0:STATE=0x3F+CBOX10C0:STATE=0x3F+CBOX11C0:STATE=0x3F+CBOX12C0:STATE=0x3F+CBOX13C0:STATE=0x3F+CBOX14C0:STATE=0x3F)*64.0/time
-System to L3 data volume [GBytes] 1.0E-09*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F+CBOX8C0:STATE=0x3F+CBOX9C0:STATE=0x3F+CBOX10C0:STATE=0x3F+CBOX11C0:STATE=0x3F+CBOX12C0:STATE=0x3F+CBOX13C0:STATE=0x3F+CBOX14C0:STATE=0x3F)*64.0
+System to L3 bandwidth [MBytes/s] 1.0E-06*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0)*64.0/time
+System to L3 data volume [GBytes] 1.0E-09*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0)*64.0
 L3 to memory bandwidth [MBytes/s] 1.0E-06*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64/time
 L3 to memory data volume [GBytes] 1.0E-09*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64
-L3 to/from system bandwidth [MBytes/s] 1.0E-06*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F+CBOX8C0:STATE=0x3F+CBOX9C0:STATE=0x3F+CBOX10C0:STATE=0x3F+CBOX11C0:STATE=0x3F+CBOX12C0:STATE=0x3F+CBOX13C0:STATE=0x3F+CBOX14C0:STATE=0x3F+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64.0/time
-L3 to/from system data volume [GBytes] 1.0E-09*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F+CBOX8C0:STATE=0x3F+CBOX9C0:STATE=0x3F+CBOX10C0:STATE=0x3F+CBOX11C0:STATE=0x3F+CBOX12C0:STATE=0x3F+CBOX13C0:STATE=0x3F+CBOX14C0:STATE=0x3F+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64.0
+L3 to/from system bandwidth [MBytes/s] 1.0E-06*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64.0/time
+L3 to/from system data volume [GBytes] 1.0E-09*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64.0
 Memory read bandwidth [MBytes/s] 1.0E-06*(MBOX0C0+MBOX1C0+MBOX2C0+MBOX3C0+MBOX4C0+MBOX5C0+MBOX6C0+MBOX7C0)*64.0/time
 Memory read data volume [GBytes] 1.0E-09*(MBOX0C0+MBOX1C0+MBOX2C0+MBOX3C0+MBOX4C0+MBOX5C0+MBOX6C0+MBOX7C0)*64.0
 Memory write bandwidth [MBytes/s] 1.0E-06*(MBOX0C1+MBOX1C1+MBOX2C1+MBOX3C1+MBOX4C1+MBOX5C1+MBOX6C1+MBOX7C1)*64.0/time

diff --git a/groups/ivybridgeEP/CBOX.txt b/groups/ivybridgeEP/CBOX.txt
@@ -41,7 +41,7 @@ Runtime unhalted [s] FIXC1*inverseClock
 Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 LLC misses per instruction (CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0+CBOX8C0+CBOX9C0+CBOX10C0+CBOX11C0+CBOX12C0+CBOX13C0+CBOX14C0)/FIXC0
-LLC data written to MEM [MBytes] 1E-6*(CBOX0C1:STATE=0x1+CBOX1C1:STATE=0x1+CBOX2C1:STATE=0x1+CBOX3C1:STATE=0x1+CBOX4C1:STATE=0x1+CBOX5C1:STATE=0x1+CBOX6C1:STATE=0x1+CBOX7C1:STATE=0x1+CBOX8C1:STATE=0x1+CBOX9C1:STATE=0x1+CBOX10C1:STATE=0x1+CBOX11C1:STATE=0x1+CBOX12C1:STATE=0x1+CBOX13C1:STATE=0x1+CBOX14C1:STATE=0x1)*64
+LLC data written to MEM [MBytes] 1E-6*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1+CBOX8C1+CBOX9C1+CBOX10C1+CBOX11C1+CBOX12C1+CBOX13C1+CBOX14C1)*64
 
 
 LONG

diff --git a/groups/ivybridgeEP/UOPS_EXEC.txt b/groups/ivybridgeEP/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridgeEP/UOPS_ISSUE.txt b/groups/ivybridgeEP/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/ivybridgeEP/UOPS_RETIRE.txt b/groups/ivybridgeEP/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/knl/HBM_OFFCORE.txt b/groups/knl/HBM_OFFCORE.txt
@@ -12,21 +12,21 @@ Runtime (RDTSC) [s] time
 Runtime unhalted [s] FIXC1*inverseClock
 Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
-Memory read bandwidth [MBytes/s] 1.0E-06*(PMC1:MATCH0=0x32F7:MATCH1=0x3F8060)*64.0/time
-Memory read data volume [GBytes] 1.0E-09*(PMC1:MATCH0=0x32F7:MATCH1=0x3F8060)*64.0
-Memory writeback bandwidth [MBytes/s] 1.0E-06*(PMC0:MATCH0=0x4908:MATCH1=0x3F8060)*64.0/time
-Memory writeback data volume [GBytes] 1.0E-09*(PMC0:MATCH0=0x4908:MATCH1=0x3F8060)*64.0
-Memory bandwidth [MBytes/s] 1.0E-06*(PMC0:MATCH0=0x4908:MATCH1=0x3F8060+PMC1:MATCH0=0x32F7:MATCH1=0x3F8060)*64.0/time
-Memory data volume [GBytes] 1.0E-09*(PMC0:MATCH0=0x4908:MATCH1=0x3F8060+PMC1:MATCH0=0x32F7:MATCH1=0x3F8060)*64.0
+Memory read bandwidth [MBytes/s] 1.0E-06*(PMC1)*64.0/time
+Memory read data volume [GBytes] 1.0E-09*(PMC1)*64.0
+Memory writeback bandwidth [MBytes/s] 1.0E-06*(PMC0)*64.0/time
+Memory writeback data volume [GBytes] 1.0E-09*(PMC0)*64.0
+Memory bandwidth [MBytes/s] 1.0E-06*(PMC0+PMC1)*64.0/time
+Memory data volume [GBytes] 1.0E-09*(PMC0+PMC1)*64.0
 
 LONG
 Formulas:
-Memory read bandwidth [MBytes/s] = 1.0E-06*(sum(EDC_RPQ_INSERTS))*64/time
-Memory read data volume [GBytes] = 1.0E-09*(sum(EDC_RPQ_INSERTS))*64
-Memory writeback bandwidth [MBytes/s] = 1.0E-06*(sum(EDC_WPQ_INSERTS))*64/time
-Memory writeback data volume [GBytes] = 1.0E-09*(sum(EDC_WPQ_INSERTS))*64
-Memory bandwidth [MBytes/s] = 1.0E-06*(sum(EDC_RPQ_INSERTS)+sum(EDC_WPQ_INSERTS))*64/time
-Memory data volume [GBytes] = 1.0E-09*(sum(EDC_RPQ_INSERTS)+sum(EDC_WPQ_INSERTS))*64
+Memory read bandwidth [MBytes/s] = 1.0E-06*(sum(OFFCORE_RESPONSE_1_OPTIONS:MATCH0=0x32F7:MATCH1=0x3F8060))*64/time
+Memory read data volume [GBytes] = 1.0E-09*(sum(OFFCORE_RESPONSE_1_OPTIONS:MATCH0=0x32F7:MATCH1=0x3F8060))*64
+Memory writeback bandwidth [MBytes/s] = 1.0E-06*(sum(OFFCORE_RESPONSE_0_OPTIONS:MATCH0=0x4908:MATCH1=0x3F8060))*64/time
+Memory writeback data volume [GBytes] = 1.0E-09*(sum(OFFCORE_RESPONSE_0_OPTIONS:MATCH0=0x4908:MATCH1=0x3F8060))*64
+Memory bandwidth [MBytes/s] = 1.0E-06*(sum(OFFCORE_RESPONSE_1_OPTIONS:MATCH0=0x32F7:MATCH1=0x3F8060)+sum(OFFCORE_RESPONSE_0_OPTIONS:MATCH0=0x4908:MATCH1=0x3F8060))*64/time
+Memory data volume [GBytes] = 1.0E-09*(sum(OFFCORE_RESPONSE_1_OPTIONS:MATCH0=0x32F7:MATCH1=0x3F8060)+sum(OFFCORE_RESPONSE_0_OPTIONS:MATCH0=0x4908:MATCH1=0x3F8060))*64
 -
 Profiling group to measure data transfers from and to the high bandwidth memory (HBM).
-
+If possible, use the HBM or HBM_CACHE group because they provide more accurate counts.
diff --git a/groups/knl/L2.txt b/groups/knl/L2.txt
@@ -14,10 +14,10 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 L2 non-RFO bandwidth [MBytes/s] 1.E-06*(PMC0)*64.0/time
 L2 non-RFO data volume [GByte] 1.E-09*PMC0*64.0
-L2 RFO bandwidth [MBytes/s] 1.E-06*(PMC1:MATCH0=0x0002:MATCH1=0x1)*64.0/time
-L2 RFO data volume [GByte] 1.E-09*(PMC1:MATCH0=0x0002:MATCH1=0x1)*64.0
-L2 bandwidth [MBytes/s] 1.E-06*(PMC0+PMC1:MATCH0=0x0002:MATCH1=0x1)*64.0/time
-L2 data volume [GByte] 1.E-06*(PMC0+PMC1:MATCH0=0x0002:MATCH1=0x1)*64.0
+L2 RFO bandwidth [MBytes/s] 1.E-06*(PMC1)*64.0/time
+L2 RFO data volume [GByte] 1.E-09*(PMC1)*64.0
+L2 bandwidth [MBytes/s] 1.E-06*(PMC0+PMC1)*64.0/time
+L2 data volume [GByte] 1.E-06*(PMC0+PMC1)*64.0
 
 LONG
 Formula:

diff --git a/groups/sandybridge/L3CACHE.txt b/groups/sandybridge/L3CACHE.txt
@@ -13,9 +13,9 @@ Runtime (RDTSC) [s] time
 Runtime unhalted [s] FIXC1*inverseClock
 Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
-L3 request rate PMC1:MATCH0=0x0081:MATCH1=0x1/FIXC0
-L3 miss rate PMC0:MATCH0=0x0081:MATCH1=0x3fffc0/FIXC0
-L3 miss ratio PMC0:MATCH0=0x0081:MATCH1=0x3fffc0/PMC1:MATCH0=0x0081:MATCH1=0x1
+L3 request rate PMC1/FIXC0
+L3 miss rate PMC0/FIXC0
+L3 miss ratio PMC0/PMC1
 
 LONG
 Formulas:

diff --git a/groups/sandybridge/UOPS_EXEC.txt b/groups/sandybridge/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/sandybridge/UOPS_ISSUE.txt b/groups/sandybridge/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/sandybridge/UOPS_RETIRE.txt b/groups/sandybridge/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/sandybridgeEP/CACHES.txt b/groups/sandybridgeEP/CACHES.txt
@@ -50,12 +50,12 @@ L2 to L3 evict bandwidth [MBytes/s]  1.0E-06*PMC3*64.0/time
 L2 to L3 evict data volume [GBytes]  1.0E-09*PMC3*64.0
 L2 to/from L3 bandwidth [MBytes/s] 1.0E-06*(PMC2+PMC3)*64.0/time
 L2 to/from L3 data volume [GBytes] 1.0E-09*(PMC2+PMC3)*64.0
-System to L3 bandwidth [MBytes/s] 1.0E-06*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F)*64.0/time
-System to L3 data volume [GBytes] 1.0E-09*(CBOX0C0:STATE=0x3F+CBOX1C0:STATE=0x3F+CBOX2C0:STATE=0x3F+CBOX3C0:STATE=0x3F+CBOX4C0:STATE=0x3F+CBOX5C0:STATE=0x3F+CBOX6C0:STATE=0x3F+CBOX7C0:STATE=0x3F)*64.0
+System to L3 bandwidth [MBytes/s] 1.0E-06*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0)*64.0/time
+System to L3 data volume [GBytes] 1.0E-09*(CBOX0C0+CBOX1C0+CBOX2C0+CBOX3C0+CBOX4C0+CBOX5C0+CBOX6C0+CBOX7C0)*64.0
 L3 to system bandwidth [MBytes/s] 1.0E-06*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0/time
 L3 to system data volume [GBytes] 1.0E-09*(CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0
-L3 to/from system bandwidth [MBytes/s] 1.0E-06*(CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0/time
-L3 to/from system data volume [GBytes] 1.0E-09*(CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C0:STATE=0x3F+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0
+L3 to/from system bandwidth [MBytes/s] 1.0E-06*(CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0/time
+L3 to/from system data volume [GBytes] 1.0E-09*(CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C0+CBOX0C1+CBOX1C1+CBOX2C1+CBOX3C1+CBOX4C1+CBOX5C1+CBOX6C1+CBOX7C1)*64.0
 Memory read bandwidth [MBytes/s] 1.0E-06*(MBOX0C0+MBOX1C0+MBOX2C0+MBOX3C0)*64.0/time
 Memory read data volume [GBytes] 1.0E-09*(MBOX0C0+MBOX1C0+MBOX2C0+MBOX3C0)*64.0
 Memory write bandwidth [MBytes/s] 1.0E-06*(MBOX0C1+MBOX1C1+MBOX2C1+MBOX3C1)*64.0/time

diff --git a/groups/sandybridgeEP/L3CACHE.txt b/groups/sandybridgeEP/L3CACHE.txt
@@ -13,9 +13,9 @@ Runtime (RDTSC) [s] time
 Runtime unhalted [s] FIXC1*inverseClock
 Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
-L3 request rate PMC1:MATCH0=0x0081:MATCH1=0x1/FIXC0
-L3 miss rate PMC0:MATCH0=0x0081:MATCH1=0x3fffc0/FIXC0
-L3 miss ratio PMC0:MATCH0=0x0081:MATCH1=0x3fffc0/PMC1:MATCH0=0x0081:MATCH1=0x1
+L3 request rate PMC1/FIXC0
+L3 miss rate PMC0/FIXC0
+L3 miss ratio PMC0/PMC1
 
 LONG
 Formulas:

diff --git a/groups/skylake/UOPS_EXEC.txt b/groups/skylake/UOPS_EXEC.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/skylake/UOPS_ISSUE.txt b/groups/skylake/UOPS_ISSUE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/groups/skylake/UOPS_RETIRE.txt b/groups/skylake/UOPS_RETIRE.txt
@@ -16,7 +16,7 @@ Clock [MHz]  1.E-06*(FIXC1/FIXC2)/inverseClock
 CPI  FIXC1/FIXC0
 Used cycles ratio [%] 100*PMC0/PMC2
 Unused cycles ratio [%] 100*PMC1/PMC2
-Avg stall duration [cycles] PMC1/PMC3:EDGEDETECT
+Avg stall duration [cycles] PMC1/PMC3
 
 
 LONG

diff --git a/src/includes/perfgroup.h b/src/includes/perfgroup.h
@@ -88,6 +88,8 @@ extern int add_to_clist(CounterList* clist, char* counter, double result);
 extern int update_clist(CounterList* clist, char* counter, double result);
 extern void destroy_clist(CounterList* clist);
 
-extern int calc_metric(char* formula, CounterList* clist, double *result);
+int add_var(char* name, char* value, char** varstr, char** varlist);
+int add_dbl_var(char* name, double value, char** varstr, char** varlist);
+extern int calc_metric(int cpu, char* formula, char* varstr, char* varlist, double *result);
 
 #endif /* PERFGROUP_H */
diff --git a/src/perfgroup.c b/src/perfgroup.c
@@ -1491,6 +1491,23 @@ int add_var(char* name, char* value, char** varstr, char** varlist)
     return 0;
 }
 
+int add_dbl_var(char* name, double value, char** varstr, char** varlist)
+{
+    int slen = 0;
+    int ret = 0;
+    if (*varstr)
+        slen = strlen(*varstr);
+    int add = strlen(name) + 45;
+    *varstr = realloc_buffer(*varstr, slen + add + 4);
+    if (!*varstr)
+    {
+        return -ENOMEM;
+    }
+    ret = snprintf(&((*varstr)[slen]), add+4, "%s = %20.20f\n", name, value);
+    ret = add_to_varlist(name, varlist);
+    return 0;
+}
+
 int add_def(char* name, char* value, int cpu, char** varlist)
 {
     int slen = 0;
@@ -1545,7 +1562,7 @@ double do_calc(int cpu, char* s, char* vars, char* varlist)
     {
         return res;
     }
-
+    printf("%s\n", t);
     ret = luaL_dostring (L, t);
     if (!ret)
     {
@@ -1586,39 +1603,23 @@ char* do_expand(int cpu, char* s, char *varlist)
 
 
 int
-calc_metric(char* formula, CounterList* clist, double *result)
+calc_metric(int cpu, char* formula, char* vars, char* varlist, double *result)
 {
     int i=0;
     char* f;
     int maxstrlen = 0, minstrlen = 10000;
-    int cpu = 0;
     char buf[128];
 
-    if ((formula == NULL) || (clist == NULL) || (result == NULL))
+    if ((formula == NULL) || (vars == NULL) || (varlist == NULL) || (result == NULL))
         return -EINVAL;
     *result = 0.0;
 
-    char* vars = NULL;
-    char* varlist = NULL;
-
-    for(i=0;i<clist->counters;i++)
-    {
-        snprintf(buf, 127, "%.25f", clist->cvalues[i]);
-        add_var(clist->cnames[i], buf, &vars, &varlist);
-        if (strcmp(clist->cnames[i], "CPU") == 0)
-            cpu = (int)clist->cvalues[i];
-        memset(buf, 0, 128 * sizeof(char));
-    }
     f = do_expand(cpu, formula, varlist);
     if (f)
     {
         *result = do_calc(cpu, f, vars, varlist);
     }
 
-    if (vars)
-        free(vars);
-    if (varlist)
-        free(varlist);
     return i;
 }