From 6dc0459613420cebe55812c23ce260fdc3f9fc4d Mon Sep 17 00:00:00 2001
From: Saurabh Verma <sauverma@amd.com>
Date: Wed, 11 Jan 2023 17:58:18 +0000
Subject: [PATCH] Adding missing MI200 metrics

Change-Id: I410f50e03d38bb03cf43e743318eb1242e7d6518


[ROCm/rocprofiler commit: 225bddf148c67725b88c63efa748390213912999]
---
 projects/rocprofiler/test/tool/metrics.xml | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/projects/rocprofiler/test/tool/metrics.xml b/projects/rocprofiler/test/tool/metrics.xml
index 85f3e6fb1b..c8e010b34c 100755
--- a/projects/rocprofiler/test/tool/metrics.xml
+++ b/projects/rocprofiler/test/tool/metrics.xml
@@ -213,6 +213,19 @@
   <metric name="FETCH_SIZE"                                             expr=(TCC_EA_RDREQ_32B_sum*32+(TCC_EA_RDREQ_sum-TCC_EA_RDREQ_32B_sum)*64)/1024 descr="The total kilobytes fetched from the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
   <metric name="WRITE_SIZE"                                             expr=((TCC_EA_WRREQ_sum-TCC_EA_WRREQ_64B_sum)*32+TCC_EA_WRREQ_64B_sum*64)/1024 descr="The total kilobytes written to the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
   <metric name="WRITE_REQ_32B"                                          expr=TCC_EA_WRREQ_64B_sum*2+(TCC_EA_WRREQ_sum-TCC_EA_WRREQ_64B_sum) descr="The total number of 32-byte effective memory writes."></metric>
+  <metric name="VFetchInsts"                                            expr=(SQ_INSTS_VMEM_RD-TA_FLAT_READ_WAVEFRONTS_sum)/SQ_WAVES descr="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>
+  <metric name="VWriteInsts"                                            expr=(SQ_INSTS_VMEM_WR-TA_FLAT_WRITE_WAVEFRONTS_sum)/SQ_WAVES descr="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>
+  <metric name="FlatVMemInsts"                                          expr=(SQ_INSTS_FLAT-SQ_INSTS_FLAT_LDS_ONLY)/SQ_WAVES descr="The average number of FLAT instructions that read from or write to the video memory executed per work item (affected by flow control). Includes FLAT instructions that read from or write to scratch."></metric>
+  <metric name="LDSInsts"                                               expr=(SQ_INSTS_LDS-SQ_INSTS_FLAT_LDS_ONLY)/SQ_WAVES descr="The average number of LDS read or LDS write instructions executed per work item (affected by flow control).  Excludes FLAT instructions that read from or write to LDS."></metric>
+  <metric name="FlatLDSInsts"                                           expr=SQ_INSTS_FLAT_LDS_ONLY/SQ_WAVES descr="The average number of FLAT instructions that read or write to LDS executed per work item (affected by flow control)."></metric>
+  <metric name="VALUUtilization"                                        expr=100*SQ_THREAD_CYCLES_VALU/(SQ_ACTIVE_INST_VALU*MAX_WAVE_SIZE) descr="The percentage of active vector ALU threads in a wave. A lower number can mean either more thread divergence in a wave or that the work-group size is not a multiple of 64. Value range: 0% (bad), 100% (ideal - no thread divergence)."></metric>
+  <metric name="VALUBusy"                                               expr=100*SQ_ACTIVE_INST_VALU*4/SIMD_NUM/GRBM_GUI_ACTIVE descr="The percentage of GPUTime vector ALU instructions are processed. Value range: 0% (bad) to 100% (optimal)."></metric>
+  <metric name="SALUBusy"                                               expr=100*SQ_INST_CYCLES_SALU*4/SIMD_NUM/GRBM_GUI_ACTIVE descr="The percentage of GPUTime scalar ALU instructions are processed. Value range: 0% (bad) to 100% (optimal)."></metric>
+  <metric name="MemWrites32B"                                           expr=WRITE_REQ_32B descr="The total number of effective 32B write transactions to the memory"></metric>
+  <metric name="L2CacheHit"                                             expr=100*sum(TCC_HIT,32)/(sum(TCC_HIT,32)+sum(TCC_MISS,32)) descr="The percentage of fetch, write, atomic, and other instructions that hit the data in L2 cache. Value range: 0% (no hit) to 100% (optimal)."></metric>
+  <metric name="MemUnitStalled"                                         expr=100*max(TCP_TCP_TA_DATA_STALL_CYCLES,16)/GRBM_GUI_ACTIVE/SE_NUM descr="The percentage of GPUTime the memory unit is stalled. Try reducing the number or size of fetches and writes if possible. Value range: 0% (optimal) to 100% (bad)."></metric>
+  <metric name="WriteUnitStalled"                                       expr=100*TCC_WRREQ_STALL_max/GRBM_GUI_ACTIVE descr="The percentage of GPUTime the Write unit is stalled. Value range: 0% to 100% (bad)."></metric>
+  <metric name="LDSBankConflict"                                        expr=100*SQ_LDS_BANK_CONFLICT/GRBM_GUI_ACTIVE/CU_NUM descr="The percentage of GPUTime LDS is stalled by bank conflicts. Value range: 0% (optimal) to 100% (bad)."></metric>
 </gfx90a_expr>
 
 <gfx10_expr>