AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Pytest add mi200 to analyze workloads (#334)

Browse Source 
* Updated links in documentation. (#328)

Updated to reflect new GitHub organization.
Fixed broken links to GitHub pages.

Signed-off-by: David Galiffi <David.Galiffi@amd.com>

* update branch for 2.x documentation builds

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* update checkout action and use concurrency instead of cancel-workflow-action

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* test addition of user option for container launch

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* remove --user option for container, try chown instead

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* fixing yaml syntax

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* reorder job step - start with checkout

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* restore missing run directive

Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>

* Update workloads to include log.txt
Add missing MI200 workloads

Signed-off-by: Jose Santos <josantos@amd.com>

* Signed-off-by: Jose Santos <josantos@amd.com>
Add vcopy workload for tests

* Change exit codes for caught failures

Signed-off-by: Jose Santos <josantos@amd.com>

* reformat

Signed-off-by: Jose Santos <josantos@amd.com>

* Add pytest-xdist for pytest -n

Signed-off-by: Jose Santos <josantos@amd.com>

---------

Signed-off-by: David Galiffi <David.Galiffi@amd.com>
Signed-off-by: Karl W. Schulz <karl.schulz@amd.com>
Signed-off-by: Jose Santos <josantos@amd.com>
Co-authored-by: David Galiffi <David.Galiffi@amd.com>
Co-authored-by: Karl W. Schulz <karl.schulz@amd.com>
This commit is contained in:
JoseSantosAMD
2024-03-25 10:20:31 -05:00
committed by Cole Ramos
parent 482fd6f2ca
commit da506ad9b5
1122 changed files with 37938 additions and 3563 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/workloads/dispatch_inv/MI100/SQ_IFETCH_LEVEL.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,GRBM_COUNT,GRBM_GUI_ACTIVE,SQ_WAVES,SQ_IFETCH,SQ_IFETCH_LEVEL,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,100378,100378,1048576,256,0,0,8,8,16,64,0x0,0x7f626043ce80,47668,47668,16384,65536,29010,1854504,194051411496730,194063713519257,194063713543257,194051419980944
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1196025,1196025,1048576,256,0,0,8,8,16,64,0x0,0x7f3d8c718ec0,48770,48770,16384,65536,14340,1831132,1409981308951962,1409993151479981,1409993151504461,1409981316706246
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj GRBM_COUNT GRBM_GUI_ACTIVE SQ_WAVES SQ_IFETCH SQ_IFETCH_LEVEL SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 100378 1196025 100378 1196025 1048576 256 0 0 8 8 16 64 0x0 0x7f626043ce80 0x7f3d8c718ec0 47668 48770 47668 48770 16384 65536 29010 14340 1854504 1831132 194051411496730 1409981308951962 194063713519257 1409993151479981 194063713543257 1409993151504461 194051419980944 1409981316706246
tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_LDS.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_LDS,SQ_INST_LEVEL_LDS,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,100568,100568,1048576,256,0,0,8,8,16,64,0x0,0x7fd4ec17ce80,0,0,0,194051899565198,194063713519257,194063713543257,194051907560356
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1196228,1196228,1048576,256,0,0,8,8,16,64,0x0,0x7fde5b1e8ec0,0,0,0,1409981797196754,1409993151479981,1409993151504461,1409981804835661
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_LDS SQ_INST_LEVEL_LDS SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 100568 1196228 100568 1196228 1048576 256 0 0 8 8 16 64 0x0 0x7fd4ec17ce80 0x7fde5b1e8ec0 0 0 0 194051899565198 1409981797196754 194063713519257 1409993151479981 194063713543257 1409993151504461 194051907560356 1409981804835661
tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_SMEM.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_SMEM,SQ_INST_LEVEL_SMEM,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,100757,100757,1048576,256,0,0,8,8,16,64,0x0,0x7f9039444e80,65536,381816,24432184,194052393039211,194063713519257,194063713543257,194052400779427
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1196431,1196431,1048576,256,0,0,8,8,16,64,0x0,0x7f4aa38c8ec0,65536,241224,30840808,1409982278411950,1409993151479981,1409993151504461,1409982286062829
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_SMEM SQ_INST_LEVEL_SMEM SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 100757 1196431 100757 1196431 1048576 256 0 0 8 8 16 64 0x0 0x7f9039444e80 0x7f4aa38c8ec0 65536 381816 241224 24432184 30840808 194052393039211 1409982278411950 194063713519257 1409993151479981 194063713543257 1409993151504461 194052400779427 1409982286062829
tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_VMEM.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_VMEM,SQ_INST_LEVEL_VMEM,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,100958,100958,1048576,256,0,0,8,8,16,64,0x0,0x7f1524968e80,32768,1335366,85465152,194052885720815,194063713519257,194063713543257,194052895100074
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1196615,1196615,1048576,256,0,0,8,8,16,64,0x0,0x7f9060c34ec0,32768,613680,78559712,1409982759228172,1409993151479981,1409993151504461,1409982766976496
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_VMEM SQ_INST_LEVEL_VMEM SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 100958 1196615 100958 1196615 1048576 256 0 0 8 8 16 64 0x0 0x7f1524968e80 0x7f9060c34ec0 32768 1335366 613680 85465152 78559712 194052885720815 1409982759228172 194063713519257 1409993151479981 194063713543257 1409993151504461 194052895100074 1409982766976496
tests/workloads/dispatch_inv/MI100/SQ_LEVEL_WAVES.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,GRBM_COUNT,GRBM_GUI_ACTIVE,CPC_ME1_BUSY_FOR_PACKET_DECODE,SQ_CYCLES,SQ_WAVES,SQ_WAVE_CYCLES,SQ_BUSY_CYCLES,SQ_LEVEL_WAVES,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,101159,101159,1048576,256,0,0,8,8,16,64,0x0,0x7f37e15b8e80,48091,48091,17290,384736,16384,25334875,235204,0,101833412,194053385787391,194063713519257,194063713543257,194053393562893
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1196818,1196818,1048576,256,0,0,8,8,16,64,0x0,0x7f782cf40ec0,49166,49166,17906,393336,16384,25994718,243072,0,104504564,1409983239161961,1409993151479981,1409993151504461,1409983246839460
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj GRBM_COUNT GRBM_GUI_ACTIVE CPC_ME1_BUSY_FOR_PACKET_DECODE SQ_CYCLES SQ_WAVES SQ_WAVE_CYCLES SQ_BUSY_CYCLES SQ_LEVEL_WAVES SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 101159 1196818 101159 1196818 1048576 256 0 0 8 8 16 64 0x0 0x7f37e15b8e80 0x7f782cf40ec0 48091 49166 48091 49166 17290 17906 384736 393336 16384 25334875 25994718 235204 243072 0 101833412 104504564 194053385787391 1409983239161961 194063713519257 1409993151479981 194063713543257 1409993151504461 194053393562893 1409983246839460
tests/workloads/dispatch_inv/MI100/log.txt
+702
View File
@@ -0,0 +1,702 @@
Omniperf version: 2.0.0-RC1
Profiler choice: rocprofv1
Path: /home1/josantos/omniperf/tests/workloads/dispatch_inv/MI100
Target: MI100
Command: ./tests/vcopy -n 1048576 -b 256 -i 3
Kernel Selection: None
Dispatch Selection: ['invalid']
IP Blocks: All
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Collecting Performance Counters
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/SQ_IFETCH_LEVEL.txt
|-> [rocprof] RPL: on '240321_155047' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/SQ_IFETCH_LEVEL.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155047_1195865'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155047_1195865/input0_results_240321_155047'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155047_1195865/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 6 metrics
|-> [rocprof] GRBM_COUNT, GRBM_GUI_ACTIVE, SQ_WAVES, SQ_IFETCH, SQ_IFETCH_LEVEL, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155047_1195865/input0_results_240321_155047
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/SQ_IFETCH_LEVEL.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_LDS.txt
|-> [rocprof] RPL: on '240321_155047' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_LDS.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155047_1196068'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155047_1196068/input0_results_240321_155047'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155047_1196068/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_LDS, SQ_INST_LEVEL_LDS, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155047_1196068/input0_results_240321_155047
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_LDS.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_SMEM.txt
|-> [rocprof] RPL: on '240321_155048' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_SMEM.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155048_1196271'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155048_1196271/input0_results_240321_155048'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155048_1196271/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_SMEM, SQ_INST_LEVEL_SMEM, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155048_1196271/input0_results_240321_155048
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_SMEM.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_VMEM.txt
|-> [rocprof] RPL: on '240321_155048' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/SQ_INST_LEVEL_VMEM.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155048_1196455'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155048_1196455/input0_results_240321_155048'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155048_1196455/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_VMEM, SQ_INST_LEVEL_VMEM, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155048_1196455/input0_results_240321_155048
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/SQ_INST_LEVEL_VMEM.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/SQ_LEVEL_WAVES.txt
|-> [rocprof] RPL: on '240321_155049' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/SQ_LEVEL_WAVES.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155049_1196658'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155049_1196658/input0_results_240321_155049'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155049_1196658/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 9 metrics
|-> [rocprof] GRBM_COUNT, GRBM_GUI_ACTIVE, CPC_ME1_BUSY_FOR_PACKET_DECODE, SQ_CYCLES, SQ_WAVES, SQ_WAVE_CYCLES, SQ_BUSY_CYCLES, SQ_LEVEL_WAVES, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155049_1196658/input0_results_240321_155049
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/SQ_LEVEL_WAVES.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_0.txt
|-> [rocprof] RPL: on '240321_155049' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_0.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155049_1196843'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155049_1196843/input0_results_240321_155049'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155049_1196843/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 28 metrics
|-> [rocprof] SQ_CYCLES, SQ_BUSY_CYCLES, SQ_BUSY_CU_CYCLES, SQ_WAVES, SQ_WAVE_CYCLES, SQC_TC_INST_REQ, SQC_TC_DATA_READ_REQ, SQC_TC_DATA_WRITE_REQ, GRBM_COUNT, GRBM_GUI_ACTIVE, TCP_GATE_EN1_sum, TCP_GATE_EN2_sum, TCP_TD_TCP_STALL_CYCLES_sum, TCP_TCR_TCP_STALL_CYCLES_sum, TA_TA_BUSY_sum, TA_BUFFER_WAVEFRONTS_sum, TD_TD_BUSY_sum, TD_TC_STALL_sum, SPI_CSN_WINDOW_VALID, SPI_CSN_BUSY, CPC_CPC_STAT_BUSY, CPC_CPC_STAT_IDLE, CPF_CPF_STAT_BUSY, CPF_CPF_STAT_STALL, TCC_CYCLE_sum, TCC_BUSY_sum, TCC_PROBE_sum, TCC_PROBE_ALL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155049_1196843/input0_results_240321_155049
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_0.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_1.txt
|-> [rocprof] RPL: on '240321_155050' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_1.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155050_1197044'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155050_1197044/input0_results_240321_155050'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155050_1197044/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 27 metrics
|-> [rocprof] SQC_TC_DATA_ATOMIC_REQ, SQC_TC_STALL, SQC_TC_REQ, SQC_DCACHE_REQ_READ_16, SQC_ICACHE_REQ, SQC_ICACHE_HITS, SQC_ICACHE_MISSES, SQC_ICACHE_MISSES_DUPLICATE, GRBM_SPI_BUSY, TCP_READ_TAGCONFLICT_STALL_CYCLES_sum, TCP_WRITE_TAGCONFLICT_STALL_CYCLES_sum, TCP_ATOMIC_TAGCONFLICT_STALL_CYCLES_sum, TCP_TA_TCP_STATE_READ_sum, TA_BUFFER_READ_WAVEFRONTS_sum, TA_BUFFER_WRITE_WAVEFRONTS_sum, TD_COALESCABLE_WAVEFRONT_sum, TD_LOAD_WAVEFRONT_sum, SPI_CSN_NUM_THREADGROUPS, SPI_CSN_WAVE, CPC_CPC_TCIU_BUSY, CPC_CPC_TCIU_IDLE, CPF_CPF_TCIU_BUSY, CPF_CPF_TCIU_STALL, TCC_NC_REQ_sum, TCC_UC_REQ_sum, TCC_CC_REQ_sum, TCC_RW_REQ_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155050_1197044/input0_results_240321_155050
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_1.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_10.txt
|-> [rocprof] RPL: on '240321_155050' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_10.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155050_1197247'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155050_1197247/input0_results_240321_155050'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155050_1197247/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 1 metrics
|-> [rocprof] TCC_EA_ATOMIC_LEVEL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155050_1197247/input0_results_240321_155050
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_10.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_11.txt
|-> [rocprof] RPL: on '240321_155051' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_11.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155051_1197449'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155051_1197449/input0_results_240321_155051'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155051_1197449/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_ATOMIC[0], TCC_CYCLE[0], TCC_EA_ATOMIC[0], TCC_EA_ATOMIC_LEVEL[0], TCC_ATOMIC[1], TCC_CYCLE[1], TCC_EA_ATOMIC[1], TCC_EA_ATOMIC_LEVEL[1], TCC_ATOMIC[2], TCC_CYCLE[2], TCC_EA_ATOMIC[2], TCC_EA_ATOMIC_LEVEL[2], TCC_ATOMIC[3], TCC_CYCLE[3], TCC_EA_ATOMIC[3], TCC_EA_ATOMIC_LEVEL[3], TCC_ATOMIC[4], TCC_CYCLE[4], TCC_EA_ATOMIC[4], TCC_EA_ATOMIC_LEVEL[4], TCC_ATOMIC[5], TCC_CYCLE[5], TCC_EA_ATOMIC[5], TCC_EA_ATOMIC_LEVEL[5], TCC_ATOMIC[6], TCC_CYCLE[6], TCC_EA_ATOMIC[6], TCC_EA_ATOMIC_LEVEL[6], TCC_ATOMIC[7], TCC_CYCLE[7], TCC_EA_ATOMIC[7], TCC_EA_ATOMIC_LEVEL[7], TCC_ATOMIC[8], TCC_CYCLE[8], TCC_EA_ATOMIC[8], TCC_EA_ATOMIC_LEVEL[8], TCC_ATOMIC[9], TCC_CYCLE[9], TCC_EA_ATOMIC[9], TCC_EA_ATOMIC_LEVEL[9], TCC_ATOMIC[10], TCC_CYCLE[10], TCC_EA_ATOMIC[10], TCC_EA_ATOMIC_LEVEL[10], TCC_ATOMIC[11], TCC_CYCLE[11], TCC_EA_ATOMIC[11], TCC_EA_ATOMIC_LEVEL[11], TCC_ATOMIC[12], TCC_CYCLE[12], TCC_EA_ATOMIC[12], TCC_EA_ATOMIC_LEVEL[12], TCC_ATOMIC[13], TCC_CYCLE[13], TCC_EA_ATOMIC[13], TCC_EA_ATOMIC_LEVEL[13], TCC_ATOMIC[14], TCC_CYCLE[14], TCC_EA_ATOMIC[14], TCC_EA_ATOMIC_LEVEL[14], TCC_ATOMIC[15], TCC_CYCLE[15], TCC_EA_ATOMIC[15], TCC_EA_ATOMIC_LEVEL[15], TCC_ATOMIC[16], TCC_CYCLE[16], TCC_EA_ATOMIC[16], TCC_EA_ATOMIC_LEVEL[16], TCC_ATOMIC[17], TCC_CYCLE[17], TCC_EA_ATOMIC[17], TCC_EA_ATOMIC_LEVEL[17], TCC_ATOMIC[18], TCC_CYCLE[18], TCC_EA_ATOMIC[18], TCC_EA_ATOMIC_LEVEL[18], TCC_ATOMIC[19], TCC_CYCLE[19], TCC_EA_ATOMIC[19], TCC_EA_ATOMIC_LEVEL[19], TCC_ATOMIC[20], TCC_CYCLE[20], TCC_EA_ATOMIC[20], TCC_EA_ATOMIC_LEVEL[20], TCC_ATOMIC[21], TCC_CYCLE[21], TCC_EA_ATOMIC[21], TCC_EA_ATOMIC_LEVEL[21], TCC_ATOMIC[22], TCC_CYCLE[22], TCC_EA_ATOMIC[22], TCC_EA_ATOMIC_LEVEL[22], TCC_ATOMIC[23], TCC_CYCLE[23], TCC_EA_ATOMIC[23], TCC_EA_ATOMIC_LEVEL[23], TCC_ATOMIC[24], TCC_CYCLE[24], TCC_EA_ATOMIC[24], TCC_EA_ATOMIC_LEVEL[24], TCC_ATOMIC[25], TCC_CYCLE[25], TCC_EA_ATOMIC[25], TCC_EA_ATOMIC_LEVEL[25], TCC_ATOMIC[26], TCC_CYCLE[26], TCC_EA_ATOMIC[26], TCC_EA_ATOMIC_LEVEL[26], TCC_ATOMIC[27], TCC_CYCLE[27], TCC_EA_ATOMIC[27], TCC_EA_ATOMIC_LEVEL[27], TCC_ATOMIC[28], TCC_CYCLE[28], TCC_EA_ATOMIC[28], TCC_EA_ATOMIC_LEVEL[28], TCC_ATOMIC[29], TCC_CYCLE[29], TCC_EA_ATOMIC[29], TCC_EA_ATOMIC_LEVEL[29], TCC_ATOMIC[30], TCC_CYCLE[30], TCC_EA_ATOMIC[30], TCC_EA_ATOMIC_LEVEL[30], TCC_ATOMIC[31], TCC_CYCLE[31], TCC_EA_ATOMIC[31], TCC_EA_ATOMIC_LEVEL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155051_1197449/input0_results_240321_155051
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_11.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_12.txt
|-> [rocprof] RPL: on '240321_155051' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_12.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155051_1197651'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155051_1197651/input0_results_240321_155051'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155051_1197651/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_RDREQ[0], TCC_EA_RDREQ_32B[0], TCC_EA_RDREQ_DRAM_CREDIT_STALL[0], TCC_EA_RDREQ_GMI_CREDIT_STALL[0], TCC_EA_RDREQ[1], TCC_EA_RDREQ_32B[1], TCC_EA_RDREQ_DRAM_CREDIT_STALL[1], TCC_EA_RDREQ_GMI_CREDIT_STALL[1], TCC_EA_RDREQ[2], TCC_EA_RDREQ_32B[2], TCC_EA_RDREQ_DRAM_CREDIT_STALL[2], TCC_EA_RDREQ_GMI_CREDIT_STALL[2], TCC_EA_RDREQ[3], TCC_EA_RDREQ_32B[3], TCC_EA_RDREQ_DRAM_CREDIT_STALL[3], TCC_EA_RDREQ_GMI_CREDIT_STALL[3], TCC_EA_RDREQ[4], TCC_EA_RDREQ_32B[4], TCC_EA_RDREQ_DRAM_CREDIT_STALL[4], TCC_EA_RDREQ_GMI_CREDIT_STALL[4], TCC_EA_RDREQ[5], TCC_EA_RDREQ_32B[5], TCC_EA_RDREQ_DRAM_CREDIT_STALL[5], TCC_EA_RDREQ_GMI_CREDIT_STALL[5], TCC_EA_RDREQ[6], TCC_EA_RDREQ_32B[6], TCC_EA_RDREQ_DRAM_CREDIT_STALL[6], TCC_EA_RDREQ_GMI_CREDIT_STALL[6], TCC_EA_RDREQ[7], TCC_EA_RDREQ_32B[7], TCC_EA_RDREQ_DRAM_CREDIT_STALL[7], TCC_EA_RDREQ_GMI_CREDIT_STALL[7], TCC_EA_RDREQ[8], TCC_EA_RDREQ_32B[8], TCC_EA_RDREQ_DRAM_CREDIT_STALL[8], TCC_EA_RDREQ_GMI_CREDIT_STALL[8], TCC_EA_RDREQ[9], TCC_EA_RDREQ_32B[9], TCC_EA_RDREQ_DRAM_CREDIT_STALL[9], TCC_EA_RDREQ_GMI_CREDIT_STALL[9], TCC_EA_RDREQ[10], TCC_EA_RDREQ_32B[10], TCC_EA_RDREQ_DRAM_CREDIT_STALL[10], TCC_EA_RDREQ_GMI_CREDIT_STALL[10], TCC_EA_RDREQ[11], TCC_EA_RDREQ_32B[11], TCC_EA_RDREQ_DRAM_CREDIT_STALL[11], TCC_EA_RDREQ_GMI_CREDIT_STALL[11], TCC_EA_RDREQ[12], TCC_EA_RDREQ_32B[12], TCC_EA_RDREQ_DRAM_CREDIT_STALL[12], TCC_EA_RDREQ_GMI_CREDIT_STALL[12], TCC_EA_RDREQ[13], TCC_EA_RDREQ_32B[13], TCC_EA_RDREQ_DRAM_CREDIT_STALL[13], TCC_EA_RDREQ_GMI_CREDIT_STALL[13], TCC_EA_RDREQ[14], TCC_EA_RDREQ_32B[14], TCC_EA_RDREQ_DRAM_CREDIT_STALL[14], TCC_EA_RDREQ_GMI_CREDIT_STALL[14], TCC_EA_RDREQ[15], TCC_EA_RDREQ_32B[15], TCC_EA_RDREQ_DRAM_CREDIT_STALL[15], TCC_EA_RDREQ_GMI_CREDIT_STALL[15], TCC_EA_RDREQ[16], TCC_EA_RDREQ_32B[16], TCC_EA_RDREQ_DRAM_CREDIT_STALL[16], TCC_EA_RDREQ_GMI_CREDIT_STALL[16], TCC_EA_RDREQ[17], TCC_EA_RDREQ_32B[17], TCC_EA_RDREQ_DRAM_CREDIT_STALL[17], TCC_EA_RDREQ_GMI_CREDIT_STALL[17], TCC_EA_RDREQ[18], TCC_EA_RDREQ_32B[18], TCC_EA_RDREQ_DRAM_CREDIT_STALL[18], TCC_EA_RDREQ_GMI_CREDIT_STALL[18], TCC_EA_RDREQ[19], TCC_EA_RDREQ_32B[19], TCC_EA_RDREQ_DRAM_CREDIT_STALL[19], TCC_EA_RDREQ_GMI_CREDIT_STALL[19], TCC_EA_RDREQ[20], TCC_EA_RDREQ_32B[20], TCC_EA_RDREQ_DRAM_CREDIT_STALL[20], TCC_EA_RDREQ_GMI_CREDIT_STALL[20], TCC_EA_RDREQ[21], TCC_EA_RDREQ_32B[21], TCC_EA_RDREQ_DRAM_CREDIT_STALL[21], TCC_EA_RDREQ_GMI_CREDIT_STALL[21], TCC_EA_RDREQ[22], TCC_EA_RDREQ_32B[22], TCC_EA_RDREQ_DRAM_CREDIT_STALL[22], TCC_EA_RDREQ_GMI_CREDIT_STALL[22], TCC_EA_RDREQ[23], TCC_EA_RDREQ_32B[23], TCC_EA_RDREQ_DRAM_CREDIT_STALL[23], TCC_EA_RDREQ_GMI_CREDIT_STALL[23], TCC_EA_RDREQ[24], TCC_EA_RDREQ_32B[24], TCC_EA_RDREQ_DRAM_CREDIT_STALL[24], TCC_EA_RDREQ_GMI_CREDIT_STALL[24], TCC_EA_RDREQ[25], TCC_EA_RDREQ_32B[25], TCC_EA_RDREQ_DRAM_CREDIT_STALL[25], TCC_EA_RDREQ_GMI_CREDIT_STALL[25], TCC_EA_RDREQ[26], TCC_EA_RDREQ_32B[26], TCC_EA_RDREQ_DRAM_CREDIT_STALL[26], TCC_EA_RDREQ_GMI_CREDIT_STALL[26], TCC_EA_RDREQ[27], TCC_EA_RDREQ_32B[27], TCC_EA_RDREQ_DRAM_CREDIT_STALL[27], TCC_EA_RDREQ_GMI_CREDIT_STALL[27], TCC_EA_RDREQ[28], TCC_EA_RDREQ_32B[28], TCC_EA_RDREQ_DRAM_CREDIT_STALL[28], TCC_EA_RDREQ_GMI_CREDIT_STALL[28], TCC_EA_RDREQ[29], TCC_EA_RDREQ_32B[29], TCC_EA_RDREQ_DRAM_CREDIT_STALL[29], TCC_EA_RDREQ_GMI_CREDIT_STALL[29], TCC_EA_RDREQ[30], TCC_EA_RDREQ_32B[30], TCC_EA_RDREQ_DRAM_CREDIT_STALL[30], TCC_EA_RDREQ_GMI_CREDIT_STALL[30], TCC_EA_RDREQ[31], TCC_EA_RDREQ_32B[31], TCC_EA_RDREQ_DRAM_CREDIT_STALL[31], TCC_EA_RDREQ_GMI_CREDIT_STALL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155051_1197651/input0_results_240321_155051
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_12.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_13.txt
|-> [rocprof] RPL: on '240321_155052' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_13.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155052_1197852'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155052_1197852/input0_results_240321_155052'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155052_1197852/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_RDREQ_IO_CREDIT_STALL[0], TCC_EA_RDREQ_LEVEL[0], TCC_EA_WRREQ[0], TCC_EA_WRREQ_64B[0], TCC_EA_RDREQ_IO_CREDIT_STALL[1], TCC_EA_RDREQ_LEVEL[1], TCC_EA_WRREQ[1], TCC_EA_WRREQ_64B[1], TCC_EA_RDREQ_IO_CREDIT_STALL[2], TCC_EA_RDREQ_LEVEL[2], TCC_EA_WRREQ[2], TCC_EA_WRREQ_64B[2], TCC_EA_RDREQ_IO_CREDIT_STALL[3], TCC_EA_RDREQ_LEVEL[3], TCC_EA_WRREQ[3], TCC_EA_WRREQ_64B[3], TCC_EA_RDREQ_IO_CREDIT_STALL[4], TCC_EA_RDREQ_LEVEL[4], TCC_EA_WRREQ[4], TCC_EA_WRREQ_64B[4], TCC_EA_RDREQ_IO_CREDIT_STALL[5], TCC_EA_RDREQ_LEVEL[5], TCC_EA_WRREQ[5], TCC_EA_WRREQ_64B[5], TCC_EA_RDREQ_IO_CREDIT_STALL[6], TCC_EA_RDREQ_LEVEL[6], TCC_EA_WRREQ[6], TCC_EA_WRREQ_64B[6], TCC_EA_RDREQ_IO_CREDIT_STALL[7], TCC_EA_RDREQ_LEVEL[7], TCC_EA_WRREQ[7], TCC_EA_WRREQ_64B[7], TCC_EA_RDREQ_IO_CREDIT_STALL[8], TCC_EA_RDREQ_LEVEL[8], TCC_EA_WRREQ[8], TCC_EA_WRREQ_64B[8], TCC_EA_RDREQ_IO_CREDIT_STALL[9], TCC_EA_RDREQ_LEVEL[9], TCC_EA_WRREQ[9], TCC_EA_WRREQ_64B[9], TCC_EA_RDREQ_IO_CREDIT_STALL[10], TCC_EA_RDREQ_LEVEL[10], TCC_EA_WRREQ[10], TCC_EA_WRREQ_64B[10], TCC_EA_RDREQ_IO_CREDIT_STALL[11], TCC_EA_RDREQ_LEVEL[11], TCC_EA_WRREQ[11], TCC_EA_WRREQ_64B[11], TCC_EA_RDREQ_IO_CREDIT_STALL[12], TCC_EA_RDREQ_LEVEL[12], TCC_EA_WRREQ[12], TCC_EA_WRREQ_64B[12], TCC_EA_RDREQ_IO_CREDIT_STALL[13], TCC_EA_RDREQ_LEVEL[13], TCC_EA_WRREQ[13], TCC_EA_WRREQ_64B[13], TCC_EA_RDREQ_IO_CREDIT_STALL[14], TCC_EA_RDREQ_LEVEL[14], TCC_EA_WRREQ[14], TCC_EA_WRREQ_64B[14], TCC_EA_RDREQ_IO_CREDIT_STALL[15], TCC_EA_RDREQ_LEVEL[15], TCC_EA_WRREQ[15], TCC_EA_WRREQ_64B[15], TCC_EA_RDREQ_IO_CREDIT_STALL[16], TCC_EA_RDREQ_LEVEL[16], TCC_EA_WRREQ[16], TCC_EA_WRREQ_64B[16], TCC_EA_RDREQ_IO_CREDIT_STALL[17], TCC_EA_RDREQ_LEVEL[17], TCC_EA_WRREQ[17], TCC_EA_WRREQ_64B[17], TCC_EA_RDREQ_IO_CREDIT_STALL[18], TCC_EA_RDREQ_LEVEL[18], TCC_EA_WRREQ[18], TCC_EA_WRREQ_64B[18], TCC_EA_RDREQ_IO_CREDIT_STALL[19], TCC_EA_RDREQ_LEVEL[19], TCC_EA_WRREQ[19], TCC_EA_WRREQ_64B[19], TCC_EA_RDREQ_IO_CREDIT_STALL[20], TCC_EA_RDREQ_LEVEL[20], TCC_EA_WRREQ[20], TCC_EA_WRREQ_64B[20], TCC_EA_RDREQ_IO_CREDIT_STALL[21], TCC_EA_RDREQ_LEVEL[21], TCC_EA_WRREQ[21], TCC_EA_WRREQ_64B[21], TCC_EA_RDREQ_IO_CREDIT_STALL[22], TCC_EA_RDREQ_LEVEL[22], TCC_EA_WRREQ[22], TCC_EA_WRREQ_64B[22], TCC_EA_RDREQ_IO_CREDIT_STALL[23], TCC_EA_RDREQ_LEVEL[23], TCC_EA_WRREQ[23], TCC_EA_WRREQ_64B[23], TCC_EA_RDREQ_IO_CREDIT_STALL[24], TCC_EA_RDREQ_LEVEL[24], TCC_EA_WRREQ[24], TCC_EA_WRREQ_64B[24], TCC_EA_RDREQ_IO_CREDIT_STALL[25], TCC_EA_RDREQ_LEVEL[25], TCC_EA_WRREQ[25], TCC_EA_WRREQ_64B[25], TCC_EA_RDREQ_IO_CREDIT_STALL[26], TCC_EA_RDREQ_LEVEL[26], TCC_EA_WRREQ[26], TCC_EA_WRREQ_64B[26], TCC_EA_RDREQ_IO_CREDIT_STALL[27], TCC_EA_RDREQ_LEVEL[27], TCC_EA_WRREQ[27], TCC_EA_WRREQ_64B[27], TCC_EA_RDREQ_IO_CREDIT_STALL[28], TCC_EA_RDREQ_LEVEL[28], TCC_EA_WRREQ[28], TCC_EA_WRREQ_64B[28], TCC_EA_RDREQ_IO_CREDIT_STALL[29], TCC_EA_RDREQ_LEVEL[29], TCC_EA_WRREQ[29], TCC_EA_WRREQ_64B[29], TCC_EA_RDREQ_IO_CREDIT_STALL[30], TCC_EA_RDREQ_LEVEL[30], TCC_EA_WRREQ[30], TCC_EA_WRREQ_64B[30], TCC_EA_RDREQ_IO_CREDIT_STALL[31], TCC_EA_RDREQ_LEVEL[31], TCC_EA_WRREQ[31], TCC_EA_WRREQ_64B[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155052_1197852/input0_results_240321_155052
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_13.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_14.txt
|-> [rocprof] RPL: on '240321_155053' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_14.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155053_1198055'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155053_1198055/input0_results_240321_155053'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155053_1198055/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_WRREQ_DRAM_CREDIT_STALL[0], TCC_EA_WRREQ_GMI_CREDIT_STALL[0], TCC_EA_WRREQ_IO_CREDIT_STALL[0], TCC_EA_WRREQ_LEVEL[0], TCC_EA_WRREQ_DRAM_CREDIT_STALL[1], TCC_EA_WRREQ_GMI_CREDIT_STALL[1], TCC_EA_WRREQ_IO_CREDIT_STALL[1], TCC_EA_WRREQ_LEVEL[1], TCC_EA_WRREQ_DRAM_CREDIT_STALL[2], TCC_EA_WRREQ_GMI_CREDIT_STALL[2], TCC_EA_WRREQ_IO_CREDIT_STALL[2], TCC_EA_WRREQ_LEVEL[2], TCC_EA_WRREQ_DRAM_CREDIT_STALL[3], TCC_EA_WRREQ_GMI_CREDIT_STALL[3], TCC_EA_WRREQ_IO_CREDIT_STALL[3], TCC_EA_WRREQ_LEVEL[3], TCC_EA_WRREQ_DRAM_CREDIT_STALL[4], TCC_EA_WRREQ_GMI_CREDIT_STALL[4], TCC_EA_WRREQ_IO_CREDIT_STALL[4], TCC_EA_WRREQ_LEVEL[4], TCC_EA_WRREQ_DRAM_CREDIT_STALL[5], TCC_EA_WRREQ_GMI_CREDIT_STALL[5], TCC_EA_WRREQ_IO_CREDIT_STALL[5], TCC_EA_WRREQ_LEVEL[5], TCC_EA_WRREQ_DRAM_CREDIT_STALL[6], TCC_EA_WRREQ_GMI_CREDIT_STALL[6], TCC_EA_WRREQ_IO_CREDIT_STALL[6], TCC_EA_WRREQ_LEVEL[6], TCC_EA_WRREQ_DRAM_CREDIT_STALL[7], TCC_EA_WRREQ_GMI_CREDIT_STALL[7], TCC_EA_WRREQ_IO_CREDIT_STALL[7], TCC_EA_WRREQ_LEVEL[7], TCC_EA_WRREQ_DRAM_CREDIT_STALL[8], TCC_EA_WRREQ_GMI_CREDIT_STALL[8], TCC_EA_WRREQ_IO_CREDIT_STALL[8], TCC_EA_WRREQ_LEVEL[8], TCC_EA_WRREQ_DRAM_CREDIT_STALL[9], TCC_EA_WRREQ_GMI_CREDIT_STALL[9], TCC_EA_WRREQ_IO_CREDIT_STALL[9], TCC_EA_WRREQ_LEVEL[9], TCC_EA_WRREQ_DRAM_CREDIT_STALL[10], TCC_EA_WRREQ_GMI_CREDIT_STALL[10], TCC_EA_WRREQ_IO_CREDIT_STALL[10], TCC_EA_WRREQ_LEVEL[10], TCC_EA_WRREQ_DRAM_CREDIT_STALL[11], TCC_EA_WRREQ_GMI_CREDIT_STALL[11], TCC_EA_WRREQ_IO_CREDIT_STALL[11], TCC_EA_WRREQ_LEVEL[11], TCC_EA_WRREQ_DRAM_CREDIT_STALL[12], TCC_EA_WRREQ_GMI_CREDIT_STALL[12], TCC_EA_WRREQ_IO_CREDIT_STALL[12], TCC_EA_WRREQ_LEVEL[12], TCC_EA_WRREQ_DRAM_CREDIT_STALL[13], TCC_EA_WRREQ_GMI_CREDIT_STALL[13], TCC_EA_WRREQ_IO_CREDIT_STALL[13], TCC_EA_WRREQ_LEVEL[13], TCC_EA_WRREQ_DRAM_CREDIT_STALL[14], TCC_EA_WRREQ_GMI_CREDIT_STALL[14], TCC_EA_WRREQ_IO_CREDIT_STALL[14], TCC_EA_WRREQ_LEVEL[14], TCC_EA_WRREQ_DRAM_CREDIT_STALL[15], TCC_EA_WRREQ_GMI_CREDIT_STALL[15], TCC_EA_WRREQ_IO_CREDIT_STALL[15], TCC_EA_WRREQ_LEVEL[15], TCC_EA_WRREQ_DRAM_CREDIT_STALL[16], TCC_EA_WRREQ_GMI_CREDIT_STALL[16], TCC_EA_WRREQ_IO_CREDIT_STALL[16], TCC_EA_WRREQ_LEVEL[16], TCC_EA_WRREQ_DRAM_CREDIT_STALL[17], TCC_EA_WRREQ_GMI_CREDIT_STALL[17], TCC_EA_WRREQ_IO_CREDIT_STALL[17], TCC_EA_WRREQ_LEVEL[17], TCC_EA_WRREQ_DRAM_CREDIT_STALL[18], TCC_EA_WRREQ_GMI_CREDIT_STALL[18], TCC_EA_WRREQ_IO_CREDIT_STALL[18], TCC_EA_WRREQ_LEVEL[18], TCC_EA_WRREQ_DRAM_CREDIT_STALL[19], TCC_EA_WRREQ_GMI_CREDIT_STALL[19], TCC_EA_WRREQ_IO_CREDIT_STALL[19], TCC_EA_WRREQ_LEVEL[19], TCC_EA_WRREQ_DRAM_CREDIT_STALL[20], TCC_EA_WRREQ_GMI_CREDIT_STALL[20], TCC_EA_WRREQ_IO_CREDIT_STALL[20], TCC_EA_WRREQ_LEVEL[20], TCC_EA_WRREQ_DRAM_CREDIT_STALL[21], TCC_EA_WRREQ_GMI_CREDIT_STALL[21], TCC_EA_WRREQ_IO_CREDIT_STALL[21], TCC_EA_WRREQ_LEVEL[21], TCC_EA_WRREQ_DRAM_CREDIT_STALL[22], TCC_EA_WRREQ_GMI_CREDIT_STALL[22], TCC_EA_WRREQ_IO_CREDIT_STALL[22], TCC_EA_WRREQ_LEVEL[22], TCC_EA_WRREQ_DRAM_CREDIT_STALL[23], TCC_EA_WRREQ_GMI_CREDIT_STALL[23], TCC_EA_WRREQ_IO_CREDIT_STALL[23], TCC_EA_WRREQ_LEVEL[23], TCC_EA_WRREQ_DRAM_CREDIT_STALL[24], TCC_EA_WRREQ_GMI_CREDIT_STALL[24], TCC_EA_WRREQ_IO_CREDIT_STALL[24], TCC_EA_WRREQ_LEVEL[24], TCC_EA_WRREQ_DRAM_CREDIT_STALL[25], TCC_EA_WRREQ_GMI_CREDIT_STALL[25], TCC_EA_WRREQ_IO_CREDIT_STALL[25], TCC_EA_WRREQ_LEVEL[25], TCC_EA_WRREQ_DRAM_CREDIT_STALL[26], TCC_EA_WRREQ_GMI_CREDIT_STALL[26], TCC_EA_WRREQ_IO_CREDIT_STALL[26], TCC_EA_WRREQ_LEVEL[26], TCC_EA_WRREQ_DRAM_CREDIT_STALL[27], TCC_EA_WRREQ_GMI_CREDIT_STALL[27], TCC_EA_WRREQ_IO_CREDIT_STALL[27], TCC_EA_WRREQ_LEVEL[27], TCC_EA_WRREQ_DRAM_CREDIT_STALL[28], TCC_EA_WRREQ_GMI_CREDIT_STALL[28], TCC_EA_WRREQ_IO_CREDIT_STALL[28], TCC_EA_WRREQ_LEVEL[28], TCC_EA_WRREQ_DRAM_CREDIT_STALL[29], TCC_EA_WRREQ_GMI_CREDIT_STALL[29], TCC_EA_WRREQ_IO_CREDIT_STALL[29], TCC_EA_WRREQ_LEVEL[29], TCC_EA_WRREQ_DRAM_CREDIT_STALL[30], TCC_EA_WRREQ_GMI_CREDIT_STALL[30], TCC_EA_WRREQ_IO_CREDIT_STALL[30], TCC_EA_WRREQ_LEVEL[30], TCC_EA_WRREQ_DRAM_CREDIT_STALL[31], TCC_EA_WRREQ_GMI_CREDIT_STALL[31], TCC_EA_WRREQ_IO_CREDIT_STALL[31], TCC_EA_WRREQ_LEVEL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155053_1198055/input0_results_240321_155053
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_14.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_15.txt
|-> [rocprof] RPL: on '240321_155053' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_15.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155053_1198258'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155053_1198258/input0_results_240321_155053'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155053_1198258/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_HIT[0], TCC_MISS[0], TCC_READ[0], TCC_REQ[0], TCC_HIT[1], TCC_MISS[1], TCC_READ[1], TCC_REQ[1], TCC_HIT[2], TCC_MISS[2], TCC_READ[2], TCC_REQ[2], TCC_HIT[3], TCC_MISS[3], TCC_READ[3], TCC_REQ[3], TCC_HIT[4], TCC_MISS[4], TCC_READ[4], TCC_REQ[4], TCC_HIT[5], TCC_MISS[5], TCC_READ[5], TCC_REQ[5], TCC_HIT[6], TCC_MISS[6], TCC_READ[6], TCC_REQ[6], TCC_HIT[7], TCC_MISS[7], TCC_READ[7], TCC_REQ[7], TCC_HIT[8], TCC_MISS[8], TCC_READ[8], TCC_REQ[8], TCC_HIT[9], TCC_MISS[9], TCC_READ[9], TCC_REQ[9], TCC_HIT[10], TCC_MISS[10], TCC_READ[10], TCC_REQ[10], TCC_HIT[11], TCC_MISS[11], TCC_READ[11], TCC_REQ[11], TCC_HIT[12], TCC_MISS[12], TCC_READ[12], TCC_REQ[12], TCC_HIT[13], TCC_MISS[13], TCC_READ[13], TCC_REQ[13], TCC_HIT[14], TCC_MISS[14], TCC_READ[14], TCC_REQ[14], TCC_HIT[15], TCC_MISS[15], TCC_READ[15], TCC_REQ[15], TCC_HIT[16], TCC_MISS[16], TCC_READ[16], TCC_REQ[16], TCC_HIT[17], TCC_MISS[17], TCC_READ[17], TCC_REQ[17], TCC_HIT[18], TCC_MISS[18], TCC_READ[18], TCC_REQ[18], TCC_HIT[19], TCC_MISS[19], TCC_READ[19], TCC_REQ[19], TCC_HIT[20], TCC_MISS[20], TCC_READ[20], TCC_REQ[20], TCC_HIT[21], TCC_MISS[21], TCC_READ[21], TCC_REQ[21], TCC_HIT[22], TCC_MISS[22], TCC_READ[22], TCC_REQ[22], TCC_HIT[23], TCC_MISS[23], TCC_READ[23], TCC_REQ[23], TCC_HIT[24], TCC_MISS[24], TCC_READ[24], TCC_REQ[24], TCC_HIT[25], TCC_MISS[25], TCC_READ[25], TCC_REQ[25], TCC_HIT[26], TCC_MISS[26], TCC_READ[26], TCC_REQ[26], TCC_HIT[27], TCC_MISS[27], TCC_READ[27], TCC_REQ[27], TCC_HIT[28], TCC_MISS[28], TCC_READ[28], TCC_REQ[28], TCC_HIT[29], TCC_MISS[29], TCC_READ[29], TCC_REQ[29], TCC_HIT[30], TCC_MISS[30], TCC_READ[30], TCC_REQ[30], TCC_HIT[31], TCC_MISS[31], TCC_READ[31], TCC_REQ[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155053_1198258/input0_results_240321_155053
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_15.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_16.txt
|-> [rocprof] RPL: on '240321_155054' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_16.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155054_1198460'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155054_1198460/input0_results_240321_155054'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155054_1198460/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 96 metrics
|-> [rocprof] TCC_RW_REQ[0], TCC_TOO_MANY_EA_WRREQS_STALL[0], TCC_WRITE[0], TCC_RW_REQ[1], TCC_TOO_MANY_EA_WRREQS_STALL[1], TCC_WRITE[1], TCC_RW_REQ[2], TCC_TOO_MANY_EA_WRREQS_STALL[2], TCC_WRITE[2], TCC_RW_REQ[3], TCC_TOO_MANY_EA_WRREQS_STALL[3], TCC_WRITE[3], TCC_RW_REQ[4], TCC_TOO_MANY_EA_WRREQS_STALL[4], TCC_WRITE[4], TCC_RW_REQ[5], TCC_TOO_MANY_EA_WRREQS_STALL[5], TCC_WRITE[5], TCC_RW_REQ[6], TCC_TOO_MANY_EA_WRREQS_STALL[6], TCC_WRITE[6], TCC_RW_REQ[7], TCC_TOO_MANY_EA_WRREQS_STALL[7], TCC_WRITE[7], TCC_RW_REQ[8], TCC_TOO_MANY_EA_WRREQS_STALL[8], TCC_WRITE[8], TCC_RW_REQ[9], TCC_TOO_MANY_EA_WRREQS_STALL[9], TCC_WRITE[9], TCC_RW_REQ[10], TCC_TOO_MANY_EA_WRREQS_STALL[10], TCC_WRITE[10], TCC_RW_REQ[11], TCC_TOO_MANY_EA_WRREQS_STALL[11], TCC_WRITE[11], TCC_RW_REQ[12], TCC_TOO_MANY_EA_WRREQS_STALL[12], TCC_WRITE[12], TCC_RW_REQ[13], TCC_TOO_MANY_EA_WRREQS_STALL[13], TCC_WRITE[13], TCC_RW_REQ[14], TCC_TOO_MANY_EA_WRREQS_STALL[14], TCC_WRITE[14], TCC_RW_REQ[15], TCC_TOO_MANY_EA_WRREQS_STALL[15], TCC_WRITE[15], TCC_RW_REQ[16], TCC_TOO_MANY_EA_WRREQS_STALL[16], TCC_WRITE[16], TCC_RW_REQ[17], TCC_TOO_MANY_EA_WRREQS_STALL[17], TCC_WRITE[17], TCC_RW_REQ[18], TCC_TOO_MANY_EA_WRREQS_STALL[18], TCC_WRITE[18], TCC_RW_REQ[19], TCC_TOO_MANY_EA_WRREQS_STALL[19], TCC_WRITE[19], TCC_RW_REQ[20], TCC_TOO_MANY_EA_WRREQS_STALL[20], TCC_WRITE[20], TCC_RW_REQ[21], TCC_TOO_MANY_EA_WRREQS_STALL[21], TCC_WRITE[21], TCC_RW_REQ[22], TCC_TOO_MANY_EA_WRREQS_STALL[22], TCC_WRITE[22], TCC_RW_REQ[23], TCC_TOO_MANY_EA_WRREQS_STALL[23], TCC_WRITE[23], TCC_RW_REQ[24], TCC_TOO_MANY_EA_WRREQS_STALL[24], TCC_WRITE[24], TCC_RW_REQ[25], TCC_TOO_MANY_EA_WRREQS_STALL[25], TCC_WRITE[25], TCC_RW_REQ[26], TCC_TOO_MANY_EA_WRREQS_STALL[26], TCC_WRITE[26], TCC_RW_REQ[27], TCC_TOO_MANY_EA_WRREQS_STALL[27], TCC_WRITE[27], TCC_RW_REQ[28], TCC_TOO_MANY_EA_WRREQS_STALL[28], TCC_WRITE[28], TCC_RW_REQ[29], TCC_TOO_MANY_EA_WRREQS_STALL[29], TCC_WRITE[29], TCC_RW_REQ[30], TCC_TOO_MANY_EA_WRREQS_STALL[30], TCC_WRITE[30], TCC_RW_REQ[31], TCC_TOO_MANY_EA_WRREQS_STALL[31], TCC_WRITE[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155054_1198460/input0_results_240321_155054
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_16.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_2.txt
|-> [rocprof] RPL: on '240321_155055' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_2.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155055_1198662'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155055_1198662/input0_results_240321_155055'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155055_1198662/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 26 metrics
|-> [rocprof] SQC_DCACHE_INPUT_VALID_READYB, SQC_DCACHE_ATOMIC, SQC_DCACHE_REQ_READ_8, SQC_DCACHE_REQ, SQC_DCACHE_HITS, SQC_DCACHE_MISSES, SQC_DCACHE_MISSES_DUPLICATE, SQC_DCACHE_REQ_READ_1, TCP_VOLATILE_sum, TCP_TOTAL_ACCESSES_sum, TCP_TOTAL_READ_sum, TCP_TOTAL_WRITE_sum, TA_BUFFER_ATOMIC_WAVEFRONTS_sum, TA_BUFFER_TOTAL_CYCLES_sum, TD_ATOMIC_WAVEFRONT_sum, TD_STORE_WAVEFRONT_sum, SPI_RA_REQ_NO_ALLOC, SPI_RA_REQ_NO_ALLOC_CSN, CPC_CPC_STAT_STALL, CPC_UTCL1_STALL_ON_TRANSLATION, CPF_CPF_STAT_IDLE, CPF_CPF_TCIU_IDLE, TCC_REQ_sum, TCC_STREAMING_REQ_sum, TCC_HIT_sum, TCC_MISS_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155055_1198662/input0_results_240321_155055
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_2.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_3.txt
|-> [rocprof] RPL: on '240321_155055' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_3.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155055_1198863'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155055_1198863/input0_results_240321_155055'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155055_1198863/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 23 metrics
|-> [rocprof] SQC_DCACHE_REQ_READ_2, SQC_DCACHE_REQ_READ_4, SQ_INSTS_VMEM_WR, SQ_INSTS_VMEM_RD, SQ_INSTS_VMEM, SQ_INSTS_SALU, SQ_INSTS_VSKIPPED, SQ_INSTS_SMEM, TCP_TOTAL_ATOMIC_WITH_RET_sum, TCP_TOTAL_ATOMIC_WITHOUT_RET_sum, TCP_TOTAL_WRITEBACK_INVALIDATES_sum, TCP_TOTAL_CACHE_ACCESSES_sum, TA_BUFFER_COALESCED_READ_CYCLES_sum, TA_BUFFER_COALESCED_WRITE_CYCLES_sum, SPI_RA_RES_STALL_CSN, SPI_RA_TMP_STALL_CSN, CPC_CPC_UTCL2IU_BUSY, CPC_CPC_UTCL2IU_IDLE, CPF_CMP_UTCL1_STALL_ON_TRANSLATION, TCC_READ_sum, TCC_WRITE_sum, TCC_ATOMIC_sum, TCC_WRITEBACK_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155055_1198863/input0_results_240321_155055
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_3.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_4.txt
|-> [rocprof] RPL: on '240321_155056' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_4.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155056_1199051'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155056_1199051/input0_results_240321_155056'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155056_1199051/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 22 metrics
|-> [rocprof] SQ_INSTS_FLAT, SQ_INSTS_LDS, SQ_INSTS_GDS, SQ_INSTS_EXP_GDS, SQ_INSTS_BRANCH, SQ_INSTS_SENDMSG, SQ_INSTS, SQ_WAIT_ANY, TCP_UTCL1_TRANSLATION_MISS_sum, TCP_UTCL1_TRANSLATION_HIT_sum, TCP_UTCL1_PERMISSION_MISS_sum, TCP_UTCL1_REQUEST_sum, TA_ADDR_STALLED_BY_TC_CYCLES_sum, TA_TOTAL_WAVEFRONTS_sum, SPI_RA_WAVE_SIMD_FULL_CSN, SPI_RA_VGPR_SIMD_FULL_CSN, CPC_CPC_UTCL2IU_STALL, CPC_ME1_BUSY_FOR_PACKET_DECODE, TCC_EA_WRREQ_sum, TCC_EA_WRREQ_64B_sum, TCC_EA_WR_UNCACHED_32B_sum, TCC_EA_WRREQ_DRAM_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155056_1199051/input0_results_240321_155056
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_4.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_5.txt
|-> [rocprof] RPL: on '240321_155056' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_5.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155056_1199253'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155056_1199253/input0_results_240321_155056'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155056_1199253/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 21 metrics
|-> [rocprof] SQ_WAIT_INST_ANY, SQ_ACTIVE_INST_ANY, SQ_INSTS_VALU, SQ_ACTIVE_INST_VMEM, SQ_ACTIVE_INST_LDS, SQ_ACTIVE_INST_VALU, SQ_ACTIVE_INST_SCA, SQ_ACTIVE_INST_EXP_GDS, TCP_TCP_LATENCY_sum, TCP_TCC_READ_REQ_LATENCY_sum, TCP_TCC_WRITE_REQ_LATENCY_sum, TCP_TCC_READ_REQ_sum, TA_ADDR_STALLED_BY_TD_CYCLES_sum, TA_DATA_STALLED_BY_TC_CYCLES_sum, SPI_RA_SGPR_SIMD_FULL_CSN, SPI_RA_LDS_CU_FULL_CSN, CPC_ME1_DC0_SPI_BUSY, TCC_EA_WRREQ_STALL_sum, TCC_EA_WRREQ_IO_CREDIT_STALL_sum, TCC_EA_WRREQ_GMI_CREDIT_STALL_sum, TCC_EA_WRREQ_DRAM_CREDIT_STALL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155056_1199253/input0_results_240321_155056
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_5.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_6.txt
|-> [rocprof] RPL: on '240321_155057' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_6.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155057_1199438'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155057_1199438/input0_results_240321_155057'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155057_1199438/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 20 metrics
|-> [rocprof] SQ_ACTIVE_INST_MISC, SQ_ACTIVE_INST_FLAT, SQ_INST_CYCLES_VMEM_WR, SQ_INST_CYCLES_VMEM_RD, SQ_INST_CYCLES_SMEM, SQ_INST_CYCLES_SALU, SQ_THREAD_CYCLES_VALU, SQ_IFETCH, TCP_TCC_WRITE_REQ_sum, TCP_TCC_ATOMIC_WITH_RET_REQ_sum, TCP_TCC_ATOMIC_WITHOUT_RET_REQ_sum, TCP_TCC_NC_READ_REQ_sum, TA_FLAT_WAVEFRONTS_sum, TA_FLAT_READ_WAVEFRONTS_sum, SPI_RA_BAR_CU_FULL_CSN, SPI_RA_TGLIM_CU_FULL_CSN, TCC_EA_RDREQ_sum, TCC_EA_RDREQ_32B_sum, TCC_EA_RD_UNCACHED_32B_sum, TCC_EA_RDREQ_DRAM_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155057_1199438/input0_results_240321_155057
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_6.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_7.txt
|-> [rocprof] RPL: on '240321_155057' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_7.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155057_1199640'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155057_1199640/input0_results_240321_155057'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155057_1199640/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 20 metrics
|-> [rocprof] SQ_LDS_BANK_CONFLICT, SQ_LDS_ADDR_CONFLICT, SQ_LDS_UNALIGNED_STALL, SQ_WAVES_EQ_64, SQ_WAVES_LT_64, SQ_WAVES_LT_48, SQ_WAVES_LT_32, SQ_WAVES_LT_16, TCP_TCC_NC_WRITE_REQ_sum, TCP_TCC_NC_ATOMIC_REQ_sum, TCP_TCC_UC_READ_REQ_sum, TCP_TCC_UC_WRITE_REQ_sum, TA_FLAT_WRITE_WAVEFRONTS_sum, TA_FLAT_ATOMIC_WAVEFRONTS_sum, SPI_RA_WVLIM_STALL_CSN, SPI_SWC_CSC_WR, TCC_EA_RDREQ_IO_CREDIT_STALL_sum, TCC_EA_RDREQ_GMI_CREDIT_STALL_sum, TCC_EA_RDREQ_DRAM_CREDIT_STALL_sum, TCC_TAG_STALL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155057_1199640/input0_results_240321_155057
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_7.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_8.txt
|-> [rocprof] RPL: on '240321_155058' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_8.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155058_1199843'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155058_1199843/input0_results_240321_155058'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155058_1199843/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 17 metrics
|-> [rocprof] SQ_ITEMS, SQ_LDS_MEM_VIOLATIONS, SQ_LDS_ATOMIC_RETURN, SQ_LDS_IDX_ACTIVE, SQ_WAVES_RESTORED, SQ_WAVES_SAVED, SQ_INSTS_SMEM_NORM, TCP_TCC_UC_ATOMIC_REQ_sum, TCP_TCC_CC_READ_REQ_sum, TCP_TCC_CC_WRITE_REQ_sum, TCP_TCC_CC_ATOMIC_REQ_sum, SPI_VWC_CSC_WR, SPI_RA_BULKY_CU_FULL_CSN, TCC_NORMAL_WRITEBACK_sum, TCC_ALL_TC_OP_WB_WRITEBACK_sum, TCC_NORMAL_EVICT_sum, TCC_ALL_TC_OP_INV_EVICT_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155058_1199843/input0_results_240321_155058
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_8.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_9.txt
|-> [rocprof] RPL: on '240321_155058' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/pmc_perf_9.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155058_1200045'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155058_1200045/input0_results_240321_155058'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155058_1200045/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 8 metrics
|-> [rocprof] TCP_TCC_RW_READ_REQ_sum, TCP_TCC_RW_WRITE_REQ_sum, TCP_TCC_RW_ATOMIC_REQ_sum, TCP_PENDING_STALL_CYCLES_sum, TCC_TOO_MANY_EA_WRREQS_STALL_sum, TCC_EA_ATOMIC_sum, TCC_EA_RDREQ_LEVEL_sum, TCC_EA_WRREQ_LEVEL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155058_1200045/input0_results_240321_155058
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/pmc_perf_9.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI100/perfmon/timestamps.txt
|-> [rocprof] RPL: on '240321_155059' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI100/perfmon/timestamps.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_155059_1200248'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_155059_1200248/input0_results_240321_155059'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_155059_1200248/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 0 metrics
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_155059_1200248/input0_results_240321_155059
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI100/timestamps.csv' is generating
|-> [rocprof]
tests/workloads/dispatch_inv/MI100/pmc_dispatch_info.csv
+2
View File
@@ -0,0 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2
1 Dispatch_ID Kernel_Name GPU_ID
2 0 vecCopy(double*, double*, double*, int, int) [clone .kd] 2
tests/workloads/dispatch_inv/MI100/pmc_perf.csv
+2 -2
View File
File diff suppressed because one or more lines are too long
tests/workloads/dispatch_inv/MI100/sysinfo.csv
+1 -1
View File
@@ -1,2 +1,2 @@
workload_name,command,ip_blocks,timestamp,version,hostname,cpu_model,sbios,linux_distro,linux_kernel_version,amd_gpu_kernel_version,cpu_memory,gpu_memory,rocm_version,vbios,compute_partition,memory_partition,gpu_model,gpu_arch,gpu_l1,gpu_l2,cu_per_gpu,simd_per_cu,se_per_gpu,wave_size,workgroup_max_size,max_waves_per_cu,max_sclk,max_mclk,cur_sclk,cur_mclk,total_l2_chan,lds_banks_per_cu,sqc_per_gpu,pipes_per_gpu,hbm_bw,num_xcd
dispatch_inv,./tests/vcopy -n 1048576 -b 256 -i 3,SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF,Thu 07 Mar 2024 01:05:16 PM (CST),2,t008-007.hpcfund,AMD EPYC 7V13 64-Core Processor,American Megatrends Inc.0602,Rocky Linux 9.1 (Blue Onyx),5.14.0-162.18.1.el9_1.x86_64,,527651100,,6.0.2-115,113-D3431401-100,NA,NA,MI100,gfx908,16,8192,120,4,8,64,1024,40,1502,1200,1502,1200,32,32,64,4,1228.8,1
dispatch_inv,./tests/vcopy -n 1048576 -b 256 -i 3,SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF,Thu 21 Mar 2024 03:50:46 PM (CDT),2,t007-001.hpcfund,AMD EPYC 7V13 64-Core Processor,American Megatrends Inc.0602,Rocky Linux 9.1 (Blue Onyx),5.14.0-162.18.1.el9_1.x86_64,,527651008,,6.0.2-115,113-D3431401-100,NA,NA,MI100,gfx908,16,8192,120,4,8,64,1024,40,1502,1200,1502,1200,32,32,64,4,1228.8,1
1 workload_name command ip_blocks timestamp version hostname cpu_model sbios linux_distro linux_kernel_version amd_gpu_kernel_version cpu_memory gpu_memory rocm_version vbios compute_partition memory_partition gpu_model gpu_arch gpu_l1 gpu_l2 cu_per_gpu simd_per_cu se_per_gpu wave_size workgroup_max_size max_waves_per_cu max_sclk max_mclk cur_sclk cur_mclk total_l2_chan lds_banks_per_cu sqc_per_gpu pipes_per_gpu hbm_bw num_xcd
2 dispatch_inv ./tests/vcopy -n 1048576 -b 256 -i 3 SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF Thu 07 Mar 2024 01:05:16 PM (CST) Thu 21 Mar 2024 03:50:46 PM (CDT) 2 t008-007.hpcfund t007-001.hpcfund AMD EPYC 7V13 64-Core Processor American Megatrends Inc.0602 Rocky Linux 9.1 (Blue Onyx) 5.14.0-162.18.1.el9_1.x86_64 527651100 527651008 6.0.2-115 113-D3431401-100 NA NA MI100 gfx908 16 8192 120 4 8 64 1024 40 1502 1200 1502 1200 32 32 64 4 1228.8 1
tests/workloads/dispatch_inv/MI100/timestamps.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,104717,104717,1048576,256,0,0,8,8,16,64,0x0,0x7fa555e9ee80,194063713493885,194063713519257,194063713543257,194063713554520
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,1200408,1200408,1048576,256,0,0,8,8,16,64,0x0,0x7f761fe58ec0,1409993151454587,1409993151479981,1409993151504461,1409993151517756
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 104717 1200408 104717 1200408 1048576 256 0 0 8 8 16 64 0x0 0x7fa555e9ee80 0x7f761fe58ec0 194063713493885 1409993151454587 194063713519257 1409993151479981 194063713543257 1409993151504461 194063713554520 1409993151517756
tests/workloads/dispatch_inv/MI200/SQ_IFETCH_LEVEL.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,GRBM_COUNT,GRBM_GUI_ACTIVE,SQ_WAVES,SQ_IFETCH,SQ_IFETCH_LEVEL,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,248270,248270,1048576,256,0,0,8,0,16,64,0x0,0x7fee6c1e8ec0,28159,28159,16384,65536,18528,2096508,197314699318636,197327446943430,197327446964070,197314715555093
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4070883,4070883,1048576,256,0,0,8,0,16,64,0x0,0x7f3290b00ec0,27605,27605,16384,65536,13878,1517264,1410706212315522,1410719177571898,1410719177592538,1410706228473689
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj GRBM_COUNT GRBM_GUI_ACTIVE SQ_WAVES SQ_IFETCH SQ_IFETCH_LEVEL SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 248270 4070883 248270 4070883 1048576 256 0 0 8 0 16 64 0x0 0x7fee6c1e8ec0 0x7f3290b00ec0 28159 27605 28159 27605 16384 65536 18528 13878 2096508 1517264 197314699318636 1410706212315522 197327446943430 1410719177571898 197327446964070 1410719177592538 197314715555093 1410706228473689
tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_LDS.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_LDS,SQ_INST_LEVEL_LDS,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,248472,248472,1048576,256,0,0,8,0,16,64,0x0,0x7f335e55cec0,0,0,0,197315182575180,197327446943430,197327446964070,197315198215128
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4071085,4071085,1048576,256,0,0,8,0,16,64,0x0,0x7fc0c4880ec0,0,0,0,1410706705601444,1410719177571898,1410719177592538,1410706721607704
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_LDS SQ_INST_LEVEL_LDS SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 248472 4071085 248472 4071085 1048576 256 0 0 8 0 16 64 0x0 0x7f335e55cec0 0x7fc0c4880ec0 0 0 0 197315182575180 1410706705601444 197327446943430 1410719177571898 197327446964070 1410719177592538 197315198215128 1410706721607704
tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_SMEM.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_SMEM,SQ_INST_LEVEL_SMEM,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,248674,248674,1048576,256,0,0,8,0,16,64,0x0,0x7f09405d0ec0,65536,187198,20922392,197315663370888,197327446943430,197327446964070,197315679715439
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4071277,4071277,1048576,256,0,0,8,0,16,64,0x0,0x7f48680b0ec0,65536,207168,23284944,1410707198038150,1410719177571898,1410719177592538,1410707213906169
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_SMEM SQ_INST_LEVEL_SMEM SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 248674 4071277 248674 4071277 1048576 256 0 0 8 0 16 64 0x0 0x7f09405d0ec0 0x7f48680b0ec0 65536 187198 207168 20922392 23284944 197315663370888 1410707198038150 197327446943430 1410719177571898 197327446964070 1410719177592538 197315679715439 1410707213906169
tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_VMEM.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,SQ_INSTS_VMEM,SQ_INST_LEVEL_VMEM,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,248877,248877,1048576,256,0,0,8,0,16,64,0x0,0x7f819db68ec0,32768,293897,32918400,197316140235970,197327446943430,197327446964070,197316156066459
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4071462,4071462,1048576,256,0,0,8,0,16,64,0x0,0x7f949b134ec0,32768,296236,33173792,1410707687433184,1410719177571898,1410719177592538,1410707703242482
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj SQ_INSTS_VMEM SQ_INST_LEVEL_VMEM SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 248877 4071462 248877 4071462 1048576 256 0 0 8 0 16 64 0x0 0x7f819db68ec0 0x7f949b134ec0 32768 293897 296236 32918400 33173792 197316140235970 1410707687433184 197327446943430 1410719177571898 197327446964070 1410719177592538 197316156066459 1410707703242482
tests/workloads/dispatch_inv/MI200/SQ_LEVEL_WAVES.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,GRBM_COUNT,GRBM_GUI_ACTIVE,CPC_ME1_BUSY_FOR_PACKET_DECODE,SQ_CYCLES,SQ_WAVES,SQ_WAVE_CYCLES,SQ_BUSY_CYCLES,SQ_LEVEL_WAVES,SQ_ACCUM_PREV_HIRES,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) ",2,0,0,249077,249077,1048576,256,0,0,8,0,16,64,0x0,0x7f2a00ac0ec0,27945,27945,11402,223568,16384,10871098,128273,0,43994300,197316621462013,197327446943430,197327446964070,197316637162384
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4071665,4071665,1048576,256,0,0,8,0,16,64,0x0,0x7fb260a8cec0,28141,28141,11099,225136,16384,11221694,132697,0,45401376,1410708182655689,1410719177571898,1410719177592538,1410708199397810
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj GRBM_COUNT GRBM_GUI_ACTIVE CPC_ME1_BUSY_FOR_PACKET_DECODE SQ_CYCLES SQ_WAVES SQ_WAVE_CYCLES SQ_BUSY_CYCLES SQ_LEVEL_WAVES SQ_ACCUM_PREV_HIRES DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 249077 4071665 249077 4071665 1048576 256 0 0 8 0 16 64 0x0 0x7f2a00ac0ec0 0x7fb260a8cec0 27945 28141 27945 28141 11402 11099 223568 225136 16384 10871098 11221694 128273 132697 0 43994300 45401376 197316621462013 1410708182655689 197327446943430 1410719177571898 197327446964070 1410719177592538 197316637162384 1410708199397810
tests/workloads/dispatch_inv/MI200/log.txt
+764
View File
@@ -0,0 +1,764 @@
Omniperf version: 2.0.0-RC1
Profiler choice: rocprofv1
Path: /home1/josantos/omniperf/tests/workloads/dispatch_inv/MI200
Target: MI200
Command: ./tests/vcopy -n 1048576 -b 256 -i 3
Kernel Selection: None
Dispatch Selection: ['invalid']
IP Blocks: All
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Collecting Performance Counters
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/SQ_IFETCH_LEVEL.txt
|-> [rocprof] RPL: on '240321_160253' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/SQ_IFETCH_LEVEL.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160253_4070723'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160253_4070723/input0_results_240321_160253'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160253_4070723/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 6 metrics
|-> [rocprof] GRBM_COUNT, GRBM_GUI_ACTIVE, SQ_WAVES, SQ_IFETCH, SQ_IFETCH_LEVEL, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160253_4070723/input0_results_240321_160253
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/SQ_IFETCH_LEVEL.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_LDS.txt
|-> [rocprof] RPL: on '240321_160254' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_LDS.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160254_4070925'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160254_4070925/input0_results_240321_160254'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160254_4070925/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_LDS, SQ_INST_LEVEL_LDS, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160254_4070925/input0_results_240321_160254
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_LDS.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_SMEM.txt
|-> [rocprof] RPL: on '240321_160254' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_SMEM.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160254_4071117'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160254_4071117/input0_results_240321_160254'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160254_4071117/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_SMEM, SQ_INST_LEVEL_SMEM, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160254_4071117/input0_results_240321_160254
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_SMEM.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_VMEM.txt
|-> [rocprof] RPL: on '240321_160255' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/SQ_INST_LEVEL_VMEM.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160255_4071302'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160255_4071302/input0_results_240321_160255'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160255_4071302/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQ_INSTS_VMEM, SQ_INST_LEVEL_VMEM, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160255_4071302/input0_results_240321_160255
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/SQ_INST_LEVEL_VMEM.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/SQ_LEVEL_WAVES.txt
|-> [rocprof] RPL: on '240321_160255' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/SQ_LEVEL_WAVES.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160255_4071505'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160255_4071505/input0_results_240321_160255'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160255_4071505/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 9 metrics
|-> [rocprof] GRBM_COUNT, GRBM_GUI_ACTIVE, CPC_ME1_BUSY_FOR_PACKET_DECODE, SQ_CYCLES, SQ_WAVES, SQ_WAVE_CYCLES, SQ_BUSY_CYCLES, SQ_LEVEL_WAVES, SQ_ACCUM_PREV_HIRES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160255_4071505/input0_results_240321_160255
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/SQ_LEVEL_WAVES.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_0.txt
|-> [rocprof] RPL: on '240321_160256' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_0.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160256_4071704'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160256_4071704/input0_results_240321_160256'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160256_4071704/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 28 metrics
|-> [rocprof] SQ_CYCLES, SQ_BUSY_CYCLES, SQ_WAVES, SQ_INSTS_VALU_CVT, SQ_INSTS_VMEM_WR, SQ_INSTS_VMEM_RD, SQ_INSTS_VMEM, SQ_INSTS_SALU, GRBM_COUNT, GRBM_GUI_ACTIVE, TCP_GATE_EN1_sum, TCP_GATE_EN2_sum, TCP_TD_TCP_STALL_CYCLES_sum, TCP_TCR_TCP_STALL_CYCLES_sum, TA_TA_BUSY_sum, TA_BUFFER_WAVEFRONTS_sum, TD_TD_BUSY_sum, TD_TC_STALL_sum, SPI_CSN_WINDOW_VALID, SPI_CSN_BUSY, CPC_CPC_STAT_BUSY, CPC_CPC_STAT_IDLE, CPF_CPF_STAT_BUSY, CPF_CPF_STAT_STALL, TCC_CYCLE_sum, TCC_BUSY_sum, TCC_PROBE_sum, TCC_PROBE_ALL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160256_4071704/input0_results_240321_160256
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_0.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_1.txt
|-> [rocprof] RPL: on '240321_160256' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_1.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160256_4071889'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160256_4071889/input0_results_240321_160256'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160256_4071889/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 27 metrics
|-> [rocprof] SQ_INSTS_VSKIPPED, SQ_INSTS, SQ_INSTS_VALU, SQ_INSTS_VALU_ADD_F16, SQ_INSTS_VALU_MUL_F16, SQ_INSTS_VALU_FMA_F16, SQ_INSTS_VALU_TRANS_F16, SQ_INSTS_VALU_ADD_F32, GRBM_SPI_BUSY, TCP_READ_TAGCONFLICT_STALL_CYCLES_sum, TCP_WRITE_TAGCONFLICT_STALL_CYCLES_sum, TCP_ATOMIC_TAGCONFLICT_STALL_CYCLES_sum, TCP_TA_TCP_STATE_READ_sum, TA_BUFFER_READ_WAVEFRONTS_sum, TA_BUFFER_WRITE_WAVEFRONTS_sum, TD_SPI_STALL_sum, TD_LOAD_WAVEFRONT_sum, SPI_CSN_NUM_THREADGROUPS, SPI_CSN_WAVE, CPC_CPC_TCIU_BUSY, CPC_CPC_TCIU_IDLE, CPF_CPF_TCIU_BUSY, CPF_CPF_TCIU_STALL, TCC_NC_REQ_sum, TCC_UC_REQ_sum, TCC_CC_REQ_sum, TCC_RW_REQ_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160256_4071889/input0_results_240321_160256
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_1.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_10.txt
|-> [rocprof] RPL: on '240321_160257' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_10.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160257_4072074'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160257_4072074/input0_results_240321_160257'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160257_4072074/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 8 metrics
|-> [rocprof] SQC_TC_DATA_WRITE_REQ, SQC_TC_DATA_ATOMIC_REQ, SQC_TC_STALL, SQC_TC_REQ, SQC_DCACHE_REQ_READ_16, SQC_ICACHE_REQ, SQC_ICACHE_HITS, SQC_ICACHE_MISSES
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160257_4072074/input0_results_240321_160257
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_10.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_11.txt
|-> [rocprof] RPL: on '240321_160257' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_11.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160257_4072259'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160257_4072259/input0_results_240321_160257'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160257_4072259/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 8 metrics
|-> [rocprof] SQC_ICACHE_MISSES_DUPLICATE, SQC_DCACHE_INPUT_VALID_READYB, SQC_DCACHE_ATOMIC, SQC_DCACHE_REQ_READ_8, SQC_DCACHE_REQ, SQC_DCACHE_HITS, SQC_DCACHE_MISSES, SQC_DCACHE_MISSES_DUPLICATE
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160257_4072259/input0_results_240321_160257
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_11.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_12.txt
|-> [rocprof] RPL: on '240321_160258' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_12.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160258_4072456'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160258_4072456/input0_results_240321_160258'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160258_4072456/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 3 metrics
|-> [rocprof] SQC_DCACHE_REQ_READ_1, SQC_DCACHE_REQ_READ_2, SQC_DCACHE_REQ_READ_4
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160258_4072456/input0_results_240321_160258
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_12.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_13.txt
|-> [rocprof] RPL: on '240321_160258' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_13.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160258_4072642'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160258_4072642/input0_results_240321_160258'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160258_4072642/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_ATOMIC[0], TCC_CYCLE[0], TCC_EA_ATOMIC[0], TCC_EA_ATOMIC_LEVEL[0], TCC_ATOMIC[1], TCC_CYCLE[1], TCC_EA_ATOMIC[1], TCC_EA_ATOMIC_LEVEL[1], TCC_ATOMIC[2], TCC_CYCLE[2], TCC_EA_ATOMIC[2], TCC_EA_ATOMIC_LEVEL[2], TCC_ATOMIC[3], TCC_CYCLE[3], TCC_EA_ATOMIC[3], TCC_EA_ATOMIC_LEVEL[3], TCC_ATOMIC[4], TCC_CYCLE[4], TCC_EA_ATOMIC[4], TCC_EA_ATOMIC_LEVEL[4], TCC_ATOMIC[5], TCC_CYCLE[5], TCC_EA_ATOMIC[5], TCC_EA_ATOMIC_LEVEL[5], TCC_ATOMIC[6], TCC_CYCLE[6], TCC_EA_ATOMIC[6], TCC_EA_ATOMIC_LEVEL[6], TCC_ATOMIC[7], TCC_CYCLE[7], TCC_EA_ATOMIC[7], TCC_EA_ATOMIC_LEVEL[7], TCC_ATOMIC[8], TCC_CYCLE[8], TCC_EA_ATOMIC[8], TCC_EA_ATOMIC_LEVEL[8], TCC_ATOMIC[9], TCC_CYCLE[9], TCC_EA_ATOMIC[9], TCC_EA_ATOMIC_LEVEL[9], TCC_ATOMIC[10], TCC_CYCLE[10], TCC_EA_ATOMIC[10], TCC_EA_ATOMIC_LEVEL[10], TCC_ATOMIC[11], TCC_CYCLE[11], TCC_EA_ATOMIC[11], TCC_EA_ATOMIC_LEVEL[11], TCC_ATOMIC[12], TCC_CYCLE[12], TCC_EA_ATOMIC[12], TCC_EA_ATOMIC_LEVEL[12], TCC_ATOMIC[13], TCC_CYCLE[13], TCC_EA_ATOMIC[13], TCC_EA_ATOMIC_LEVEL[13], TCC_ATOMIC[14], TCC_CYCLE[14], TCC_EA_ATOMIC[14], TCC_EA_ATOMIC_LEVEL[14], TCC_ATOMIC[15], TCC_CYCLE[15], TCC_EA_ATOMIC[15], TCC_EA_ATOMIC_LEVEL[15], TCC_ATOMIC[16], TCC_CYCLE[16], TCC_EA_ATOMIC[16], TCC_EA_ATOMIC_LEVEL[16], TCC_ATOMIC[17], TCC_CYCLE[17], TCC_EA_ATOMIC[17], TCC_EA_ATOMIC_LEVEL[17], TCC_ATOMIC[18], TCC_CYCLE[18], TCC_EA_ATOMIC[18], TCC_EA_ATOMIC_LEVEL[18], TCC_ATOMIC[19], TCC_CYCLE[19], TCC_EA_ATOMIC[19], TCC_EA_ATOMIC_LEVEL[19], TCC_ATOMIC[20], TCC_CYCLE[20], TCC_EA_ATOMIC[20], TCC_EA_ATOMIC_LEVEL[20], TCC_ATOMIC[21], TCC_CYCLE[21], TCC_EA_ATOMIC[21], TCC_EA_ATOMIC_LEVEL[21], TCC_ATOMIC[22], TCC_CYCLE[22], TCC_EA_ATOMIC[22], TCC_EA_ATOMIC_LEVEL[22], TCC_ATOMIC[23], TCC_CYCLE[23], TCC_EA_ATOMIC[23], TCC_EA_ATOMIC_LEVEL[23], TCC_ATOMIC[24], TCC_CYCLE[24], TCC_EA_ATOMIC[24], TCC_EA_ATOMIC_LEVEL[24], TCC_ATOMIC[25], TCC_CYCLE[25], TCC_EA_ATOMIC[25], TCC_EA_ATOMIC_LEVEL[25], TCC_ATOMIC[26], TCC_CYCLE[26], TCC_EA_ATOMIC[26], TCC_EA_ATOMIC_LEVEL[26], TCC_ATOMIC[27], TCC_CYCLE[27], TCC_EA_ATOMIC[27], TCC_EA_ATOMIC_LEVEL[27], TCC_ATOMIC[28], TCC_CYCLE[28], TCC_EA_ATOMIC[28], TCC_EA_ATOMIC_LEVEL[28], TCC_ATOMIC[29], TCC_CYCLE[29], TCC_EA_ATOMIC[29], TCC_EA_ATOMIC_LEVEL[29], TCC_ATOMIC[30], TCC_CYCLE[30], TCC_EA_ATOMIC[30], TCC_EA_ATOMIC_LEVEL[30], TCC_ATOMIC[31], TCC_CYCLE[31], TCC_EA_ATOMIC[31], TCC_EA_ATOMIC_LEVEL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160258_4072642/input0_results_240321_160258
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_13.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_14.txt
|-> [rocprof] RPL: on '240321_160259' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_14.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160259_4072850'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160259_4072850/input0_results_240321_160259'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160259_4072850/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_RDREQ[0], TCC_EA_RDREQ_32B[0], TCC_EA_RDREQ_DRAM_CREDIT_STALL[0], TCC_EA_RDREQ_GMI_CREDIT_STALL[0], TCC_EA_RDREQ[1], TCC_EA_RDREQ_32B[1], TCC_EA_RDREQ_DRAM_CREDIT_STALL[1], TCC_EA_RDREQ_GMI_CREDIT_STALL[1], TCC_EA_RDREQ[2], TCC_EA_RDREQ_32B[2], TCC_EA_RDREQ_DRAM_CREDIT_STALL[2], TCC_EA_RDREQ_GMI_CREDIT_STALL[2], TCC_EA_RDREQ[3], TCC_EA_RDREQ_32B[3], TCC_EA_RDREQ_DRAM_CREDIT_STALL[3], TCC_EA_RDREQ_GMI_CREDIT_STALL[3], TCC_EA_RDREQ[4], TCC_EA_RDREQ_32B[4], TCC_EA_RDREQ_DRAM_CREDIT_STALL[4], TCC_EA_RDREQ_GMI_CREDIT_STALL[4], TCC_EA_RDREQ[5], TCC_EA_RDREQ_32B[5], TCC_EA_RDREQ_DRAM_CREDIT_STALL[5], TCC_EA_RDREQ_GMI_CREDIT_STALL[5], TCC_EA_RDREQ[6], TCC_EA_RDREQ_32B[6], TCC_EA_RDREQ_DRAM_CREDIT_STALL[6], TCC_EA_RDREQ_GMI_CREDIT_STALL[6], TCC_EA_RDREQ[7], TCC_EA_RDREQ_32B[7], TCC_EA_RDREQ_DRAM_CREDIT_STALL[7], TCC_EA_RDREQ_GMI_CREDIT_STALL[7], TCC_EA_RDREQ[8], TCC_EA_RDREQ_32B[8], TCC_EA_RDREQ_DRAM_CREDIT_STALL[8], TCC_EA_RDREQ_GMI_CREDIT_STALL[8], TCC_EA_RDREQ[9], TCC_EA_RDREQ_32B[9], TCC_EA_RDREQ_DRAM_CREDIT_STALL[9], TCC_EA_RDREQ_GMI_CREDIT_STALL[9], TCC_EA_RDREQ[10], TCC_EA_RDREQ_32B[10], TCC_EA_RDREQ_DRAM_CREDIT_STALL[10], TCC_EA_RDREQ_GMI_CREDIT_STALL[10], TCC_EA_RDREQ[11], TCC_EA_RDREQ_32B[11], TCC_EA_RDREQ_DRAM_CREDIT_STALL[11], TCC_EA_RDREQ_GMI_CREDIT_STALL[11], TCC_EA_RDREQ[12], TCC_EA_RDREQ_32B[12], TCC_EA_RDREQ_DRAM_CREDIT_STALL[12], TCC_EA_RDREQ_GMI_CREDIT_STALL[12], TCC_EA_RDREQ[13], TCC_EA_RDREQ_32B[13], TCC_EA_RDREQ_DRAM_CREDIT_STALL[13], TCC_EA_RDREQ_GMI_CREDIT_STALL[13], TCC_EA_RDREQ[14], TCC_EA_RDREQ_32B[14], TCC_EA_RDREQ_DRAM_CREDIT_STALL[14], TCC_EA_RDREQ_GMI_CREDIT_STALL[14], TCC_EA_RDREQ[15], TCC_EA_RDREQ_32B[15], TCC_EA_RDREQ_DRAM_CREDIT_STALL[15], TCC_EA_RDREQ_GMI_CREDIT_STALL[15], TCC_EA_RDREQ[16], TCC_EA_RDREQ_32B[16], TCC_EA_RDREQ_DRAM_CREDIT_STALL[16], TCC_EA_RDREQ_GMI_CREDIT_STALL[16], TCC_EA_RDREQ[17], TCC_EA_RDREQ_32B[17], TCC_EA_RDREQ_DRAM_CREDIT_STALL[17], TCC_EA_RDREQ_GMI_CREDIT_STALL[17], TCC_EA_RDREQ[18], TCC_EA_RDREQ_32B[18], TCC_EA_RDREQ_DRAM_CREDIT_STALL[18], TCC_EA_RDREQ_GMI_CREDIT_STALL[18], TCC_EA_RDREQ[19], TCC_EA_RDREQ_32B[19], TCC_EA_RDREQ_DRAM_CREDIT_STALL[19], TCC_EA_RDREQ_GMI_CREDIT_STALL[19], TCC_EA_RDREQ[20], TCC_EA_RDREQ_32B[20], TCC_EA_RDREQ_DRAM_CREDIT_STALL[20], TCC_EA_RDREQ_GMI_CREDIT_STALL[20], TCC_EA_RDREQ[21], TCC_EA_RDREQ_32B[21], TCC_EA_RDREQ_DRAM_CREDIT_STALL[21], TCC_EA_RDREQ_GMI_CREDIT_STALL[21], TCC_EA_RDREQ[22], TCC_EA_RDREQ_32B[22], TCC_EA_RDREQ_DRAM_CREDIT_STALL[22], TCC_EA_RDREQ_GMI_CREDIT_STALL[22], TCC_EA_RDREQ[23], TCC_EA_RDREQ_32B[23], TCC_EA_RDREQ_DRAM_CREDIT_STALL[23], TCC_EA_RDREQ_GMI_CREDIT_STALL[23], TCC_EA_RDREQ[24], TCC_EA_RDREQ_32B[24], TCC_EA_RDREQ_DRAM_CREDIT_STALL[24], TCC_EA_RDREQ_GMI_CREDIT_STALL[24], TCC_EA_RDREQ[25], TCC_EA_RDREQ_32B[25], TCC_EA_RDREQ_DRAM_CREDIT_STALL[25], TCC_EA_RDREQ_GMI_CREDIT_STALL[25], TCC_EA_RDREQ[26], TCC_EA_RDREQ_32B[26], TCC_EA_RDREQ_DRAM_CREDIT_STALL[26], TCC_EA_RDREQ_GMI_CREDIT_STALL[26], TCC_EA_RDREQ[27], TCC_EA_RDREQ_32B[27], TCC_EA_RDREQ_DRAM_CREDIT_STALL[27], TCC_EA_RDREQ_GMI_CREDIT_STALL[27], TCC_EA_RDREQ[28], TCC_EA_RDREQ_32B[28], TCC_EA_RDREQ_DRAM_CREDIT_STALL[28], TCC_EA_RDREQ_GMI_CREDIT_STALL[28], TCC_EA_RDREQ[29], TCC_EA_RDREQ_32B[29], TCC_EA_RDREQ_DRAM_CREDIT_STALL[29], TCC_EA_RDREQ_GMI_CREDIT_STALL[29], TCC_EA_RDREQ[30], TCC_EA_RDREQ_32B[30], TCC_EA_RDREQ_DRAM_CREDIT_STALL[30], TCC_EA_RDREQ_GMI_CREDIT_STALL[30], TCC_EA_RDREQ[31], TCC_EA_RDREQ_32B[31], TCC_EA_RDREQ_DRAM_CREDIT_STALL[31], TCC_EA_RDREQ_GMI_CREDIT_STALL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160259_4072850/input0_results_240321_160259
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_14.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_15.txt
|-> [rocprof] RPL: on '240321_160300' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_15.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160300_4073035'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160300_4073035/input0_results_240321_160300'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160300_4073035/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_RDREQ_IO_CREDIT_STALL[0], TCC_EA_RDREQ_LEVEL[0], TCC_EA_WRREQ[0], TCC_EA_WRREQ_64B[0], TCC_EA_RDREQ_IO_CREDIT_STALL[1], TCC_EA_RDREQ_LEVEL[1], TCC_EA_WRREQ[1], TCC_EA_WRREQ_64B[1], TCC_EA_RDREQ_IO_CREDIT_STALL[2], TCC_EA_RDREQ_LEVEL[2], TCC_EA_WRREQ[2], TCC_EA_WRREQ_64B[2], TCC_EA_RDREQ_IO_CREDIT_STALL[3], TCC_EA_RDREQ_LEVEL[3], TCC_EA_WRREQ[3], TCC_EA_WRREQ_64B[3], TCC_EA_RDREQ_IO_CREDIT_STALL[4], TCC_EA_RDREQ_LEVEL[4], TCC_EA_WRREQ[4], TCC_EA_WRREQ_64B[4], TCC_EA_RDREQ_IO_CREDIT_STALL[5], TCC_EA_RDREQ_LEVEL[5], TCC_EA_WRREQ[5], TCC_EA_WRREQ_64B[5], TCC_EA_RDREQ_IO_CREDIT_STALL[6], TCC_EA_RDREQ_LEVEL[6], TCC_EA_WRREQ[6], TCC_EA_WRREQ_64B[6], TCC_EA_RDREQ_IO_CREDIT_STALL[7], TCC_EA_RDREQ_LEVEL[7], TCC_EA_WRREQ[7], TCC_EA_WRREQ_64B[7], TCC_EA_RDREQ_IO_CREDIT_STALL[8], TCC_EA_RDREQ_LEVEL[8], TCC_EA_WRREQ[8], TCC_EA_WRREQ_64B[8], TCC_EA_RDREQ_IO_CREDIT_STALL[9], TCC_EA_RDREQ_LEVEL[9], TCC_EA_WRREQ[9], TCC_EA_WRREQ_64B[9], TCC_EA_RDREQ_IO_CREDIT_STALL[10], TCC_EA_RDREQ_LEVEL[10], TCC_EA_WRREQ[10], TCC_EA_WRREQ_64B[10], TCC_EA_RDREQ_IO_CREDIT_STALL[11], TCC_EA_RDREQ_LEVEL[11], TCC_EA_WRREQ[11], TCC_EA_WRREQ_64B[11], TCC_EA_RDREQ_IO_CREDIT_STALL[12], TCC_EA_RDREQ_LEVEL[12], TCC_EA_WRREQ[12], TCC_EA_WRREQ_64B[12], TCC_EA_RDREQ_IO_CREDIT_STALL[13], TCC_EA_RDREQ_LEVEL[13], TCC_EA_WRREQ[13], TCC_EA_WRREQ_64B[13], TCC_EA_RDREQ_IO_CREDIT_STALL[14], TCC_EA_RDREQ_LEVEL[14], TCC_EA_WRREQ[14], TCC_EA_WRREQ_64B[14], TCC_EA_RDREQ_IO_CREDIT_STALL[15], TCC_EA_RDREQ_LEVEL[15], TCC_EA_WRREQ[15], TCC_EA_WRREQ_64B[15], TCC_EA_RDREQ_IO_CREDIT_STALL[16], TCC_EA_RDREQ_LEVEL[16], TCC_EA_WRREQ[16], TCC_EA_WRREQ_64B[16], TCC_EA_RDREQ_IO_CREDIT_STALL[17], TCC_EA_RDREQ_LEVEL[17], TCC_EA_WRREQ[17], TCC_EA_WRREQ_64B[17], TCC_EA_RDREQ_IO_CREDIT_STALL[18], TCC_EA_RDREQ_LEVEL[18], TCC_EA_WRREQ[18], TCC_EA_WRREQ_64B[18], TCC_EA_RDREQ_IO_CREDIT_STALL[19], TCC_EA_RDREQ_LEVEL[19], TCC_EA_WRREQ[19], TCC_EA_WRREQ_64B[19], TCC_EA_RDREQ_IO_CREDIT_STALL[20], TCC_EA_RDREQ_LEVEL[20], TCC_EA_WRREQ[20], TCC_EA_WRREQ_64B[20], TCC_EA_RDREQ_IO_CREDIT_STALL[21], TCC_EA_RDREQ_LEVEL[21], TCC_EA_WRREQ[21], TCC_EA_WRREQ_64B[21], TCC_EA_RDREQ_IO_CREDIT_STALL[22], TCC_EA_RDREQ_LEVEL[22], TCC_EA_WRREQ[22], TCC_EA_WRREQ_64B[22], TCC_EA_RDREQ_IO_CREDIT_STALL[23], TCC_EA_RDREQ_LEVEL[23], TCC_EA_WRREQ[23], TCC_EA_WRREQ_64B[23], TCC_EA_RDREQ_IO_CREDIT_STALL[24], TCC_EA_RDREQ_LEVEL[24], TCC_EA_WRREQ[24], TCC_EA_WRREQ_64B[24], TCC_EA_RDREQ_IO_CREDIT_STALL[25], TCC_EA_RDREQ_LEVEL[25], TCC_EA_WRREQ[25], TCC_EA_WRREQ_64B[25], TCC_EA_RDREQ_IO_CREDIT_STALL[26], TCC_EA_RDREQ_LEVEL[26], TCC_EA_WRREQ[26], TCC_EA_WRREQ_64B[26], TCC_EA_RDREQ_IO_CREDIT_STALL[27], TCC_EA_RDREQ_LEVEL[27], TCC_EA_WRREQ[27], TCC_EA_WRREQ_64B[27], TCC_EA_RDREQ_IO_CREDIT_STALL[28], TCC_EA_RDREQ_LEVEL[28], TCC_EA_WRREQ[28], TCC_EA_WRREQ_64B[28], TCC_EA_RDREQ_IO_CREDIT_STALL[29], TCC_EA_RDREQ_LEVEL[29], TCC_EA_WRREQ[29], TCC_EA_WRREQ_64B[29], TCC_EA_RDREQ_IO_CREDIT_STALL[30], TCC_EA_RDREQ_LEVEL[30], TCC_EA_WRREQ[30], TCC_EA_WRREQ_64B[30], TCC_EA_RDREQ_IO_CREDIT_STALL[31], TCC_EA_RDREQ_LEVEL[31], TCC_EA_WRREQ[31], TCC_EA_WRREQ_64B[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160300_4073035/input0_results_240321_160300
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_15.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_16.txt
|-> [rocprof] RPL: on '240321_160301' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_16.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160301_4073239'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160301_4073239/input0_results_240321_160301'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160301_4073239/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_EA_WRREQ_DRAM_CREDIT_STALL[0], TCC_EA_WRREQ_GMI_CREDIT_STALL[0], TCC_EA_WRREQ_IO_CREDIT_STALL[0], TCC_EA_WRREQ_LEVEL[0], TCC_EA_WRREQ_DRAM_CREDIT_STALL[1], TCC_EA_WRREQ_GMI_CREDIT_STALL[1], TCC_EA_WRREQ_IO_CREDIT_STALL[1], TCC_EA_WRREQ_LEVEL[1], TCC_EA_WRREQ_DRAM_CREDIT_STALL[2], TCC_EA_WRREQ_GMI_CREDIT_STALL[2], TCC_EA_WRREQ_IO_CREDIT_STALL[2], TCC_EA_WRREQ_LEVEL[2], TCC_EA_WRREQ_DRAM_CREDIT_STALL[3], TCC_EA_WRREQ_GMI_CREDIT_STALL[3], TCC_EA_WRREQ_IO_CREDIT_STALL[3], TCC_EA_WRREQ_LEVEL[3], TCC_EA_WRREQ_DRAM_CREDIT_STALL[4], TCC_EA_WRREQ_GMI_CREDIT_STALL[4], TCC_EA_WRREQ_IO_CREDIT_STALL[4], TCC_EA_WRREQ_LEVEL[4], TCC_EA_WRREQ_DRAM_CREDIT_STALL[5], TCC_EA_WRREQ_GMI_CREDIT_STALL[5], TCC_EA_WRREQ_IO_CREDIT_STALL[5], TCC_EA_WRREQ_LEVEL[5], TCC_EA_WRREQ_DRAM_CREDIT_STALL[6], TCC_EA_WRREQ_GMI_CREDIT_STALL[6], TCC_EA_WRREQ_IO_CREDIT_STALL[6], TCC_EA_WRREQ_LEVEL[6], TCC_EA_WRREQ_DRAM_CREDIT_STALL[7], TCC_EA_WRREQ_GMI_CREDIT_STALL[7], TCC_EA_WRREQ_IO_CREDIT_STALL[7], TCC_EA_WRREQ_LEVEL[7], TCC_EA_WRREQ_DRAM_CREDIT_STALL[8], TCC_EA_WRREQ_GMI_CREDIT_STALL[8], TCC_EA_WRREQ_IO_CREDIT_STALL[8], TCC_EA_WRREQ_LEVEL[8], TCC_EA_WRREQ_DRAM_CREDIT_STALL[9], TCC_EA_WRREQ_GMI_CREDIT_STALL[9], TCC_EA_WRREQ_IO_CREDIT_STALL[9], TCC_EA_WRREQ_LEVEL[9], TCC_EA_WRREQ_DRAM_CREDIT_STALL[10], TCC_EA_WRREQ_GMI_CREDIT_STALL[10], TCC_EA_WRREQ_IO_CREDIT_STALL[10], TCC_EA_WRREQ_LEVEL[10], TCC_EA_WRREQ_DRAM_CREDIT_STALL[11], TCC_EA_WRREQ_GMI_CREDIT_STALL[11], TCC_EA_WRREQ_IO_CREDIT_STALL[11], TCC_EA_WRREQ_LEVEL[11], TCC_EA_WRREQ_DRAM_CREDIT_STALL[12], TCC_EA_WRREQ_GMI_CREDIT_STALL[12], TCC_EA_WRREQ_IO_CREDIT_STALL[12], TCC_EA_WRREQ_LEVEL[12], TCC_EA_WRREQ_DRAM_CREDIT_STALL[13], TCC_EA_WRREQ_GMI_CREDIT_STALL[13], TCC_EA_WRREQ_IO_CREDIT_STALL[13], TCC_EA_WRREQ_LEVEL[13], TCC_EA_WRREQ_DRAM_CREDIT_STALL[14], TCC_EA_WRREQ_GMI_CREDIT_STALL[14], TCC_EA_WRREQ_IO_CREDIT_STALL[14], TCC_EA_WRREQ_LEVEL[14], TCC_EA_WRREQ_DRAM_CREDIT_STALL[15], TCC_EA_WRREQ_GMI_CREDIT_STALL[15], TCC_EA_WRREQ_IO_CREDIT_STALL[15], TCC_EA_WRREQ_LEVEL[15], TCC_EA_WRREQ_DRAM_CREDIT_STALL[16], TCC_EA_WRREQ_GMI_CREDIT_STALL[16], TCC_EA_WRREQ_IO_CREDIT_STALL[16], TCC_EA_WRREQ_LEVEL[16], TCC_EA_WRREQ_DRAM_CREDIT_STALL[17], TCC_EA_WRREQ_GMI_CREDIT_STALL[17], TCC_EA_WRREQ_IO_CREDIT_STALL[17], TCC_EA_WRREQ_LEVEL[17], TCC_EA_WRREQ_DRAM_CREDIT_STALL[18], TCC_EA_WRREQ_GMI_CREDIT_STALL[18], TCC_EA_WRREQ_IO_CREDIT_STALL[18], TCC_EA_WRREQ_LEVEL[18], TCC_EA_WRREQ_DRAM_CREDIT_STALL[19], TCC_EA_WRREQ_GMI_CREDIT_STALL[19], TCC_EA_WRREQ_IO_CREDIT_STALL[19], TCC_EA_WRREQ_LEVEL[19], TCC_EA_WRREQ_DRAM_CREDIT_STALL[20], TCC_EA_WRREQ_GMI_CREDIT_STALL[20], TCC_EA_WRREQ_IO_CREDIT_STALL[20], TCC_EA_WRREQ_LEVEL[20], TCC_EA_WRREQ_DRAM_CREDIT_STALL[21], TCC_EA_WRREQ_GMI_CREDIT_STALL[21], TCC_EA_WRREQ_IO_CREDIT_STALL[21], TCC_EA_WRREQ_LEVEL[21], TCC_EA_WRREQ_DRAM_CREDIT_STALL[22], TCC_EA_WRREQ_GMI_CREDIT_STALL[22], TCC_EA_WRREQ_IO_CREDIT_STALL[22], TCC_EA_WRREQ_LEVEL[22], TCC_EA_WRREQ_DRAM_CREDIT_STALL[23], TCC_EA_WRREQ_GMI_CREDIT_STALL[23], TCC_EA_WRREQ_IO_CREDIT_STALL[23], TCC_EA_WRREQ_LEVEL[23], TCC_EA_WRREQ_DRAM_CREDIT_STALL[24], TCC_EA_WRREQ_GMI_CREDIT_STALL[24], TCC_EA_WRREQ_IO_CREDIT_STALL[24], TCC_EA_WRREQ_LEVEL[24], TCC_EA_WRREQ_DRAM_CREDIT_STALL[25], TCC_EA_WRREQ_GMI_CREDIT_STALL[25], TCC_EA_WRREQ_IO_CREDIT_STALL[25], TCC_EA_WRREQ_LEVEL[25], TCC_EA_WRREQ_DRAM_CREDIT_STALL[26], TCC_EA_WRREQ_GMI_CREDIT_STALL[26], TCC_EA_WRREQ_IO_CREDIT_STALL[26], TCC_EA_WRREQ_LEVEL[26], TCC_EA_WRREQ_DRAM_CREDIT_STALL[27], TCC_EA_WRREQ_GMI_CREDIT_STALL[27], TCC_EA_WRREQ_IO_CREDIT_STALL[27], TCC_EA_WRREQ_LEVEL[27], TCC_EA_WRREQ_DRAM_CREDIT_STALL[28], TCC_EA_WRREQ_GMI_CREDIT_STALL[28], TCC_EA_WRREQ_IO_CREDIT_STALL[28], TCC_EA_WRREQ_LEVEL[28], TCC_EA_WRREQ_DRAM_CREDIT_STALL[29], TCC_EA_WRREQ_GMI_CREDIT_STALL[29], TCC_EA_WRREQ_IO_CREDIT_STALL[29], TCC_EA_WRREQ_LEVEL[29], TCC_EA_WRREQ_DRAM_CREDIT_STALL[30], TCC_EA_WRREQ_GMI_CREDIT_STALL[30], TCC_EA_WRREQ_IO_CREDIT_STALL[30], TCC_EA_WRREQ_LEVEL[30], TCC_EA_WRREQ_DRAM_CREDIT_STALL[31], TCC_EA_WRREQ_GMI_CREDIT_STALL[31], TCC_EA_WRREQ_IO_CREDIT_STALL[31], TCC_EA_WRREQ_LEVEL[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160301_4073239/input0_results_240321_160301
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_16.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_17.txt
|-> [rocprof] RPL: on '240321_160301' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_17.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160301_4073460'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160301_4073460/input0_results_240321_160301'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160301_4073460/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 128 metrics
|-> [rocprof] TCC_HIT[0], TCC_MISS[0], TCC_READ[0], TCC_REQ[0], TCC_HIT[1], TCC_MISS[1], TCC_READ[1], TCC_REQ[1], TCC_HIT[2], TCC_MISS[2], TCC_READ[2], TCC_REQ[2], TCC_HIT[3], TCC_MISS[3], TCC_READ[3], TCC_REQ[3], TCC_HIT[4], TCC_MISS[4], TCC_READ[4], TCC_REQ[4], TCC_HIT[5], TCC_MISS[5], TCC_READ[5], TCC_REQ[5], TCC_HIT[6], TCC_MISS[6], TCC_READ[6], TCC_REQ[6], TCC_HIT[7], TCC_MISS[7], TCC_READ[7], TCC_REQ[7], TCC_HIT[8], TCC_MISS[8], TCC_READ[8], TCC_REQ[8], TCC_HIT[9], TCC_MISS[9], TCC_READ[9], TCC_REQ[9], TCC_HIT[10], TCC_MISS[10], TCC_READ[10], TCC_REQ[10], TCC_HIT[11], TCC_MISS[11], TCC_READ[11], TCC_REQ[11], TCC_HIT[12], TCC_MISS[12], TCC_READ[12], TCC_REQ[12], TCC_HIT[13], TCC_MISS[13], TCC_READ[13], TCC_REQ[13], TCC_HIT[14], TCC_MISS[14], TCC_READ[14], TCC_REQ[14], TCC_HIT[15], TCC_MISS[15], TCC_READ[15], TCC_REQ[15], TCC_HIT[16], TCC_MISS[16], TCC_READ[16], TCC_REQ[16], TCC_HIT[17], TCC_MISS[17], TCC_READ[17], TCC_REQ[17], TCC_HIT[18], TCC_MISS[18], TCC_READ[18], TCC_REQ[18], TCC_HIT[19], TCC_MISS[19], TCC_READ[19], TCC_REQ[19], TCC_HIT[20], TCC_MISS[20], TCC_READ[20], TCC_REQ[20], TCC_HIT[21], TCC_MISS[21], TCC_READ[21], TCC_REQ[21], TCC_HIT[22], TCC_MISS[22], TCC_READ[22], TCC_REQ[22], TCC_HIT[23], TCC_MISS[23], TCC_READ[23], TCC_REQ[23], TCC_HIT[24], TCC_MISS[24], TCC_READ[24], TCC_REQ[24], TCC_HIT[25], TCC_MISS[25], TCC_READ[25], TCC_REQ[25], TCC_HIT[26], TCC_MISS[26], TCC_READ[26], TCC_REQ[26], TCC_HIT[27], TCC_MISS[27], TCC_READ[27], TCC_REQ[27], TCC_HIT[28], TCC_MISS[28], TCC_READ[28], TCC_REQ[28], TCC_HIT[29], TCC_MISS[29], TCC_READ[29], TCC_REQ[29], TCC_HIT[30], TCC_MISS[30], TCC_READ[30], TCC_REQ[30], TCC_HIT[31], TCC_MISS[31], TCC_READ[31], TCC_REQ[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160301_4073460/input0_results_240321_160301
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_17.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_18.txt
|-> [rocprof] RPL: on '240321_160302' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_18.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160302_4073655'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160302_4073655/input0_results_240321_160302'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160302_4073655/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 96 metrics
|-> [rocprof] TCC_RW_REQ[0], TCC_TOO_MANY_EA_WRREQS_STALL[0], TCC_WRITE[0], TCC_RW_REQ[1], TCC_TOO_MANY_EA_WRREQS_STALL[1], TCC_WRITE[1], TCC_RW_REQ[2], TCC_TOO_MANY_EA_WRREQS_STALL[2], TCC_WRITE[2], TCC_RW_REQ[3], TCC_TOO_MANY_EA_WRREQS_STALL[3], TCC_WRITE[3], TCC_RW_REQ[4], TCC_TOO_MANY_EA_WRREQS_STALL[4], TCC_WRITE[4], TCC_RW_REQ[5], TCC_TOO_MANY_EA_WRREQS_STALL[5], TCC_WRITE[5], TCC_RW_REQ[6], TCC_TOO_MANY_EA_WRREQS_STALL[6], TCC_WRITE[6], TCC_RW_REQ[7], TCC_TOO_MANY_EA_WRREQS_STALL[7], TCC_WRITE[7], TCC_RW_REQ[8], TCC_TOO_MANY_EA_WRREQS_STALL[8], TCC_WRITE[8], TCC_RW_REQ[9], TCC_TOO_MANY_EA_WRREQS_STALL[9], TCC_WRITE[9], TCC_RW_REQ[10], TCC_TOO_MANY_EA_WRREQS_STALL[10], TCC_WRITE[10], TCC_RW_REQ[11], TCC_TOO_MANY_EA_WRREQS_STALL[11], TCC_WRITE[11], TCC_RW_REQ[12], TCC_TOO_MANY_EA_WRREQS_STALL[12], TCC_WRITE[12], TCC_RW_REQ[13], TCC_TOO_MANY_EA_WRREQS_STALL[13], TCC_WRITE[13], TCC_RW_REQ[14], TCC_TOO_MANY_EA_WRREQS_STALL[14], TCC_WRITE[14], TCC_RW_REQ[15], TCC_TOO_MANY_EA_WRREQS_STALL[15], TCC_WRITE[15], TCC_RW_REQ[16], TCC_TOO_MANY_EA_WRREQS_STALL[16], TCC_WRITE[16], TCC_RW_REQ[17], TCC_TOO_MANY_EA_WRREQS_STALL[17], TCC_WRITE[17], TCC_RW_REQ[18], TCC_TOO_MANY_EA_WRREQS_STALL[18], TCC_WRITE[18], TCC_RW_REQ[19], TCC_TOO_MANY_EA_WRREQS_STALL[19], TCC_WRITE[19], TCC_RW_REQ[20], TCC_TOO_MANY_EA_WRREQS_STALL[20], TCC_WRITE[20], TCC_RW_REQ[21], TCC_TOO_MANY_EA_WRREQS_STALL[21], TCC_WRITE[21], TCC_RW_REQ[22], TCC_TOO_MANY_EA_WRREQS_STALL[22], TCC_WRITE[22], TCC_RW_REQ[23], TCC_TOO_MANY_EA_WRREQS_STALL[23], TCC_WRITE[23], TCC_RW_REQ[24], TCC_TOO_MANY_EA_WRREQS_STALL[24], TCC_WRITE[24], TCC_RW_REQ[25], TCC_TOO_MANY_EA_WRREQS_STALL[25], TCC_WRITE[25], TCC_RW_REQ[26], TCC_TOO_MANY_EA_WRREQS_STALL[26], TCC_WRITE[26], TCC_RW_REQ[27], TCC_TOO_MANY_EA_WRREQS_STALL[27], TCC_WRITE[27], TCC_RW_REQ[28], TCC_TOO_MANY_EA_WRREQS_STALL[28], TCC_WRITE[28], TCC_RW_REQ[29], TCC_TOO_MANY_EA_WRREQS_STALL[29], TCC_WRITE[29], TCC_RW_REQ[30], TCC_TOO_MANY_EA_WRREQS_STALL[30], TCC_WRITE[30], TCC_RW_REQ[31], TCC_TOO_MANY_EA_WRREQS_STALL[31], TCC_WRITE[31]
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160302_4073655/input0_results_240321_160302
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_18.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_2.txt
|-> [rocprof] RPL: on '240321_160303' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_2.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160303_4073839'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160303_4073839/input0_results_240321_160303'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160303_4073839/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 26 metrics
|-> [rocprof] SQ_INSTS_VALU_MUL_F32, SQ_INSTS_VALU_FMA_F32, SQ_INSTS_VALU_TRANS_F32, SQ_INSTS_VALU_ADD_F64, SQ_INSTS_VALU_MUL_F64, SQ_INSTS_VALU_FMA_F64, SQ_INSTS_VALU_TRANS_F64, SQ_INSTS_VALU_INT32, TCP_VOLATILE_sum, TCP_TOTAL_ACCESSES_sum, TCP_TOTAL_READ_sum, TCP_TOTAL_WRITE_sum, TA_BUFFER_ATOMIC_WAVEFRONTS_sum, TA_BUFFER_TOTAL_CYCLES_sum, TD_ATOMIC_WAVEFRONT_sum, TD_STORE_WAVEFRONT_sum, SPI_RA_REQ_NO_ALLOC, SPI_RA_REQ_NO_ALLOC_CSN, CPC_CPC_STAT_STALL, CPC_UTCL1_STALL_ON_TRANSLATION, CPF_CPF_STAT_IDLE, CPF_CPF_TCIU_IDLE, TCC_REQ_sum, TCC_STREAMING_REQ_sum, TCC_HIT_sum, TCC_MISS_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160303_4073839/input0_results_240321_160303
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_2.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_3.txt
|-> [rocprof] RPL: on '240321_160303' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_3.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160303_4074032'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160303_4074032/input0_results_240321_160303'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160303_4074032/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 24 metrics
|-> [rocprof] SQ_INSTS_VALU_INT64, SQ_INSTS_SMEM, SQ_INSTS_FLAT, SQ_INSTS_LDS, SQ_INSTS_GDS, SQ_INSTS_EXP_GDS, SQ_INSTS_BRANCH, SQ_INSTS_SENDMSG, TCP_TOTAL_ATOMIC_WITH_RET_sum, TCP_TOTAL_ATOMIC_WITHOUT_RET_sum, TCP_TOTAL_WRITEBACK_INVALIDATES_sum, TCP_TOTAL_CACHE_ACCESSES_sum, TA_BUFFER_COALESCED_READ_CYCLES_sum, TA_BUFFER_COALESCED_WRITE_CYCLES_sum, TD_COALESCABLE_WAVEFRONT_sum, SPI_RA_RES_STALL_CSN, SPI_RA_TMP_STALL_CSN, CPC_CPC_UTCL2IU_BUSY, CPC_CPC_UTCL2IU_IDLE, CPF_CMP_UTCL1_STALL_ON_TRANSLATION, TCC_READ_sum, TCC_WRITE_sum, TCC_ATOMIC_sum, TCC_WRITEBACK_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160303_4074032/input0_results_240321_160303
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_3.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_4.txt
|-> [rocprof] RPL: on '240321_160304' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_4.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160304_4074228'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160304_4074228/input0_results_240321_160304'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160304_4074228/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 22 metrics
|-> [rocprof] SQ_WAVE_CYCLES, SQ_WAIT_ANY, SQ_WAIT_INST_ANY, SQ_ACTIVE_INST_ANY, SQ_BUSY_CU_CYCLES, SQ_ACTIVE_INST_VMEM, SQ_ACTIVE_INST_LDS, SQ_ACTIVE_INST_VALU, TCP_UTCL1_TRANSLATION_MISS_sum, TCP_UTCL1_TRANSLATION_HIT_sum, TCP_UTCL1_PERMISSION_MISS_sum, TCP_UTCL1_REQUEST_sum, TA_ADDR_STALLED_BY_TC_CYCLES_sum, TA_TOTAL_WAVEFRONTS_sum, SPI_RA_WAVE_SIMD_FULL_CSN, SPI_RA_VGPR_SIMD_FULL_CSN, CPC_CPC_UTCL2IU_STALL, CPC_ME1_BUSY_FOR_PACKET_DECODE, TCC_EA_WRREQ_sum, TCC_EA_WRREQ_64B_sum, TCC_EA_WR_UNCACHED_32B_sum, TCC_EA_WRREQ_DRAM_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160304_4074228/input0_results_240321_160304
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_4.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_5.txt
|-> [rocprof] RPL: on '240321_160304' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_5.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160304_4074428'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160304_4074428/input0_results_240321_160304'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160304_4074428/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 21 metrics
|-> [rocprof] SQ_ACTIVE_INST_SCA, SQ_ACTIVE_INST_EXP_GDS, SQ_ACTIVE_INST_MISC, SQ_ACTIVE_INST_FLAT, SQ_INST_CYCLES_VMEM_WR, SQ_INST_CYCLES_VMEM_RD, SQ_INST_CYCLES_SMEM, SQ_INST_CYCLES_SALU, TCP_TCP_LATENCY_sum, TCP_TCC_READ_REQ_LATENCY_sum, TCP_TCC_WRITE_REQ_LATENCY_sum, TCP_TCC_READ_REQ_sum, TA_ADDR_STALLED_BY_TD_CYCLES_sum, TA_DATA_STALLED_BY_TC_CYCLES_sum, SPI_RA_SGPR_SIMD_FULL_CSN, SPI_RA_LDS_CU_FULL_CSN, CPC_ME1_DC0_SPI_BUSY, TCC_EA_WRREQ_STALL_sum, TCC_EA_RDREQ_sum, TCC_EA_RDREQ_32B_sum, TCC_EA_RD_UNCACHED_32B_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160304_4074428/input0_results_240321_160304
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_5.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_6.txt
|-> [rocprof] RPL: on '240321_160305' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_6.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160305_4074630'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160305_4074630/input0_results_240321_160305'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160305_4074630/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 20 metrics
|-> [rocprof] SQ_THREAD_CYCLES_VALU, SQ_IFETCH, SQ_LDS_BANK_CONFLICT, SQ_LDS_ADDR_CONFLICT, SQ_LDS_UNALIGNED_STALL, SQ_WAVES_EQ_64, SQ_WAVES_LT_64, SQ_WAVES_LT_48, TCP_TCC_WRITE_REQ_sum, TCP_TCC_ATOMIC_WITH_RET_REQ_sum, TCP_TCC_ATOMIC_WITHOUT_RET_REQ_sum, TCP_TCC_NC_READ_REQ_sum, TA_FLAT_WAVEFRONTS_sum, TA_FLAT_READ_WAVEFRONTS_sum, SPI_RA_BAR_CU_FULL_CSN, SPI_RA_TGLIM_CU_FULL_CSN, TCC_EA_RDREQ_DRAM_sum, TCC_TAG_STALL_sum, TCC_NORMAL_WRITEBACK_sum, TCC_ALL_TC_OP_WB_WRITEBACK_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160305_4074630/input0_results_240321_160305
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_6.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_7.txt
|-> [rocprof] RPL: on '240321_160305' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_7.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160305_4074832'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160305_4074832/input0_results_240321_160305'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160305_4074832/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 20 metrics
|-> [rocprof] SQ_WAVES_LT_32, SQ_WAVES_LT_16, SQ_ITEMS, SQ_LDS_MEM_VIOLATIONS, SQ_LDS_ATOMIC_RETURN, SQ_LDS_IDX_ACTIVE, SQ_WAVES_RESTORED, SQ_WAVES_SAVED, TCP_TCC_NC_WRITE_REQ_sum, TCP_TCC_NC_ATOMIC_REQ_sum, TCP_TCC_UC_READ_REQ_sum, TCP_TCC_UC_WRITE_REQ_sum, TA_FLAT_WRITE_WAVEFRONTS_sum, TA_FLAT_ATOMIC_WAVEFRONTS_sum, SPI_RA_WVLIM_STALL_CSN, SPI_SWC_CSC_WR, TCC_NORMAL_EVICT_sum, TCC_ALL_TC_OP_INV_EVICT_sum, TCC_TOO_MANY_EA_WRREQS_STALL_sum, TCC_EA_ATOMIC_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160305_4074832/input0_results_240321_160305
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_7.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_8.txt
|-> [rocprof] RPL: on '240321_160305' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_8.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160305_4075016'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160305_4075016/input0_results_240321_160305'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160305_4075016/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 17 metrics
|-> [rocprof] SQ_INSTS_SMEM_NORM, SQ_INSTS_MFMA, SQ_INSTS_VALU_MFMA_I8, SQ_INSTS_VALU_MFMA_F16, SQ_INSTS_VALU_MFMA_BF16, SQ_INSTS_VALU_MFMA_F32, SQ_INSTS_VALU_MFMA_F64, SQ_VALU_MFMA_BUSY_CYCLES, TCP_TCC_UC_ATOMIC_REQ_sum, TCP_TCC_CC_READ_REQ_sum, TCP_TCC_CC_WRITE_REQ_sum, TCP_TCC_CC_ATOMIC_REQ_sum, SPI_VWC_CSC_WR, SPI_RA_BULKY_CU_FULL_CSN, TCC_EA_RDREQ_LEVEL_sum, TCC_EA_WRREQ_LEVEL_sum, TCC_EA_ATOMIC_LEVEL_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160305_4075016/input0_results_240321_160305
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_8.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_9.txt
|-> [rocprof] RPL: on '240321_160306' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/pmc_perf_9.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160306_4075219'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160306_4075219/input0_results_240321_160306'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160306_4075219/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 12 metrics
|-> [rocprof] SQ_INSTS_FLAT_LDS_ONLY, SQ_INSTS_VALU_MFMA_MOPS_I8, SQ_INSTS_VALU_MFMA_MOPS_F16, SQ_INSTS_VALU_MFMA_MOPS_BF16, SQ_INSTS_VALU_MFMA_MOPS_F32, SQ_INSTS_VALU_MFMA_MOPS_F64, SQC_TC_INST_REQ, SQC_TC_DATA_READ_REQ, TCP_TCC_RW_READ_REQ_sum, TCP_TCC_RW_WRITE_REQ_sum, TCP_TCC_RW_ATOMIC_REQ_sum, TCP_PENDING_STALL_CYCLES_sum
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160306_4075219/input0_results_240321_160306
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/pmc_perf_9.csv' is generating
|-> [rocprof]
[profiling] Current input file: tests/workloads/dispatch_inv/MI200/perfmon/timestamps.txt
|-> [rocprof] RPL: on '240321_160306' from '/opt/rocm-6.0.2' in '/home1/josantos/omniperf'
|-> [rocprof] RPL: profiling '""./tests/vcopy -n 1048576 -b 256 -i 3""'
|-> [rocprof] RPL: input file 'tests/workloads/dispatch_inv/MI200/perfmon/timestamps.txt'
|-> [rocprof] RPL: output dir '/tmp/rpl_data_240321_160306_4075410'
|-> [rocprof] RPL: result dir '/tmp/rpl_data_240321_160306_4075410/input0_results_240321_160306'
|-> [rocprof] ROCProfiler: input from "/tmp/rpl_data_240321_160306_4075410/input0.xml"
|-> [rocprof] gpu_index =
|-> [rocprof] kernel =
|-> [rocprof] range = invalid
|-> [rocprof] 0 metrics
|-> [rocprof] vcopy testing on GCD 0
|-> [rocprof] Finished allocating vectors on the CPU
|-> [rocprof] Finished allocating vectors on the GPU
|-> [rocprof] Finished copying vectors to the GPU
|-> [rocprof] sw thinks it moved 1.000000 KB per wave
|-> [rocprof] Total threads: 1048576, Grid Size: 4096 block Size:256, Wavefronts:16384:
|-> [rocprof] Launching the kernel on the GPU
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished executing kernel
|-> [rocprof] Finished copying the output vector from the GPU to the CPU
|-> [rocprof] Releasing GPU memory
|-> [rocprof] Releasing CPU memory
|-> [rocprof]
|-> [rocprof] ROCPRofiler: 1 contexts collected, output directory /tmp/rpl_data_240321_160306_4075410/input0_results_240321_160306
|-> [rocprof] File 'tests/workloads/dispatch_inv/MI200/timestamps.csv' is generating
|-> [rocprof]
[roofline] Checking for roofline.csv in tests/workloads/dispatch_inv/MI200
[roofline] No roofline data found. Generating...
tests/workloads/dispatch_inv/MI200/pmc_dispatch_info.csv
+2
View File
@@ -0,0 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2
1 Dispatch_ID Kernel_Name GPU_ID
2 0 vecCopy(double*, double*, double*, int, int) [clone .kd] 2
tests/workloads/dispatch_inv/MI200/pmc_perf.csv
+2 -2
View File
File diff suppressed because one or more lines are too long
tests/workloads/dispatch_inv/MI200/roofline.csv
+4 -4
View File
@@ -1,5 +1,5 @@
device,HBMBw,HBMBwLow,hbmBwHigh,L2Bw,L2BwLow,L2BwHigh,L1Bw,L1BwLow,L1BwHigh,LDSBw,LDSBwLow,LDSBwHigh,FP32Flops,FP32FlopsLow,FP32FlopsHigh,FP64Flops,FP64FlopsLow,FP64FlopsHigh,MFMABF16Flops,MFMABF16FlopsLow,MFMABF16FlopsHigh,MFMAF16Flops,MFMAF16FlopsLow,MFMAF16FlopsHigh,MFMAF32Flops,MFMAF32FlopsLow,MFMAF32FlopsHigh,MFMAF64Flops,MFMAF64FlopsLow,MFMAF64FlopsHigh,MFMAI8Ops,MFMAFI8OpsLow,MFMAI8OpsHigh
0,1388.5352,1387.9612,1389.1091,5015.9067,5012.9795,5018.834,9222.8076,9222.2178,9223.3975,18052.672,18051.719,18053.625,20935.363,20875.119,20995.607,20266.707,20265.77,20267.645,170447.69,170443.23,170452.14,164876.22,164871.3,164881.14,41420.414,41419.844,41420.984,41477.77,41477.254,41478.285,166338.45,165842.38,166834.53
1,1389.2339,1388.7003,1389.7675,5027.8267,5025.6963,5029.957,9235.0215,9234.4062,9235.6367,18093.281,18089.494,18097.068,20974.221,20906.723,21041.719,20295.463,20294.809,20296.117,170853.08,170848.84,170857.31,165136.75,165133,165140.5,41495.602,41494.711,41496.492,41552.312,41551.637,41552.988,166910.98,166908.33,166913.64
2,1388.1539,1387.5933,1388.7146,5035.959,5033.6694,5038.2485,9263.0859,9262.3916,9263.7803,18110.701,18109.838,18111.564,21052.15,21051.787,21052.514,20358.631,20358.045,20359.217,171151.48,171147.84,171155.12,165561.08,165557.56,165564.59,41593.766,41592.574,41594.957,41652.289,41651.02,41653.559,167010.91,166416.05,167605.77
3,1388.6646,1388.1143,1389.2148,5031.4497,5030.1416,5032.7578,9240.3379,9239.8271,9240.8486,18023.123,18019.518,18026.729,21031.461,21031.051,21031.871,20311.738,20311.039,20312.438,171003.95,171001.52,171006.39,165217.39,165214.22,165220.56,41438.289,41279.395,41597.184,41580.961,41580.465,41581.457,167015.88,167013.78,167017.97
0,1388.9141,1388.3721,1389.4561,5017.4395,5014.228,5020.6509,9226.0234,9225.335,9226.7119,17713.227,17710.121,17716.332,20948.057,20887.705,21008.408,20275.275,20274.727,20275.824,170525.36,170521.61,170529.11,164925.19,164920.83,164929.55,41434.961,41434.453,41435.469,41492.777,41492.191,41493.363,166378.45,165887.44,166869.47
1,1388.9977,1388.3805,1389.6149,5029.1138,5027.1401,5031.0874,9236.625,9235.9473,9237.3027,18235.252,18234.104,18236.4,20982.668,20915.91,21049.426,20298.568,20298.055,20299.082,170906.36,170901.52,170911.2,165152.56,165148.75,165156.38,41497.047,41496.234,41497.859,41549.789,41548.727,41550.852,166915.34,166911.5,166919.19
2,1388.4413,1387.9117,1388.9708,5036.0142,5033.8271,5038.2012,9260.9375,9260.2217,9261.6533,18218.748,18217.816,18219.68,21050.432,21049.973,21050.891,20353.637,20353.145,20354.129,171133.27,171129.09,171137.44,165528.48,165525.61,165531.36,41586.836,41585.996,41587.676,41640.574,41639.734,41641.414,166959.62,166363.45,167555.8
3,1389.6438,1388.9714,1390.3162,5031.4707,5029.8926,5033.0488,9237.8076,9237.0107,9238.6045,18328.922,18328.092,18329.752,21031.645,21031.092,21032.197,20307.449,20306.998,20307.9,171022.23,171016.14,171028.33,165129.09,165124.02,165134.17,41416.918,41262.441,41571.395,41555.602,41554.879,41556.324,166912.22,166909.77,166914.67
1 device HBMBw HBMBwLow hbmBwHigh L2Bw L2BwLow L2BwHigh L1Bw L1BwLow L1BwHigh LDSBw LDSBwLow LDSBwHigh FP32Flops FP32FlopsLow FP32FlopsHigh FP64Flops FP64FlopsLow FP64FlopsHigh MFMABF16Flops MFMABF16FlopsLow MFMABF16FlopsHigh MFMAF16Flops MFMAF16FlopsLow MFMAF16FlopsHigh MFMAF32Flops MFMAF32FlopsLow MFMAF32FlopsHigh MFMAF64Flops MFMAF64FlopsLow MFMAF64FlopsHigh MFMAI8Ops MFMAFI8OpsLow MFMAI8OpsHigh
2 0 1388.5352 1388.9141 1387.9612 1388.3721 1389.1091 1389.4561 5015.9067 5017.4395 5012.9795 5014.228 5018.834 5020.6509 9222.8076 9226.0234 9222.2178 9225.335 9223.3975 9226.7119 18052.672 17713.227 18051.719 17710.121 18053.625 17716.332 20935.363 20948.057 20875.119 20887.705 20995.607 21008.408 20266.707 20275.275 20265.77 20274.727 20267.645 20275.824 170447.69 170525.36 170443.23 170521.61 170452.14 170529.11 164876.22 164925.19 164871.3 164920.83 164881.14 164929.55 41420.414 41434.961 41419.844 41434.453 41420.984 41435.469 41477.77 41492.777 41477.254 41492.191 41478.285 41493.363 166338.45 166378.45 165842.38 165887.44 166834.53 166869.47
3 1 1389.2339 1388.9977 1388.7003 1388.3805 1389.7675 1389.6149 5027.8267 5029.1138 5025.6963 5027.1401 5029.957 5031.0874 9235.0215 9236.625 9234.4062 9235.9473 9235.6367 9237.3027 18093.281 18235.252 18089.494 18234.104 18097.068 18236.4 20974.221 20982.668 20906.723 20915.91 21041.719 21049.426 20295.463 20298.568 20294.809 20298.055 20296.117 20299.082 170853.08 170906.36 170848.84 170901.52 170857.31 170911.2 165136.75 165152.56 165133 165148.75 165140.5 165156.38 41495.602 41497.047 41494.711 41496.234 41496.492 41497.859 41552.312 41549.789 41551.637 41548.727 41552.988 41550.852 166910.98 166915.34 166908.33 166911.5 166913.64 166919.19
4 2 1388.1539 1388.4413 1387.5933 1387.9117 1388.7146 1388.9708 5035.959 5036.0142 5033.6694 5033.8271 5038.2485 5038.2012 9263.0859 9260.9375 9262.3916 9260.2217 9263.7803 9261.6533 18110.701 18218.748 18109.838 18217.816 18111.564 18219.68 21052.15 21050.432 21051.787 21049.973 21052.514 21050.891 20358.631 20353.637 20358.045 20353.145 20359.217 20354.129 171151.48 171133.27 171147.84 171129.09 171155.12 171137.44 165561.08 165528.48 165557.56 165525.61 165564.59 165531.36 41593.766 41586.836 41592.574 41585.996 41594.957 41587.676 41652.289 41640.574 41651.02 41639.734 41653.559 41641.414 167010.91 166959.62 166416.05 166363.45 167605.77 167555.8
5 3 1388.6646 1389.6438 1388.1143 1388.9714 1389.2148 1390.3162 5031.4497 5031.4707 5030.1416 5029.8926 5032.7578 5033.0488 9240.3379 9237.8076 9239.8271 9237.0107 9240.8486 9238.6045 18023.123 18328.922 18019.518 18328.092 18026.729 18329.752 21031.461 21031.645 21031.051 21031.092 21031.871 21032.197 20311.738 20307.449 20311.039 20306.998 20312.438 20307.9 171003.95 171022.23 171001.52 171016.14 171006.39 171028.33 165217.39 165129.09 165214.22 165124.02 165220.56 165134.17 41438.289 41416.918 41279.395 41262.441 41597.184 41571.395 41580.961 41555.602 41580.465 41554.879 41581.457 41556.324 167015.88 166912.22 167013.78 166909.77 167017.97 166914.67
tests/workloads/dispatch_inv/MI200/sysinfo.csv
+1 -1
View File
@@ -1,2 +1,2 @@
workload_name,command,ip_blocks,timestamp,version,hostname,cpu_model,sbios,linux_distro,linux_kernel_version,amd_gpu_kernel_version,cpu_memory,gpu_memory,rocm_version,vbios,compute_partition,memory_partition,gpu_model,gpu_arch,gpu_l1,gpu_l2,cu_per_gpu,simd_per_cu,se_per_gpu,wave_size,workgroup_max_size,max_waves_per_cu,max_sclk,max_mclk,cur_sclk,cur_mclk,total_l2_chan,lds_banks_per_cu,sqc_per_gpu,pipes_per_gpu,hbm_bw,num_xcd
dispatch_inv,./sample/vcopy -n 1048576 -b 256 -i 3,SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF|roofline,Thu 07 Mar 2024 01:59:41 PM (CST),2,t007-002.hpcfund,AMD EPYC 7V13 64-Core Processor,American Megatrends Inc.0602,Rocky Linux 9.1 (Blue Onyx),5.14.0-162.18.1.el9_1.x86_64,,527650760,,6.0.2-115,113-D67301-059,NA,NA,MI200,gfx90a,16,8192,104,4,8,64,1024,32,1700,1600,1700,1600,32,32,56,4,1638.4,1
dispatch_inv,./tests/vcopy -n 1048576 -b 256 -i 3,SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF|roofline,Thu 21 Mar 2024 04:02:53 PM (CDT),2,t007-002.hpcfund,AMD EPYC 7V13 64-Core Processor,American Megatrends Inc.0602,Rocky Linux 9.1 (Blue Onyx),5.14.0-162.18.1.el9_1.x86_64,,527650760,,6.0.2-115,113-D67301-059,NA,NA,MI200,gfx90a,16,8192,104,4,8,64,1024,32,1700,1600,1700,1600,32,32,56,4,1638.4,1
1 workload_name command ip_blocks timestamp version hostname cpu_model sbios linux_distro linux_kernel_version amd_gpu_kernel_version cpu_memory gpu_memory rocm_version vbios compute_partition memory_partition gpu_model gpu_arch gpu_l1 gpu_l2 cu_per_gpu simd_per_cu se_per_gpu wave_size workgroup_max_size max_waves_per_cu max_sclk max_mclk cur_sclk cur_mclk total_l2_chan lds_banks_per_cu sqc_per_gpu pipes_per_gpu hbm_bw num_xcd
2 dispatch_inv ./sample/vcopy -n 1048576 -b 256 -i 3 ./tests/vcopy -n 1048576 -b 256 -i 3 SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF|roofline Thu 07 Mar 2024 01:59:41 PM (CST) Thu 21 Mar 2024 04:02:53 PM (CDT) 2 t007-002.hpcfund AMD EPYC 7V13 64-Core Processor American Megatrends Inc.0602 Rocky Linux 9.1 (Blue Onyx) 5.14.0-162.18.1.el9_1.x86_64 527650760 6.0.2-115 113-D67301-059 NA NA MI200 gfx90a 16 8192 104 4 8 64 1024 32 1700 1600 1700 1600 32 32 56 4 1638.4 1
tests/workloads/dispatch_inv/MI200/timestamps.csv
+1 -1
View File
@@ -1,2 +1,2 @@
Dispatch_ID,Kernel_Name,GPU_ID,queue-id,queue-index,pid,tid,Grid_Size,Workgroup_Size,LDS_Per_Workgroup,Scratch_Per_Workitem,Arch_VGPR,Accum_VGPR,SGPR,wave_size,sig,obj,DispatchNs,Start_Timestamp,End_Timestamp,CompleteNs
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,253003,253003,1048576,256,0,0,8,0,16,64,0x0,0x7f70cdd84ec0,197327446918845,197327446943430,197327446964070,197327446982486
0,"vecCopy(double*, double*, double*, int, int) [clone .kd]",2,0,0,4075570,4075570,1048576,256,0,0,8,0,16,64,0x0,0x7ff93b844ec0,1410719177547494,1410719177571898,1410719177592538,1410719177604342
1 Dispatch_ID Kernel_Name GPU_ID queue-id queue-index pid tid Grid_Size Workgroup_Size LDS_Per_Workgroup Scratch_Per_Workitem Arch_VGPR Accum_VGPR SGPR wave_size sig obj DispatchNs Start_Timestamp End_Timestamp CompleteNs
2 0 vecCopy(double*, double*, double*, int, int) [clone .kd] 2 0 0 253003 4075570 253003 4075570 1048576 256 0 0 8 0 16 64 0x0 0x7f70cdd84ec0 0x7ff93b844ec0 197327446918845 1410719177547494 197327446943430 1410719177571898 197327446964070 1410719177592538 197327446982486 1410719177604342