[rocprofiler-compute] [Bugfix] Fix analysis not working with rocpd (#704)

* fix rocpd roofline

* Improve rocpd test by using dynamic workload folder

* bugfix

* fix ruff format
Этот коммит содержится в:
vedithal-amd
2025-08-25 11:46:55 -04:00
коммит произвёл GitHub
родитель b645010655
Коммит 9a02dae75f
25 изменённых файлов: 10 добавлений и 19143 удалений
+3 -4
Просмотреть файл
@@ -693,6 +693,9 @@ class Roofline:
console_error("roofline", "{} does not exist".format(pmc_perf_csv))
t_df = OrderedDict()
t_df["pmc_perf"] = pd.read_csv(pmc_perf_csv)
profiling_config = file_io.load_profiling_config(self.__args.path[0][0])
if profiling_config.get("format_rocprof_output") == "rocpd":
t_df["pmc_perf"] = rocpd_data.process_rocpd_csv(t_df["pmc_perf"])
self.__ai_data = calc_ai_profile(
self.__mspec, self.__run_parameters["sort_type"], t_df
@@ -715,10 +718,6 @@ class Roofline:
self.__run_parameters["mem_level"].remove("vL1D")
self.__run_parameters["mem_level"].append("L1")
profiling_config = file_io.load_profiling_config(self.__args.path[0][0])
if profiling_config.get("format_rocprof_output") == "rocpd":
t_df["pmc_perf"] = rocpd_data.process_rocpd_csv(t_df["pmc_perf"])
color_scheme = {
"HBM": "blue+",
"L2": "green+",
+1 -6
Просмотреть файл
@@ -39,7 +39,6 @@ indirs = [
"tests/workloads/vcopy/MI200",
"tests/workloads/vcopy/MI300A_A1",
"tests/workloads/vcopy/MI300X_A1",
"tests/workloads/vcopy/MI300X_A1_rocpd",
"tests/workloads/vcopy/MI350",
]
@@ -350,11 +349,7 @@ def test_dispatch_5(binary_handler_analyze_rocprof_compute):
@pytest.mark.misc
def test_gpu_ids(binary_handler_analyze_rocprof_compute):
for dir in indirs:
# if dir.endswith("MI350") or dir.endswith("MI300X_A1_rocpd"):
if dir in (
"tests/workloads/vcopy/MI350",
"tests/workloads/vcopy/MI300X_A1_rocpd",
):
if dir == "tests/workloads/vcopy/MI350":
gpu_id = "0"
else:
gpu_id = "2"
-9
Просмотреть файл
@@ -46,15 +46,6 @@ def test_analyze_vcopy_MI200(binary_handler_analyze_rocprof_compute):
assert code == 0
def test_analyze_vcopy_MI300_rocpd(binary_handler_analyze_rocprof_compute):
code = binary_handler_analyze_rocprof_compute([
"analyze",
"--path",
"tests/workloads/vcopy/MI300X_A1_rocpd",
])
assert code == 0
def test_analyze_ipblocks_TCP_MI300X_A1(binary_handler_analyze_rocprof_compute):
code = binary_handler_analyze_rocprof_compute([
"analyze",
+6 -1
Просмотреть файл
@@ -593,7 +593,9 @@ def test_path(binary_handler_profile_rocprof_compute):
@pytest.mark.misc
def test_path_rocpd(binary_handler_profile_rocprof_compute):
def test_path_rocpd(
binary_handler_profile_rocprof_compute, binary_handler_analyze_rocprof_compute
):
workload_dir = test_utils.get_output_dir()
options = ["--format-rocprof-output", "rocpd"]
binary_handler_profile_rocprof_compute(config, workload_dir, options)
@@ -604,6 +606,9 @@ def test_path_rocpd(binary_handler_profile_rocprof_compute):
)
assert test_utils.check_file_pattern("Counter_Name", f"{workload_dir}/pmc_perf.csv")
code = binary_handler_analyze_rocprof_compute(["analyze", "--path", workload_dir])
assert code == 0
test_utils.clean_output_dir(config["cleanup"], workload_dir)
-399
Просмотреть файл
@@ -1,399 +0,0 @@
[profiling] pre-processing using rocprofv3 profiler
[gen_sysinfo]
Incomplete class definition for gfx942. Expecting populated max_mclk but detected None.
Incomplete class definition for gfx942. Expecting populated cur_mclk but detected None.
Missing specs fields for gfx942
starting "run_profiling" and about to start rocprof's workload
[profiling] performing profiling using rocprofv3 profiler
Rocprofiler-Compute version: 3.2.0
Profiler choice: rocprofv3
Path: /app/workloads/vcopy/MI300X_A1
Target: MI300X_A1
Command: sample/vcopy -n 1048576 -b 256 -i 3
Kernel Selection: None
Dispatch Selection: None
Hardware Blocks: []
Report Sections: []
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Collecting Performance Counters
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[Run 1/12][Approximate profiling time left: pending first measurement...]
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/SQ_IFETCH_LEVEL.txt
pmc file: SQ_IFETCH_LEVEL.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_w2ranh0p
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/SQ_IFETCH_LEVEL.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/SQ_IFETCH_LEVEL.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:25.713947 140172743004992 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.166549 sec
|-> [rocprofv3] W20250710 21:33:25.714215 140172743004992 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:25.941891 140172743004992 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:26.335688 140172743004992 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.621474 sec
|-> [rocprofv3] W20250710 21:33:26.351137 140172743004992 generateRocpd.cpp:580] writing SQL database for process 3554 on node 3224294684
|-> [rocprofv3] E20250710 21:33:26.351912 140172743004992 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3554_results.db (UUID=00031de8-fbce-7bce-9912-6a8f9645ae7f)
|-> [rocprofv3] W20250710 21:33:26.391902 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.014565 sec
|-> [rocprofv3] W20250710 21:33:26.392284 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000363 sec
|-> [rocprofv3] W20250710 21:33:26.393046 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000752 sec
|-> [rocprofv3] W20250710 21:33:26.406145 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012446 sec
|-> [rocprofv3] W20250710 21:33:27.791204 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 1.385042 sec
|-> [rocprofv3] W20250710 21:33:27.792304 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001065 sec
|-> [rocprofv3] W20250710 21:33:27.792315 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:27.798288 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005969 sec
|-> [rocprofv3] W20250710 21:33:27.827005 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.028702 sec
|-> [rocprofv3] W20250710 21:33:27.827022 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:27.827026 140172743004992 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:27.827218 140172743004992 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000190 sec
|-> [rocprofv3] W20250710 21:33:27.827762 140172743004992 simple_timer.cpp:55] SQLite3 generation :: total :: 1.476625 sec
|-> [rocprofv3] W20250710 21:33:27.832297 140172743004992 simple_timer.cpp:55] [rocprofv3] tool finalization :: 1.495785 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/SQ_IFETCH_LEVEL.txt, the time it takes was 0 m 2.7194788455963135 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/SQ_IFETCH_LEVEL.txt is 0 m 3.8476221561431885 sec
[Run 2/12][Approximate profiling time left: 38 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_LDS.txt
pmc file: SQ_INST_LEVEL_LDS.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_wp1tt6p6
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_LDS.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_LDS.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:29.552773 140238077793088 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.168769 sec
|-> [rocprofv3] W20250710 21:33:29.553044 140238077793088 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:29.736798 140238077793088 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:30.154960 140238077793088 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.601916 sec
|-> [rocprofv3] W20250710 21:33:30.170439 140238077793088 generateRocpd.cpp:580] writing SQL database for process 3561 on node 3224294684
|-> [rocprofv3] E20250710 21:33:30.171201 140238077793088 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3561_results.db (UUID=00031de9-0ace-7ace-9f47-5cc55ac49931)
|-> [rocprofv3] W20250710 21:33:30.211098 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.014766 sec
|-> [rocprofv3] W20250710 21:33:30.211463 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000347 sec
|-> [rocprofv3] W20250710 21:33:30.212224 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000751 sec
|-> [rocprofv3] W20250710 21:33:30.224678 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011904 sec
|-> [rocprofv3] W20250710 21:33:31.620420 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 1.395725 sec
|-> [rocprofv3] W20250710 21:33:31.621558 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001104 sec
|-> [rocprofv3] W20250710 21:33:31.621569 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:31.627417 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005843 sec
|-> [rocprofv3] W20250710 21:33:31.658664 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.031232 sec
|-> [rocprofv3] W20250710 21:33:31.658684 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:31.658688 140238077793088 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:31.658887 140238077793088 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000198 sec
|-> [rocprofv3] W20250710 21:33:31.659456 140238077793088 simple_timer.cpp:55] SQLite3 generation :: total :: 1.489016 sec
|-> [rocprofv3] W20250710 21:33:31.664027 140238077793088 simple_timer.cpp:55] [rocprofv3] tool finalization :: 1.508410 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_LDS.txt, the time it takes was 0 m 2.7174251079559326 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_LDS.txt is 0 m 3.833453416824341 sec
[Run 3/12][Approximate profiling time left: 34 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_SMEM.txt
pmc file: SQ_INST_LEVEL_SMEM.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_dl9hsbp8
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_SMEM.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_SMEM.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:33.415380 140711413728064 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.170888 sec
|-> [rocprofv3] W20250710 21:33:33.415655 140711413728064 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:33.649099 140711413728064 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:34.063816 140711413728064 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.648161 sec
|-> [rocprofv3] W20250710 21:33:34.079514 140711413728064 generateRocpd.cpp:580] writing SQL database for process 3568 on node 3224294684
|-> [rocprofv3] E20250710 21:33:34.080280 140711413728064 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3568_results.db (UUID=00031de9-19e2-79e2-befc-6c52d81fce09)
|-> [rocprofv3] W20250710 21:33:34.120796 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.014830 sec
|-> [rocprofv3] W20250710 21:33:34.121112 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000299 sec
|-> [rocprofv3] W20250710 21:33:34.121859 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000738 sec
|-> [rocprofv3] W20250710 21:33:34.134532 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012161 sec
|-> [rocprofv3] W20250710 21:33:35.536345 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 1.401796 sec
|-> [rocprofv3] W20250710 21:33:35.537554 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001175 sec
|-> [rocprofv3] W20250710 21:33:35.537565 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:35.543474 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005905 sec
|-> [rocprofv3] W20250710 21:33:35.574172 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.030682 sec
|-> [rocprofv3] W20250710 21:33:35.574191 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:35.574194 140711413728064 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:35.574389 140711413728064 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000193 sec
|-> [rocprofv3] W20250710 21:33:35.574956 140711413728064 simple_timer.cpp:55] SQLite3 generation :: total :: 1.495442 sec
|-> [rocprofv3] W20250710 21:33:35.579445 140711413728064 simple_timer.cpp:55] [rocprofv3] tool finalization :: 1.514702 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_SMEM.txt, the time it takes was 0 m 2.7634034156799316 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_SMEM.txt is 0 m 3.9122095108032227 sec
[Run 4/12][Approximate profiling time left: 30 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_VMEM.txt
pmc file: SQ_INST_LEVEL_VMEM.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_20mmgyh1
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_VMEM.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_VMEM.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:36.918470 140619354510144 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.104106 sec
|-> [rocprofv3] W20250710 21:33:36.918753 140619354510144 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:37.102945 140619354510144 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:37.299363 140619354510144 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.380610 sec
|-> [rocprofv3] W20250710 21:33:37.314710 140619354510144 generateRocpd.cpp:580] writing SQL database for process 3575 on node 3224294684
|-> [rocprofv3] E20250710 21:33:37.315474 140619354510144 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3575_results.db (UUID=00031de9-27d0-77d0-99c1-6bed5f3fcae1)
|-> [rocprofv3] W20250710 21:33:37.350755 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010948 sec
|-> [rocprofv3] W20250710 21:33:37.351102 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000328 sec
|-> [rocprofv3] W20250710 21:33:37.351614 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000503 sec
|-> [rocprofv3] W20250710 21:33:37.364570 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012404 sec
|-> [rocprofv3] W20250710 21:33:38.033478 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.668889 sec
|-> [rocprofv3] W20250710 21:33:38.034553 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001043 sec
|-> [rocprofv3] W20250710 21:33:38.034565 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:38.040348 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005778 sec
|-> [rocprofv3] W20250710 21:33:38.053614 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.013255 sec
|-> [rocprofv3] W20250710 21:33:38.053627 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:38.053630 140619354510144 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:38.053815 140619354510144 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000183 sec
|-> [rocprofv3] W20250710 21:33:38.054183 140619354510144 simple_timer.cpp:55] SQLite3 generation :: total :: 0.739472 sec
|-> [rocprofv3] W20250710 21:33:38.057863 140619354510144 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.757852 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_VMEM.txt, the time it takes was 0 m 1.6674315929412842 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/SQ_INST_LEVEL_VMEM.txt is 0 m 2.43190336227417 sec
[Run 5/12][Approximate profiling time left: 24 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/SQ_LEVEL_WAVES.txt
pmc file: SQ_LEVEL_WAVES.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_4_rrldsg
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/SQ_LEVEL_WAVES.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/SQ_LEVEL_WAVES.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:39.561464 139705878659904 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.135631 sec
|-> [rocprofv3] W20250710 21:33:39.561722 139705878659904 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:39.744744 139705878659904 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:40.044125 139705878659904 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.482403 sec
|-> [rocprofv3] W20250710 21:33:40.059697 139705878659904 generateRocpd.cpp:580] writing SQL database for process 3582 on node 3224294684
|-> [rocprofv3] E20250710 21:33:40.060465 139705878659904 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3582_results.db (UUID=00031de9-3205-7205-869c-ec5f02149d7e)
|-> [rocprofv3] W20250710 21:33:40.097027 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.012617 sec
|-> [rocprofv3] W20250710 21:33:40.097312 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000266 sec
|-> [rocprofv3] W20250710 21:33:40.097962 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000641 sec
|-> [rocprofv3] W20250710 21:33:40.110168 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011707 sec
|-> [rocprofv3] W20250710 21:33:41.116521 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 1.006337 sec
|-> [rocprofv3] W20250710 21:33:41.117628 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001074 sec
|-> [rocprofv3] W20250710 21:33:41.117640 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:41.123520 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005876 sec
|-> [rocprofv3] W20250710 21:33:41.143166 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.019632 sec
|-> [rocprofv3] W20250710 21:33:41.143182 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:41.143186 139705878659904 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:41.143382 139705878659904 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000194 sec
|-> [rocprofv3] W20250710 21:33:41.143922 139705878659904 simple_timer.cpp:55] SQLite3 generation :: total :: 1.084224 sec
|-> [rocprofv3] W20250710 21:33:41.148481 139705878659904 simple_timer.cpp:55] [rocprofv3] tool finalization :: 1.103637 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/SQ_LEVEL_WAVES.txt, the time it takes was 0 m 2.1514015197753906 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/SQ_LEVEL_WAVES.txt is 0 m 3.113880157470703 sec
[Run 6/12][Approximate profiling time left: 20 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_0.txt
pmc file: pmc_perf_0.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_b7cf79sp
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_0.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_0.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:42.470522 139873853399872 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.104173 sec
|-> [rocprofv3] W20250710 21:33:42.470776 139873853399872 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:42.655197 139873853399872 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:42.849277 139873853399872 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.378502 sec
|-> [rocprofv3] W20250710 21:33:42.865054 139873853399872 generateRocpd.cpp:580] writing SQL database for process 3589 on node 3224294684
|-> [rocprofv3] E20250710 21:33:42.865845 139873853399872 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3589_results.db (UUID=00031de9-3d82-7d82-96dc-2dcf0a0631dc)
|-> [rocprofv3] W20250710 21:33:42.903272 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.011102 sec
|-> [rocprofv3] W20250710 21:33:42.903557 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000267 sec
|-> [rocprofv3] W20250710 21:33:42.904080 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000512 sec
|-> [rocprofv3] W20250710 21:33:42.916637 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012050 sec
|-> [rocprofv3] W20250710 21:33:43.580464 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.663812 sec
|-> [rocprofv3] W20250710 21:33:43.581638 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001143 sec
|-> [rocprofv3] W20250710 21:33:43.581649 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:43.587455 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005801 sec
|-> [rocprofv3] W20250710 21:33:43.599109 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011640 sec
|-> [rocprofv3] W20250710 21:33:43.599125 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:43.599128 139873853399872 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:43.599317 139873853399872 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000186 sec
|-> [rocprofv3] W20250710 21:33:43.599679 139873853399872 simple_timer.cpp:55] SQLite3 generation :: total :: 0.734625 sec
|-> [rocprofv3] W20250710 21:33:43.603395 139873853399872 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.753253 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_0.txt, the time it takes was 0 m 1.6624979972839355 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_0.txt is 0 m 2.4318082332611084 sec
[Run 7/12][Approximate profiling time left: 16 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_1.txt
pmc file: pmc_perf_1.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_i5zi8g1c
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_1.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_1.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:44.921375 140329810051904 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.104611 sec
|-> [rocprofv3] W20250710 21:33:44.921646 140329810051904 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:45.127446 140329810051904 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:45.322745 140329810051904 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.401100 sec
|-> [rocprofv3] W20250710 21:33:45.338315 140329810051904 generateRocpd.cpp:580] writing SQL database for process 3596 on node 3224294684
|-> [rocprofv3] E20250710 21:33:45.339069 140329810051904 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3596_results.db (UUID=00031de9-4713-7713-aca4-263b5bf68657)
|-> [rocprofv3] W20250710 21:33:45.374491 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010790 sec
|-> [rocprofv3] W20250710 21:33:45.374878 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000369 sec
|-> [rocprofv3] W20250710 21:33:45.375436 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000547 sec
|-> [rocprofv3] W20250710 21:33:45.388444 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012406 sec
|-> [rocprofv3] W20250710 21:33:46.060238 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.671777 sec
|-> [rocprofv3] W20250710 21:33:46.061333 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001063 sec
|-> [rocprofv3] W20250710 21:33:46.061344 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:46.067223 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005874 sec
|-> [rocprofv3] W20250710 21:33:46.078638 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011402 sec
|-> [rocprofv3] W20250710 21:33:46.078650 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:46.078653 140329810051904 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:46.078838 140329810051904 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000182 sec
|-> [rocprofv3] W20250710 21:33:46.079192 140329810051904 simple_timer.cpp:55] SQLite3 generation :: total :: 0.740877 sec
|-> [rocprofv3] W20250710 21:33:46.082891 140329810051904 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.759142 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_1.txt, the time it takes was 0 m 1.6950774192810059 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_1.txt is 0 m 2.480154514312744 sec
[Run 8/12][Approximate profiling time left: 12 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_2.txt
pmc file: pmc_perf_2.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_tv_uqwoh
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_2.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_2.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:47.392035 140613043073856 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.103348 sec
|-> [rocprofv3] W20250710 21:33:47.392296 140613043073856 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:47.575114 140613043073856 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:47.769792 140613043073856 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.377496 sec
|-> [rocprofv3] W20250710 21:33:47.785435 140613043073856 generateRocpd.cpp:580] writing SQL database for process 3603 on node 3224294684
|-> [rocprofv3] E20250710 21:33:47.786191 140613043073856 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3603_results.db (UUID=00031de9-50bd-70bd-ae26-e3dd40c28d4c)
|-> [rocprofv3] W20250710 21:33:47.820358 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010878 sec
|-> [rocprofv3] W20250710 21:33:47.820688 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000313 sec
|-> [rocprofv3] W20250710 21:33:47.821208 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000510 sec
|-> [rocprofv3] W20250710 21:33:47.833615 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011855 sec
|-> [rocprofv3] W20250710 21:33:48.485409 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.651779 sec
|-> [rocprofv3] W20250710 21:33:48.486280 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.000839 sec
|-> [rocprofv3] W20250710 21:33:48.486291 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:48.492148 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005851 sec
|-> [rocprofv3] W20250710 21:33:48.503742 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011578 sec
|-> [rocprofv3] W20250710 21:33:48.503758 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:48.503762 140613043073856 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:48.503952 140613043073856 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000188 sec
|-> [rocprofv3] W20250710 21:33:48.504347 140613043073856 simple_timer.cpp:55] SQLite3 generation :: total :: 0.718912 sec
|-> [rocprofv3] W20250710 21:33:48.508036 140613043073856 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.737530 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_2.txt, the time it takes was 0 m 1.6442956924438477 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_2.txt is 0 m 2.423243761062622 sec
[Run 9/12][Approximate profiling time left: 9 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_3.txt
pmc file: pmc_perf_3.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_d6edxy7z
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_3.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_3.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:49.825653 139970175718208 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.103156 sec
|-> [rocprofv3] W20250710 21:33:49.825903 139970175718208 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:50.011954 139970175718208 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:50.207081 139970175718208 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.381179 sec
|-> [rocprofv3] W20250710 21:33:50.222580 139970175718208 generateRocpd.cpp:580] writing SQL database for process 3610 on node 3224294684
|-> [rocprofv3] E20250710 21:33:50.223351 139970175718208 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3610_results.db (UUID=00031de9-5a3e-7a3e-8845-b724df0c5328)
|-> [rocprofv3] W20250710 21:33:50.277220 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.030272 sec
|-> [rocprofv3] W20250710 21:33:50.277672 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000428 sec
|-> [rocprofv3] W20250710 21:33:50.278189 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000509 sec
|-> [rocprofv3] W20250710 21:33:50.290370 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011683 sec
|-> [rocprofv3] W20250710 21:33:50.954577 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.664191 sec
|-> [rocprofv3] W20250710 21:33:50.955670 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.001063 sec
|-> [rocprofv3] W20250710 21:33:50.955682 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:50.961629 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005942 sec
|-> [rocprofv3] W20250710 21:33:50.973364 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011722 sec
|-> [rocprofv3] W20250710 21:33:50.973378 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:50.973381 139970175718208 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:50.973569 139970175718208 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000187 sec
|-> [rocprofv3] W20250710 21:33:50.973960 139970175718208 simple_timer.cpp:55] SQLite3 generation :: total :: 0.751380 sec
|-> [rocprofv3] W20250710 21:33:50.977619 139970175718208 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.769843 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_3.txt, the time it takes was 0 m 1.6794917583465576 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_3.txt is 0 m 2.4662797451019287 sec
[Run 10/12][Approximate profiling time left: 5 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_4.txt
pmc file: pmc_perf_4.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_7pmy1aeg
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_4.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_4.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:52.291220 140318446199616 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.102577 sec
|-> [rocprofv3] W20250710 21:33:52.291470 140318446199616 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:52.492822 140318446199616 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:52.684795 140318446199616 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.393325 sec
|-> [rocprofv3] W20250710 21:33:52.700354 140318446199616 generateRocpd.cpp:580] writing SQL database for process 3617 on node 3224294684
|-> [rocprofv3] E20250710 21:33:52.701113 140318446199616 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3617_results.db (UUID=00031de9-63e2-73e2-b1de-4720cbe72cf4)
|-> [rocprofv3] W20250710 21:33:52.757125 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010891 sec
|-> [rocprofv3] W20250710 21:33:52.757435 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000290 sec
|-> [rocprofv3] W20250710 21:33:52.757948 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000505 sec
|-> [rocprofv3] W20250710 21:33:52.771027 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.012490 sec
|-> [rocprofv3] W20250710 21:33:53.432385 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.661342 sec
|-> [rocprofv3] W20250710 21:33:53.433363 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.000947 sec
|-> [rocprofv3] W20250710 21:33:53.433374 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:53.439104 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005725 sec
|-> [rocprofv3] W20250710 21:33:53.450240 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011123 sec
|-> [rocprofv3] W20250710 21:33:53.450253 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:53.450257 140318446199616 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:53.450440 140318446199616 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000181 sec
|-> [rocprofv3] W20250710 21:33:53.450806 140318446199616 simple_timer.cpp:55] SQLite3 generation :: total :: 0.750452 sec
|-> [rocprofv3] W20250710 21:33:53.454365 140318446199616 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.768517 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_4.txt, the time it takes was 0 m 1.689640760421753 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_4.txt is 0 m 2.470271110534668 sec
[Run 11/12][Approximate profiling time left: 2 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_5.txt
pmc file: pmc_perf_5.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_i1z59usc
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_5.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_5.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:54.741603 140680114004800 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.103912 sec
|-> [rocprofv3] W20250710 21:33:54.741856 140680114004800 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:54.934468 140680114004800 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:55.127192 140680114004800 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.385336 sec
|-> [rocprofv3] W20250710 21:33:55.142986 140680114004800 generateRocpd.cpp:580] writing SQL database for process 3624 on node 3224294684
|-> [rocprofv3] E20250710 21:33:55.143809 140680114004800 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3624_results.db (UUID=00031de9-6d70-7d70-801f-957d2f8ca86a)
|-> [rocprofv3] W20250710 21:33:55.177382 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010768 sec
|-> [rocprofv3] W20250710 21:33:55.177741 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000341 sec
|-> [rocprofv3] W20250710 21:33:55.178266 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000516 sec
|-> [rocprofv3] W20250710 21:33:55.190617 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011831 sec
|-> [rocprofv3] W20250710 21:33:55.867254 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.676620 sec
|-> [rocprofv3] W20250710 21:33:55.868249 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.000964 sec
|-> [rocprofv3] W20250710 21:33:55.868260 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:55.874104 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005839 sec
|-> [rocprofv3] W20250710 21:33:55.885401 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.011285 sec
|-> [rocprofv3] W20250710 21:33:55.885413 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:55.885417 140680114004800 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:55.885599 140680114004800 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000180 sec
|-> [rocprofv3] W20250710 21:33:55.885979 140680114004800 simple_timer.cpp:55] SQLite3 generation :: total :: 0.742993 sec
|-> [rocprofv3] W20250710 21:33:55.889649 140680114004800 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.761385 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_5.txt, the time it takes was 0 m 1.6850097179412842 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_5.txt is 0 m 2.4484713077545166 sec
[Run 12/12][Approximate profiling time left: 0 seconds]...
[profiling] Current input file: /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_6.txt
pmc file: pmc_perf_6.txt
Adding env var for counter definitions: ROCPROFILER_METRICS_PATH=/tmp/rocprof_counter_defs_8km5kicc
rocprof command: ['rocprofv3', '-A', 'absolute', '-i', '/app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_6.txt', '-d', '/app/workloads/vcopy/MI300X_A1/out', '--kernel-trace', '--output-format', 'rocpd', '--', 'sample/vcopy', '-n', '1048576', '-b', '256', '-i', '3']
[subprocess] Running: rocprofv3 -A absolute -i /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_6.txt -d /app/workloads/vcopy/MI300X_A1/out --kernel-trace --output-format rocpd -- sample/vcopy -n 1048576 -b 256 -i 3
|-> [rocprofv3] W20250710 21:33:57.212530 139907727749952 simple_timer.cpp:55] [rocprofv3] tool initialization :: 0.103224 sec
|-> [rocprofv3] W20250710 21:33:57.212781 139907727749952 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:57.411536 139907727749952 tool.cpp:2150] HSA version 8.18.0 initialized (instance=0)
|-> [rocprofv3] W20250710 21:33:57.602202 139907727749952 simple_timer.cpp:55] [rocprofv3] 'sample/vcopy -n 1048576 -b 256 -i 3' :: 0.389421 sec
|-> [rocprofv3] W20250710 21:33:57.617817 139907727749952 generateRocpd.cpp:580] writing SQL database for process 3631 on node 3224294684
|-> [rocprofv3] E20250710 21:33:57.618592 139907727749952 generateRocpd.cpp:603] Opened result file: /app/workloads/vcopy/MI300X_A1/out/pmc_1/7f6eaef84eaf/3631_results.db (UUID=00031de9-771a-771a-95de-02702fbe6904)
|-> [rocprofv3] W20250710 21:33:57.682260 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_string :: 0.010946 sec
|-> [rocprofv3] W20250710 21:33:57.682562 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_node :: 0.000283 sec
|-> [rocprofv3] W20250710 21:33:57.683093 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_process :: 0.000521 sec
|-> [rocprofv3] W20250710 21:33:57.695274 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_agent :: 0.011651 sec
|-> [rocprofv3] W20250710 21:33:58.358124 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_info_pmc :: 0.662833 sec
|-> [rocprofv3] W20250710 21:33:58.358983 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd kernel info :: 0.000831 sec
|-> [rocprofv3] W20250710 21:33:58.358994 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_region :: 0.000002 sec
|-> [rocprofv3] W20250710 21:33:58.364913 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_kernel_dispatch :: 0.005914 sec
|-> [rocprofv3] W20250710 21:33:58.372546 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_pmc_event :: 0.007617 sec
|-> [rocprofv3] W20250710 21:33:58.372562 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_copy :: 0.000000 sec
|-> [rocprofv3] W20250710 21:33:58.372566 139907727749952 simple_timer.cpp:55] SQLite3 generation :: rocpd_memory_allocate :: 0.000001 sec
|-> [rocprofv3] W20250710 21:33:58.372750 139907727749952 simple_timer.cpp:55] SQLite3 generation :: SQL indexing :: 0.000182 sec
|-> [rocprofv3] W20250710 21:33:58.373145 139907727749952 simple_timer.cpp:55] SQLite3 generation :: total :: 0.755328 sec
|-> [rocprofv3] W20250710 21:33:58.376827 139907727749952 simple_timer.cpp:55] [rocprofv3] tool finalization :: 0.773574 sec
|-> [rocprofv3] vcopy testing on GCD 0
|-> [rocprofv3] Finished allocating vectors on the CPU
Finishing subprocess of fname /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_6.txt, the time it takes was 0 m 1.6893463134765625 sec
The time of run_prof of /app/workloads/vcopy/MI300X_A1/perfmon/pmc_perf_6.txt is 0 m 2.4789364337921143 sec
finished "run_profiling" and finished rocprof's workload, time taken was 0 m 34.34300780296326 sec
[profiling] performing post-processing using rocprofv3 profiler
Created file: /app/workloads/vcopy/MI300X_A1/pmc_perf.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_6.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_SQ_IFETCH_LEVEL.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_2.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_SQ_INST_LEVEL_VMEM.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_SQ_INST_LEVEL_LDS.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_SQ_INST_LEVEL_SMEM.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_5.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_3.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_0.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_4.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_pmc_perf_1.csv
Deleted file: /app/workloads/vcopy/MI300X_A1/results_SQ_LEVEL_WAVES.csv
time taken for "post_processing" was 0 seconds
[profiling] perform SoC post processing for gfx942
[roofline] Skipping roofline
Различия файлов скрыты, потому что одна или несколько строк слишком длинны
Разница между файлами не показана из-за своего большого размера Загрузить разницу
Различия файлов скрыты, потому что одна или несколько строк слишком длинны
Разница между файлами не показана из-за своего большого размера Загрузить разницу
Различия файлов скрыты, потому что одна или несколько строк слишком длинны
Разница между файлами не показана из-за своего большого размера Загрузить разницу
@@ -1,5 +0,0 @@
pmc: SQ_INST_LEVEL_VMEM SQ_INST_LEVEL_VMEM_ACCUM SQ_LDS_UNALIGNED_STALL SQ_INSTS_VALU_MFMA_F32 SQ_INSTS SQ_WAIT_ANY SQC_ICACHE_MISSES SQ_INSTS_VALU_MFMA_BF16 TA_TOTAL_WAVEFRONTS_sum TA_BUFFER_READ_WAVEFRONTS_sum TCP_TCC_RW_READ_REQ_sum TCP_UTCL1_TRANSLATION_MISS_sum TCP_TCC_CC_READ_REQ_sum TCP_UTCL1_PERMISSION_MISS_sum TCC_EA0_RDREQ_LEVEL_sum TCC_ALL_TC_OP_INV_EVICT_sum TCC_EA0_WRREQ_64B_sum TCC_MISS_sum CPC_ME1_DC0_SPI_BUSY CPC_CPC_STAT_BUSY
gpu:
range:
kernel:
@@ -1,10 +0,0 @@
rocprofiler-sdk:
counters-schema-version: 1
counters:
- name: SQ_INST_LEVEL_VMEM_ACCUM
description: SQ_INST_LEVEL_VMEM accumulation
properties: []
definitions:
- architectures:
- gfx942
expression: accumulate(SQ_INST_LEVEL_VMEM, HIGH_RES)
@@ -1,5 +0,0 @@
pmc: SQ_LEVEL_WAVES SQ_LEVEL_WAVES_ACCUM SQ_INSTS_VMEM SQC_TC_DATA_READ_REQ SQ_INSTS_VALU_MFMA_F64 SQ_INSTS_VALU_MFMA_F16 SQ_ACTIVE_INST_FLAT SQC_DCACHE_HITS TA_BUFFER_ATOMIC_WAVEFRONTS_sum TA_ADDR_STALLED_BY_TC_CYCLES_sum TD_SPI_STALL_sum TCP_TCC_UC_ATOMIC_REQ_sum TCP_ATOMIC_TAGCONFLICT_STALL_CYCLES_sum TCP_TOTAL_READ_sum TCP_TOTAL_CACHE_ACCESSES_sum TCC_WRITE[99] TCC_REQ[28] TCC_REQ[31] TCC_WRITE[100] TCC_WRITE[112] TCC_WRITE[75] TCC_WRITE[114] TCC_EA0_WRREQ_DRAM_sum TCC_REQ[9] TCC_WRITE[39] TCC_REQ[122] TCC_WRITE[76] TCC_NORMAL_EVICT_sum TCC_WRITE[66] TCC_REQ[90] TCC_WRITE[61] TCC_REQ[45] TCC_WRITE[74] TCC_WRITE[108] TCC_REQ[76] TCC_WRITE[85] TCC_WRITE[111] TCC_REQ[46] TCC_WRITE[33] TCC_WRITE[0] TCC_WRITE[113] TCC_WRITE[86] TCC_REQ[37] TCC_REQ[77] TCC_REQ[13] TCC_WRITE[55] TCC_WRITE[101] TCC_REQ[27] TCC_REQ[89] TCC_WRITE[126] TCC_WRITE[34] TCC_REQ[112] TCC_WRITE[105] TCC_WRITE[2] TCC_REQ[39] TCC_WRITE[31] TCC_REQ[18] TCC_REQ[120] TCC_WRITE[15] TCC_WRITE[77] TCC_WRITE[68] TCC_REQ[51] TCC_WRITE[115] TCC_WRITE[63] TCC_REQ[23] TCC_REQ[32] TCC_REQ[85] TCC_REQ[95] TCC_WRITE[8] TCC_WRITE[98] TCC_WRITE[53] TCC_WRITE[51] TCC_WRITE[67] TCC_WRITE[1] TCC_REQ[110] TCC_REQ[65] TCC_WRITE[7] TCC_REQ[107] TCC_WRITE[89] TCC_WRITE[123] TCC_WRITE[23] TCC_REQ[116] TCC_WRITE[110] TCC_REQ[36] TCC_WRITE[107] TCC_REQ[74] TCC_REQ[12] TCC_WRITE[124] TCC_REQ[71] TCC_WRITE[4] TCC_WRITE[71] TCC_WRITE[103] TCC_WRITE[93] TCC_REQ[6] TCC_WRITE[6] TCC_REQ[96] TCC_REQ[99] TCC_WRITE[37] TCC_WRITE[22] TCC_WRITE[12] TCC_WRITE[122] TCC_REQ[127] TCC_REQ[68] TCC_WRITE[82] TCC_REQ[30] TCC_WRITE[45] TCC_REQ[38] TCC_WRITE[11] TCC_REQ[35] TCC_REQ[126] TCC_REQ[69] TCC_REQ[98] TCC_WRITE[3] TCC_WRITE[38] TCC_REQ[15] TCC_WRITE[49] TCC_WRITE[72] TCC_WRITE[116] TCC_REQ[59] TCC_WRITE[47] TCC_REQ[40] TCC_REQ[78] TCC_WRITE[70] TCC_WRITE[88] TCC_WRITE[17] TCC_WRITE[48] TCC_REQ[123] TCC_WRITE[102] TCC_REQ[53] TCC_REQ[19] TCC_REQ[108] TCC_REQ[11] TCC_REQ[64] TCC_WRITE[95] TCC_REQ[58] TCC_REQ[106] TCC_WRITE[81] TCC_WRITE[32] TCC_REQ[117] TCC_REQ[24] TCC_REQ[82] TCC_WRITE[65] TCC_WRITE[24] TCC_WRITE[19] TCC_WRITE[60] TCC_WRITE[58] TCC_REQ[66] TCC_WRITE[30] TCC_WRITE[118] TCC_REQ[80] TCC_REQ[94] TCC_REQ[20] TCC_REQ[102] TCC_WRITE[5] TCC_REQ[33] TCC_REQ[3] TCC_WRITE[25] TCC_REQ[118] TCC_REQ[52] TCC_REQ[17] TCC_REQ[8] TCC_WRITE[56] TCC_REQ[125] TCC_WRITE[97] TCC_WRITE[90] TCC_WRITE[20] TCC_WRITE[64] TCC_REQ[54] TCC_REQ[4] TCC_WRITE[84] TCC_REQ[29] TCC_REQ[92] TCC_REQ[101] TCC_WRITE[13] TCC_WRITE[96] TCC_REQ[7] TCC_WRITE[80] TCC_WRITE[46] TCC_REQ[62] TCC_WRITE[83] TCC_REQ[49] TCC_REQ[88] TCC_WRITE[104] TCC_REQ[84] TCC_REQ[41] TCC_REQ[44] TCC_REQ[60] TCC_REQ[119] TCC_REQ[0] TCC_REQ[55] TCC_REQ[26] TCC_REQ[81] TCC_WRITE[59] TCC_WRITE[42] TCC_WRITE[119] TCC_REQ[109] TCC_WRITE[50] TCC_WRITE[121] TCC_WRITE[18] TCC_REQ[70] TCC_WRITE[125] TCC_REQ[48] TCC_REQ[16] TCC_REQ[72] TCC_REQ[91] TCC_REQ[22] TCC_WRITE[40] TCC_REQ[75] TCC_WRITE[79] TCC_REQ[93] TCC_WRITE[78] TCC_WRITE[92] TCC_REQ[25] TCC_WRITE[9] TCC_REQ[43] TCC_WRITE[28] TCC_REQ[124] TCC_WRITE[52] TCC_WRITE[27] TCC_WRITE[87] TCC_WRITE[29] TCC_WRITE[73] TCC_WRITE[44] TCC_WRITE[41] TCC_REQ[73] TCC_REQ[111] TCC_REQ[87] TCC_REQ[86] TCC_REQ[5] TCC_WRITE[120] TCC_REQ[114] TCC_REQ[1] TCC_REQ[67] TCC_REQ[42] TCC_WRITE[10] TCC_REQ[2] TCC_WRITE[62] TCC_WRITE[106] TCC_REQ[97] TCC_REQ[57] TCC_WRITE[54] TCC_WRITE[117] TCC_WRITE[109] TCC_WRITE[43] TCC_REQ[115] TCC_REQ[100] TCC_REQ[121] TCC_WRITE[127] TCC_REQ[10] TCC_REQ[61] TCC_REQ[50] TCC_REQ[56] TCC_WRITE[94] TCC_REQ[103] TCC_REQ[21] TCC_WRITE[36] TCC_REQ[47] TCC_WRITE[26] TCC_REQ[79] TCC_WRITE[14] TCC_WRITE[91] TCC_WRITE[35] TCC_WRITE[21] TCC_WRITE[69] TCC_REQ[34] TCC_REQ[14] TCC_WRITE[57] TCC_REQ[105] TCC_REQ[83] TCC_REQ[113] TCC_REQ[63] TCC_REQ[104] TCC_WRITE[16] CPC_CPC_STAT_IDLE CPC_UTCL1_STALL_ON_TRANSLATION CPF_CPF_TCIU_IDLE
gpu:
range:
kernel:
Разница между файлами не показана из-за своего большого размера Загрузить разницу
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_ACTIVE_INST_VMEM SQC_DCACHE_REQ_READ_1 SQ_ACTIVE_INST_ANY SQC_TC_DATA_ATOMIC_REQ SQ_INSTS_LDS SQ_LDS_IDX_ACTIVE SQ_LDS_MEM_VIOLATIONS SQ_INSTS_SENDMSG TA_BUFFER_COALESCED_READ_CYCLES_sum TA_FLAT_WRITE_WAVEFRONTS_sum TCP_GATE_EN2_sum TCP_TCC_WRITE_REQ_sum TCP_UTCL1_REQUEST_sum TCP_TCC_ATOMIC_WITH_RET_REQ_sum TCC_TOO_MANY_EA_WRREQS_STALL_sum TCC_HIT_sum TCC_NORMAL_WRITEBACK_sum TCC_UC_REQ_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_INSTS_VALU_INT64 SQ_INSTS_SALU SQ_INSTS_VALU_MFMA_F8 SQ_INSTS_VALU_MUL_F64 SQ_WAVES SQ_WAVES_RESTORED SQC_DCACHE_MISSES_DUPLICATE SQ_INSTS_VSKIPPED TA_FLAT_READ_WAVEFRONTS_sum TA_BUFFER_WAVEFRONTS_sum TCP_TOTAL_WRITEBACK_INVALIDATES_sum TCP_TOTAL_ACCESSES_sum TCP_TCR_TCP_STALL_CYCLES_sum TCP_TCC_NC_WRITE_REQ_sum TCC_EA0_RDREQ_sum TCC_WRITEBACK_sum TCC_READ_sum TCC_EA0_RD_UNCACHED_32B_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQC_TC_INST_REQ SQ_ACTIVE_INST_LDS SQC_DCACHE_MISSES SQ_BUSY_CU_CYCLES SQ_INSTS_VALU_ADD_F32 SQ_INSTS_VALU_MFMA_I8 SQ_LDS_ADDR_CONFLICT SQ_IFETCH TA_DATA_STALLED_BY_TC_CYCLES_sum TA_TA_BUSY_sum TCP_TCC_RW_ATOMIC_REQ_sum TCP_GATE_EN1_sum TCP_TA_TCP_STATE_READ_sum TCP_WRITE_TAGCONFLICT_STALL_CYCLES_sum TCC_BUSY_sum TCC_BUBBLE_sum TCC_WRITE_sum TCC_STREAMING_REQ_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_INSTS_VALU_FMA_F64 SQ_INSTS_VALU_ADD_F16 SQ_INSTS_VALU_MUL_F32 SQ_INSTS_VALU_MFMA_MOPS_BF16 SQC_DCACHE_REQ_READ_8 SQC_ICACHE_HITS SQ_ACTIVE_INST_VALU SQ_INSTS_VALU_FMA_F16 TCP_TCC_CC_ATOMIC_REQ_sum TCP_TCC_CC_WRITE_REQ_sum TCP_TOTAL_WRITE_sum TCC_EA0_RDREQ_32B_sum TCC_EA0_WR_UNCACHED_32B_sum TCC_NC_REQ_sum TCC_EA0_WRREQ_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_WAIT_INST_ANY SQ_INSTS_GDS SQ_INSTS_VALU_ADD_F64 SQ_INSTS_VALU_MFMA_MOPS_F64 SQ_INSTS_VALU_MFMA_MOPS_F16 SQ_INSTS_VALU_INT32 SQC_DCACHE_REQ_READ_2 SQC_DCACHE_REQ TCC_EA0_ATOMIC_sum TCC_CC_REQ_sum TCC_ATOMIC_sum TCC_PROBE_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_INSTS_MFMA SQ_INSTS_VALU_TRANS_F16 SQ_BUSY_CYCLES SQC_DCACHE_ATOMIC SQ_INSTS_VALU_CVT SQC_TC_DATA_WRITE_REQ SQ_INSTS_VALU_FMA_F32 SQ_INSTS_BRANCH TCC_EA0_ATOMIC_LEVEL_sum TCC_EA0_RDREQ_DRAM_sum TCC_REQ_sum TCC_ALL_TC_OP_WB_WRITEBACK_sum
gpu:
range:
kernel:
-5
Просмотреть файл
@@ -1,5 +0,0 @@
pmc: SQ_INSTS_VALU_MFMA_MOPS_F8 SQ_LDS_ATOMIC_RETURN SQ_INSTS_VALU_MFMA_MOPS_I8 TCC_RW_REQ_sum TCC_EA0_WRREQ_LEVEL_sum
gpu:
range:
kernel:
Разница между файлами не показана из-за своего большого размера Загрузить разницу
-40
Просмотреть файл
@@ -1,40 +0,0 @@
config_dir: /app/src/rocprof_compute_soc/analysis_configs
device: -1
dispatch: null
filter_blocks: {}
format_rocprof_output: rocpd
hip_trace: false
join_type: grid
kernel: null
kernel_names: false
kokkos_trace: false
list_metrics: null
loglevel: 10
mem_level: ALL
mode: profile
name: vcopy
no_roof: true
path: /app/workloads/vcopy/MI300X_A1
pc_sampling_interval: 1048576
pc_sampling_method: stochastic
quiet: false
remaining:
- --
- sample/vcopy
- -n
- '1048576'
- -b
- '256'
- -i
- '3'
retain_rocpd_output: false
rocprofiler_sdk_library_path: /opt/rocm/lib/librocprofiler-sdk.so
roof_only: false
roofline_data_type:
- FP32
sort: kernels
spatial_multiplexing: null
specs: false
subpath: gpu
target: null
verbose: 1
-2
Просмотреть файл
@@ -1,2 +0,0 @@
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_series,gpu_model,gpu_arch,gpu_chip_id,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,num_xcd,num_hbm_channels
vcopy,sample/vcopy -n 1048576 -b 256 -i 3,SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF,Thu Jul 10 21:33:20 2025 (UTC),3,7f6eaef84eaf,AMD EPYC 9354 32-Core Processor,"American Megatrends International, LLC.1.8",Ubuntu 22.04.5 LTS,5.15.0-70-generic,,1584988420,,7.0.0,113-M3000100-103,SPX,NPS1,MI300,MI300X_A1,gfx942,29857,32,4096,304,4,32,64,1024,32,2100,,2100,,128,32,160,4,8,128
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_series gpu_model gpu_arch gpu_chip_id 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 num_xcd num_hbm_channels
2 vcopy sample/vcopy -n 1048576 -b 256 -i 3 SQ|LDS|SQC|TA|TD|TCP|TCC|SPI|CPC|CPF Thu Jul 10 21:33:20 2025 (UTC) 3 7f6eaef84eaf AMD EPYC 9354 32-Core Processor American Megatrends International, LLC.1.8 Ubuntu 22.04.5 LTS 5.15.0-70-generic 1584988420 7.0.0 113-M3000100-103 SPX NPS1 MI300 MI300X_A1 gfx942 29857 32 4096 304 4 32 64 1024 32 2100 2100 128 32 160 4 8 128