adding pseudo realistic metrics

CMakeLists.txt

@@ -87,7 +87,7 @@ set ( CPACK_PACKAGE_VERSION_MINOR ${BUILD_VERSION_MINOR} )
 set ( CPACK_PACKAGE_VERSION_PATCH ${BUILD_VERSION_PATCH} )
 set ( CPACK_PACKAGE_CONTACT "Advanced Micro Devices Inc." )
 set ( CPACK_PACKAGE_DESCRIPTION_SUMMARY "ROCPROFILER library for AMD HSA runtime API extension support" )
-set ( CPACK_RESOURCE_FILE_LICENSE "${CMAKE_CURRENT_SOURCE_DIR}/LICENSE.md" )
+set ( CPACK_RESOURCE_FILE_LICENSE "${CMAKE_CURRENT_SOURCE_DIR}/LICENSE" )
 
 ## Debian package specific variables
 set ( CPACK_DEBIAN_PACKAGE_HOMEPAGE "https://github.com/RadeonOpenCompute/HSA-RocProfiler" )

test/input.xml

@@ -1,4 +1,4 @@
-<metric name=CPC_ME1_STALL_WAIT_ON_RCIU_READ,SQ_WAVES,SQ_WAVE_READY,SQ_CYCLES,SQ_ITEMS,WAVE_STALLS_RATE ></metric>
+<metric name=CPC_ME1_STALL_WAIT_ON_RCIU_READ,SQ_WAVES,SQ_WAVE_READY,SQ_CYCLES,SQ_ITEMS,SQ_INSTS_VMEM_WR,VWRITE_INSTS,SQ_INSTS_SMEM,SFETCH_INSTS ></metric>
 <trace name=SQTT copy=true >
   <parameters
     HSA_VEN_AMD_AQLPROFILE_PARAMETER_NAME_MASK=0xf

test/metrics.xml

@@ -2,14 +2,42 @@
   <metric name=SQ_CYCLES block=SQ event=2 ></metric>
   <metric name=SQ_WAVES block=SQ event=4 ></metric>
   <metric name=SQ_ITEMS block=SQ event=14 ></metric>
+  <metric name=SQ_INSTS_VMEM_WR block=SQ event=27 ></metric>
+  <metric name=SQ_INSTS_VMEM_RD block=SQ event=28 ></metric>
+  <metric name=SQ_INSTS_SMEM block=SQ event=31 ></metric>
   <metric name=SQ_WAVE_READY block=SQ event=47 ></metric>
+
+  <metric name=TA_FLAT_READ_WAVEFRONTS block=TA event=101 ></metric>
+  <metric name=TA_FLAT_WRITE_WAVEFRONTS block=TA event=102 ></metric>
+
   <metric name=TCC_CYCLE block=TCC event=1 ></metric>
   <metric name=TCC_REQ block=TCC event=3 ></metric>
   <metric name=TCC_WRITEBACK block=TCC event=22 ></metric>
+
   <metric name=CPC_ALWAYS_COUNT block=CPC event=0 ></metric>
   <metric name=CPC_ME1_STALL_WAIT_ON_RCIU_READ block=CPC event=8 ></metric>
+
+  <metric
+    name=TA_FLAT_READ_WAVEFRONTS_sum
+    expr=TA_FLAT_READ_WAVEFRONTS[0]+TA_FLAT_READ_WAVEFRONTS[1]
+  ></metric>
+  <metric
+    name=TA_FLAT_WRITE_WAVEFRONTS_sum
+    expr=TA_FLAT_WRITE_WAVEFRONTS[0]+TA_FLAT_WRITE_WAVEFRONTS[1]
+  ></metric>
 </gfx8>
 
 <global>
-  <metric name=WAVE_STALLS_RATE expr=CPC_ME1_STALL_WAIT_ON_RCIU_READ*(SQ_WAVES+SQ_WAVE_READY)*100/(SQ_CYCLES/SQ_ITEMS) ></metric>
+  # VFETCH_INSTS
+  # The average number of vector fetch instructions from the video memory executed per work-item (affected by flow control).
+  # Excludes FLAT instructions that fetch from video memory.
+  <metric name=VFETCH_INSTS expr=1000*(SQ_INSTS_VMEM_RD-TA_FLAT_READ_WAVEFRONTS_sum)/SQ_WAVES ></metric>
+
+  # VWRITE_INSTS
+  # The average number of vector write instructions to the video memory executed per work-item (affected by flow control). Excludes FLAT instructions that write to video memory.
+  <metric name=VWRITE_INSTS expr=1000*(SQ_INSTS_VMEM_WR-TA_FLAT_WRITE_WAVEFRONTS_sum)/SQ_WAVES ></metric>
+
+  # SFETCH_INSTS
+  # The average number of scalar fetch instructions from the video memory executed per work-item (affected by flow control).
+  <metric name=SFETCH_INSTS expr=1000*SQ_INSTS_SMEM/SQ_WAVES ></metric>
 </global>